diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10962 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.99712,
+  "eval_steps": 500,
+  "global_step": 1560,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00192,
+      "grad_norm": 5.0316033363342285,
+      "learning_rate": 6.41025641025641e-08,
+      "loss": 0.4618,
+      "step": 1
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 4.99949312210083,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 0.483,
+      "step": 2
+    },
+    {
+      "epoch": 0.00576,
+      "grad_norm": 4.812772274017334,
+      "learning_rate": 1.9230769230769234e-07,
+      "loss": 0.4064,
+      "step": 3
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 4.554108142852783,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 0.4284,
+      "step": 4
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 4.5176897048950195,
+      "learning_rate": 3.205128205128205e-07,
+      "loss": 0.4042,
+      "step": 5
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 4.5790114402771,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 0.4145,
+      "step": 6
+    },
+    {
+      "epoch": 0.01344,
+      "grad_norm": 4.345544338226318,
+      "learning_rate": 4.4871794871794876e-07,
+      "loss": 0.4246,
+      "step": 7
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 4.502335071563721,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 0.3834,
+      "step": 8
+    },
+    {
+      "epoch": 0.01728,
+      "grad_norm": 4.529841899871826,
+      "learning_rate": 5.76923076923077e-07,
+      "loss": 0.3825,
+      "step": 9
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 3.8888132572174072,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 0.3867,
+      "step": 10
+    },
+    {
+      "epoch": 0.02112,
+      "grad_norm": 3.8011980056762695,
+      "learning_rate": 7.051282051282052e-07,
+      "loss": 0.3923,
+      "step": 11
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 4.1597580909729,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.4069,
+      "step": 12
+    },
+    {
+      "epoch": 0.02496,
+      "grad_norm": 3.1295435428619385,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.4416,
+      "step": 13
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 3.0165700912475586,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 0.4148,
+      "step": 14
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 2.9410784244537354,
+      "learning_rate": 9.615384615384617e-07,
+      "loss": 0.3897,
+      "step": 15
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 3.0560102462768555,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.4194,
+      "step": 16
+    },
+    {
+      "epoch": 0.03264,
+      "grad_norm": 2.7734267711639404,
+      "learning_rate": 1.0897435897435899e-06,
+      "loss": 0.3597,
+      "step": 17
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 2.4571542739868164,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 0.349,
+      "step": 18
+    },
+    {
+      "epoch": 0.03648,
+      "grad_norm": 2.3121178150177,
+      "learning_rate": 1.217948717948718e-06,
+      "loss": 0.3187,
+      "step": 19
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 2.293428421020508,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 0.3572,
+      "step": 20
+    },
+    {
+      "epoch": 0.04032,
+      "grad_norm": 2.083855390548706,
+      "learning_rate": 1.3461538461538462e-06,
+      "loss": 0.2936,
+      "step": 21
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 2.1424598693847656,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 0.2985,
+      "step": 22
+    },
+    {
+      "epoch": 0.04416,
+      "grad_norm": 1.910278558731079,
+      "learning_rate": 1.4743589743589745e-06,
+      "loss": 0.3498,
+      "step": 23
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 1.3667981624603271,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.2896,
+      "step": 24
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 1.3955644369125366,
+      "learning_rate": 1.602564102564103e-06,
+      "loss": 0.3037,
+      "step": 25
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 1.39174485206604,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.2743,
+      "step": 26
+    },
+    {
+      "epoch": 0.05184,
+      "grad_norm": 1.3554505109786987,
+      "learning_rate": 1.7307692307692308e-06,
+      "loss": 0.2585,
+      "step": 27
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 1.2093051671981812,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 0.2949,
+      "step": 28
+    },
+    {
+      "epoch": 0.05568,
+      "grad_norm": 1.2416698932647705,
+      "learning_rate": 1.8589743589743592e-06,
+      "loss": 0.2928,
+      "step": 29
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 1.0989071130752563,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.2819,
+      "step": 30
+    },
+    {
+      "epoch": 0.05952,
+      "grad_norm": 1.1074273586273193,
+      "learning_rate": 1.987179487179487e-06,
+      "loss": 0.2601,
+      "step": 31
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 1.0452232360839844,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 0.2686,
+      "step": 32
+    },
+    {
+      "epoch": 0.06336,
+      "grad_norm": 0.9162420630455017,
+      "learning_rate": 2.1153846153846155e-06,
+      "loss": 0.2743,
+      "step": 33
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 1.119637370109558,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 0.278,
+      "step": 34
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 0.9774894118309021,
+      "learning_rate": 2.243589743589744e-06,
+      "loss": 0.2777,
+      "step": 35
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 0.9908045530319214,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.2868,
+      "step": 36
+    },
+    {
+      "epoch": 0.07104,
+      "grad_norm": 0.903928279876709,
+      "learning_rate": 2.371794871794872e-06,
+      "loss": 0.2571,
+      "step": 37
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 0.962845504283905,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 0.2757,
+      "step": 38
+    },
+    {
+      "epoch": 0.07488,
+      "grad_norm": 0.8145211338996887,
+      "learning_rate": 2.5e-06,
+      "loss": 0.2654,
+      "step": 39
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.9562869071960449,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 0.2486,
+      "step": 40
+    },
+    {
+      "epoch": 0.07872,
+      "grad_norm": 0.8484691381454468,
+      "learning_rate": 2.6282051282051286e-06,
+      "loss": 0.2655,
+      "step": 41
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 0.8951556086540222,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 0.2766,
+      "step": 42
+    },
+    {
+      "epoch": 0.08256,
+      "grad_norm": 0.8147677183151245,
+      "learning_rate": 2.756410256410257e-06,
+      "loss": 0.2311,
+      "step": 43
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 0.6778334975242615,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 0.2313,
+      "step": 44
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 0.8270885348320007,
+      "learning_rate": 2.8846153846153845e-06,
+      "loss": 0.2652,
+      "step": 45
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 0.7811056971549988,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 0.2983,
+      "step": 46
+    },
+    {
+      "epoch": 0.09024,
+      "grad_norm": 0.7773110270500183,
+      "learning_rate": 3.012820512820513e-06,
+      "loss": 0.24,
+      "step": 47
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 0.8252728581428528,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.2302,
+      "step": 48
+    },
+    {
+      "epoch": 0.09408,
+      "grad_norm": 0.7708638310432434,
+      "learning_rate": 3.141025641025641e-06,
+      "loss": 0.2325,
+      "step": 49
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.7612317204475403,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 0.2701,
+      "step": 50
+    },
+    {
+      "epoch": 0.09792,
+      "grad_norm": 0.7089570164680481,
+      "learning_rate": 3.2692307692307696e-06,
+      "loss": 0.2185,
+      "step": 51
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 0.7958459258079529,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.2866,
+      "step": 52
+    },
+    {
+      "epoch": 0.10176,
+      "grad_norm": 0.7239428758621216,
+      "learning_rate": 3.397435897435898e-06,
+      "loss": 0.2723,
+      "step": 53
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 0.7608776688575745,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 0.2189,
+      "step": 54
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 0.78788822889328,
+      "learning_rate": 3.5256410256410263e-06,
+      "loss": 0.2734,
+      "step": 55
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 0.700120210647583,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 0.2855,
+      "step": 56
+    },
+    {
+      "epoch": 0.10944,
+      "grad_norm": 0.7720645070075989,
+      "learning_rate": 3.653846153846154e-06,
+      "loss": 0.2582,
+      "step": 57
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 0.6845154166221619,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 0.2714,
+      "step": 58
+    },
+    {
+      "epoch": 0.11328,
+      "grad_norm": 0.6915775537490845,
+      "learning_rate": 3.782051282051282e-06,
+      "loss": 0.2351,
+      "step": 59
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.719704806804657,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.2699,
+      "step": 60
+    },
+    {
+      "epoch": 0.11712,
+      "grad_norm": 0.701739490032196,
+      "learning_rate": 3.910256410256411e-06,
+      "loss": 0.2103,
+      "step": 61
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 0.6719866394996643,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 0.2462,
+      "step": 62
+    },
+    {
+      "epoch": 0.12096,
+      "grad_norm": 0.6752523183822632,
+      "learning_rate": 4.0384615384615385e-06,
+      "loss": 0.2367,
+      "step": 63
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 0.7783870697021484,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 0.2428,
+      "step": 64
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 0.6974590420722961,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.2759,
+      "step": 65
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 0.7807461619377136,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 0.2896,
+      "step": 66
+    },
+    {
+      "epoch": 0.12864,
+      "grad_norm": 0.6923171281814575,
+      "learning_rate": 4.294871794871795e-06,
+      "loss": 0.2328,
+      "step": 67
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 0.7844114303588867,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 0.2442,
+      "step": 68
+    },
+    {
+      "epoch": 0.13248,
+      "grad_norm": 0.786483883857727,
+      "learning_rate": 4.423076923076924e-06,
+      "loss": 0.2461,
+      "step": 69
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.8369628190994263,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 0.2856,
+      "step": 70
+    },
+    {
+      "epoch": 0.13632,
+      "grad_norm": 0.7627629637718201,
+      "learning_rate": 4.551282051282052e-06,
+      "loss": 0.2682,
+      "step": 71
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 0.7634851932525635,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.283,
+      "step": 72
+    },
+    {
+      "epoch": 0.14016,
+      "grad_norm": 0.78012615442276,
+      "learning_rate": 4.6794871794871795e-06,
+      "loss": 0.2003,
+      "step": 73
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 0.7295075058937073,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 0.2997,
+      "step": 74
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 0.7868279814720154,
+      "learning_rate": 4.807692307692308e-06,
+      "loss": 0.2764,
+      "step": 75
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 0.7268662452697754,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 0.2019,
+      "step": 76
+    },
+    {
+      "epoch": 0.14784,
+      "grad_norm": 0.8158905506134033,
+      "learning_rate": 4.935897435897436e-06,
+      "loss": 0.2833,
+      "step": 77
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 0.6991612911224365,
+      "learning_rate": 5e-06,
+      "loss": 0.2661,
+      "step": 78
+    },
+    {
+      "epoch": 0.15168,
+      "grad_norm": 0.8402227759361267,
+      "learning_rate": 5.064102564102565e-06,
+      "loss": 0.2571,
+      "step": 79
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.8766093850135803,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.265,
+      "step": 80
+    },
+    {
+      "epoch": 0.15552,
+      "grad_norm": 0.7950035929679871,
+      "learning_rate": 5.192307692307693e-06,
+      "loss": 0.2673,
+      "step": 81
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 0.8521929979324341,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 0.2732,
+      "step": 82
+    },
+    {
+      "epoch": 0.15936,
+      "grad_norm": 0.7410707473754883,
+      "learning_rate": 5.320512820512821e-06,
+      "loss": 0.2384,
+      "step": 83
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 0.8307905793190002,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.343,
+      "step": 84
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 0.76188063621521,
+      "learning_rate": 5.448717948717949e-06,
+      "loss": 0.2167,
+      "step": 85
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 0.6742937564849854,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 0.2275,
+      "step": 86
+    },
+    {
+      "epoch": 0.16704,
+      "grad_norm": 0.727579653263092,
+      "learning_rate": 5.576923076923077e-06,
+      "loss": 0.2692,
+      "step": 87
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 0.6823550462722778,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 0.2393,
+      "step": 88
+    },
+    {
+      "epoch": 0.17088,
+      "grad_norm": 0.6587774753570557,
+      "learning_rate": 5.705128205128206e-06,
+      "loss": 0.2178,
+      "step": 89
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.7439016699790955,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 0.2299,
+      "step": 90
+    },
+    {
+      "epoch": 0.17472,
+      "grad_norm": 0.7538898587226868,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.2436,
+      "step": 91
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 0.7701774835586548,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 0.271,
+      "step": 92
+    },
+    {
+      "epoch": 0.17856,
+      "grad_norm": 0.708919107913971,
+      "learning_rate": 5.961538461538462e-06,
+      "loss": 0.2324,
+      "step": 93
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 0.7497438788414001,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 0.2531,
+      "step": 94
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 0.6407055258750916,
+      "learning_rate": 6.08974358974359e-06,
+      "loss": 0.2309,
+      "step": 95
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 0.6531554460525513,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.2381,
+      "step": 96
+    },
+    {
+      "epoch": 0.18624,
+      "grad_norm": 0.6916120648384094,
+      "learning_rate": 6.217948717948718e-06,
+      "loss": 0.2709,
+      "step": 97
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 0.7124315500259399,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 0.2953,
+      "step": 98
+    },
+    {
+      "epoch": 0.19008,
+      "grad_norm": 0.7291994094848633,
+      "learning_rate": 6.3461538461538466e-06,
+      "loss": 0.2753,
+      "step": 99
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.6929478049278259,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 0.2432,
+      "step": 100
+    },
+    {
+      "epoch": 0.19392,
+      "grad_norm": 0.6739209890365601,
+      "learning_rate": 6.474358974358975e-06,
+      "loss": 0.204,
+      "step": 101
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 0.7181931734085083,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 0.2434,
+      "step": 102
+    },
+    {
+      "epoch": 0.19776,
+      "grad_norm": 0.715926468372345,
+      "learning_rate": 6.602564102564103e-06,
+      "loss": 0.2282,
+      "step": 103
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 0.8047579526901245,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.2794,
+      "step": 104
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 0.6431995630264282,
+      "learning_rate": 6.730769230769232e-06,
+      "loss": 0.2408,
+      "step": 105
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 0.7323590517044067,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 0.263,
+      "step": 106
+    },
+    {
+      "epoch": 0.20544,
+      "grad_norm": 0.7388311624526978,
+      "learning_rate": 6.858974358974359e-06,
+      "loss": 0.2788,
+      "step": 107
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 0.7855584025382996,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.2113,
+      "step": 108
+    },
+    {
+      "epoch": 0.20928,
+      "grad_norm": 0.6835598349571228,
+      "learning_rate": 6.9871794871794876e-06,
+      "loss": 0.2056,
+      "step": 109
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.7709442973136902,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 0.2483,
+      "step": 110
+    },
+    {
+      "epoch": 0.21312,
+      "grad_norm": 0.6934860348701477,
+      "learning_rate": 7.115384615384616e-06,
+      "loss": 0.219,
+      "step": 111
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 0.707187294960022,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.2574,
+      "step": 112
+    },
+    {
+      "epoch": 0.21696,
+      "grad_norm": 0.7831624150276184,
+      "learning_rate": 7.243589743589744e-06,
+      "loss": 0.2466,
+      "step": 113
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 0.6905331015586853,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 0.2323,
+      "step": 114
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 0.6960011124610901,
+      "learning_rate": 7.371794871794873e-06,
+      "loss": 0.2307,
+      "step": 115
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 0.7618747353553772,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 0.2128,
+      "step": 116
+    },
+    {
+      "epoch": 0.22464,
+      "grad_norm": 0.8016367554664612,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2699,
+      "step": 117
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 0.7424344420433044,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.2514,
+      "step": 118
+    },
+    {
+      "epoch": 0.22848,
+      "grad_norm": 0.8300876617431641,
+      "learning_rate": 7.6282051282051286e-06,
+      "loss": 0.2757,
+      "step": 119
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.7565605044364929,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.2476,
+      "step": 120
+    },
+    {
+      "epoch": 0.23232,
+      "grad_norm": 0.726492702960968,
+      "learning_rate": 7.756410256410258e-06,
+      "loss": 0.3399,
+      "step": 121
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 0.6594569683074951,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 0.2776,
+      "step": 122
+    },
+    {
+      "epoch": 0.23616,
+      "grad_norm": 0.6827067732810974,
+      "learning_rate": 7.884615384615384e-06,
+      "loss": 0.2598,
+      "step": 123
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 0.7000330090522766,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 0.2496,
+      "step": 124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7726544737815857,
+      "learning_rate": 8.012820512820515e-06,
+      "loss": 0.2426,
+      "step": 125
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 0.7081710696220398,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 0.2309,
+      "step": 126
+    },
+    {
+      "epoch": 0.24384,
+      "grad_norm": 0.6654150485992432,
+      "learning_rate": 8.141025641025641e-06,
+      "loss": 0.2449,
+      "step": 127
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 0.8198509216308594,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.3192,
+      "step": 128
+    },
+    {
+      "epoch": 0.24768,
+      "grad_norm": 0.6441714763641357,
+      "learning_rate": 8.26923076923077e-06,
+      "loss": 0.2463,
+      "step": 129
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.7069028615951538,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.3115,
+      "step": 130
+    },
+    {
+      "epoch": 0.25152,
+      "grad_norm": 0.6577409505844116,
+      "learning_rate": 8.397435897435898e-06,
+      "loss": 0.2362,
+      "step": 131
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 0.7841039299964905,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.2945,
+      "step": 132
+    },
+    {
+      "epoch": 0.25536,
+      "grad_norm": 0.6786310076713562,
+      "learning_rate": 8.525641025641026e-06,
+      "loss": 0.2249,
+      "step": 133
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 0.7534569501876831,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 0.2482,
+      "step": 134
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 0.6911149621009827,
+      "learning_rate": 8.653846153846155e-06,
+      "loss": 0.2297,
+      "step": 135
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 0.753036618232727,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 0.3128,
+      "step": 136
+    },
+    {
+      "epoch": 0.26304,
+      "grad_norm": 0.5862494707107544,
+      "learning_rate": 8.782051282051283e-06,
+      "loss": 0.2022,
+      "step": 137
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 0.7523276209831238,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.2522,
+      "step": 138
+    },
+    {
+      "epoch": 0.26688,
+      "grad_norm": 0.7222450375556946,
+      "learning_rate": 8.910256410256411e-06,
+      "loss": 0.2475,
+      "step": 139
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.6471385359764099,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.2063,
+      "step": 140
+    },
+    {
+      "epoch": 0.27072,
+      "grad_norm": 0.706000030040741,
+      "learning_rate": 9.03846153846154e-06,
+      "loss": 0.2541,
+      "step": 141
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 0.6517675518989563,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 0.2469,
+      "step": 142
+    },
+    {
+      "epoch": 0.27456,
+      "grad_norm": 0.6817055940628052,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.2339,
+      "step": 143
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 0.6987940073013306,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.2771,
+      "step": 144
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 0.685405969619751,
+      "learning_rate": 9.294871794871796e-06,
+      "loss": 0.2478,
+      "step": 145
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 0.6707216501235962,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 0.2553,
+      "step": 146
+    },
+    {
+      "epoch": 0.28224,
+      "grad_norm": 0.6278187036514282,
+      "learning_rate": 9.423076923076923e-06,
+      "loss": 0.2381,
+      "step": 147
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 0.6790494322776794,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 0.2321,
+      "step": 148
+    },
+    {
+      "epoch": 0.28608,
+      "grad_norm": 0.732184886932373,
+      "learning_rate": 9.551282051282053e-06,
+      "loss": 0.2486,
+      "step": 149
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.6243297457695007,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 0.2405,
+      "step": 150
+    },
+    {
+      "epoch": 0.28992,
+      "grad_norm": 0.6482105851173401,
+      "learning_rate": 9.67948717948718e-06,
+      "loss": 0.2417,
+      "step": 151
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 0.6503863334655762,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 0.2483,
+      "step": 152
+    },
+    {
+      "epoch": 0.29376,
+      "grad_norm": 0.6719767451286316,
+      "learning_rate": 9.807692307692308e-06,
+      "loss": 0.2428,
+      "step": 153
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 0.6568065285682678,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 0.2695,
+      "step": 154
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 0.6765230298042297,
+      "learning_rate": 9.935897435897437e-06,
+      "loss": 0.224,
+      "step": 155
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 0.6482672095298767,
+      "learning_rate": 1e-05,
+      "loss": 0.2664,
+      "step": 156
+    },
+    {
+      "epoch": 0.30144,
+      "grad_norm": 0.621073305606842,
+      "learning_rate": 9.999987482852865e-06,
+      "loss": 0.212,
+      "step": 157
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 0.7035403251647949,
+      "learning_rate": 9.99994993147413e-06,
+      "loss": 0.2247,
+      "step": 158
+    },
+    {
+      "epoch": 0.30528,
+      "grad_norm": 0.5862267017364502,
+      "learning_rate": 9.99988734605181e-06,
+      "loss": 0.23,
+      "step": 159
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.6753953099250793,
+      "learning_rate": 9.999799726899261e-06,
+      "loss": 0.251,
+      "step": 160
+    },
+    {
+      "epoch": 0.30912,
+      "grad_norm": 0.6806496977806091,
+      "learning_rate": 9.999687074455182e-06,
+      "loss": 0.2649,
+      "step": 161
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 0.6245251297950745,
+      "learning_rate": 9.999549389283605e-06,
+      "loss": 0.2551,
+      "step": 162
+    },
+    {
+      "epoch": 0.31296,
+      "grad_norm": 0.6482666730880737,
+      "learning_rate": 9.999386672073902e-06,
+      "loss": 0.2669,
+      "step": 163
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 0.5988757610321045,
+      "learning_rate": 9.999198923640774e-06,
+      "loss": 0.216,
+      "step": 164
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 0.7105097770690918,
+      "learning_rate": 9.998986144924253e-06,
+      "loss": 0.2466,
+      "step": 165
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 0.7155487537384033,
+      "learning_rate": 9.998748336989687e-06,
+      "loss": 0.2555,
+      "step": 166
+    },
+    {
+      "epoch": 0.32064,
+      "grad_norm": 0.6257727146148682,
+      "learning_rate": 9.998485501027752e-06,
+      "loss": 0.2482,
+      "step": 167
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 0.6200207471847534,
+      "learning_rate": 9.998197638354428e-06,
+      "loss": 0.2684,
+      "step": 168
+    },
+    {
+      "epoch": 0.32448,
+      "grad_norm": 0.7291122078895569,
+      "learning_rate": 9.997884750411004e-06,
+      "loss": 0.236,
+      "step": 169
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.6901133060455322,
+      "learning_rate": 9.997546838764066e-06,
+      "loss": 0.2183,
+      "step": 170
+    },
+    {
+      "epoch": 0.32832,
+      "grad_norm": 0.6270235180854797,
+      "learning_rate": 9.997183905105487e-06,
+      "loss": 0.2517,
+      "step": 171
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 0.6173262596130371,
+      "learning_rate": 9.996795951252427e-06,
+      "loss": 0.2128,
+      "step": 172
+    },
+    {
+      "epoch": 0.33216,
+      "grad_norm": 0.7116228342056274,
+      "learning_rate": 9.996382979147316e-06,
+      "loss": 0.245,
+      "step": 173
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 0.6399964094161987,
+      "learning_rate": 9.995944990857848e-06,
+      "loss": 0.2194,
+      "step": 174
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 0.658183753490448,
+      "learning_rate": 9.995481988576968e-06,
+      "loss": 0.2249,
+      "step": 175
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 0.6090811491012573,
+      "learning_rate": 9.994993974622863e-06,
+      "loss": 0.2524,
+      "step": 176
+    },
+    {
+      "epoch": 0.33984,
+      "grad_norm": 0.6520531177520752,
+      "learning_rate": 9.994480951438948e-06,
+      "loss": 0.2426,
+      "step": 177
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 0.60045325756073,
+      "learning_rate": 9.993942921593858e-06,
+      "loss": 0.2705,
+      "step": 178
+    },
+    {
+      "epoch": 0.34368,
+      "grad_norm": 0.664503812789917,
+      "learning_rate": 9.993379887781436e-06,
+      "loss": 0.2892,
+      "step": 179
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.6904479265213013,
+      "learning_rate": 9.992791852820709e-06,
+      "loss": 0.2796,
+      "step": 180
+    },
+    {
+      "epoch": 0.34752,
+      "grad_norm": 0.6302940845489502,
+      "learning_rate": 9.992178819655889e-06,
+      "loss": 0.2224,
+      "step": 181
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 0.6948865652084351,
+      "learning_rate": 9.991540791356342e-06,
+      "loss": 0.226,
+      "step": 182
+    },
+    {
+      "epoch": 0.35136,
+      "grad_norm": 0.5956205725669861,
+      "learning_rate": 9.990877771116588e-06,
+      "loss": 0.2259,
+      "step": 183
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 0.6850026845932007,
+      "learning_rate": 9.990189762256275e-06,
+      "loss": 0.2296,
+      "step": 184
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 0.6686530113220215,
+      "learning_rate": 9.98947676822017e-06,
+      "loss": 0.2123,
+      "step": 185
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 0.628671407699585,
+      "learning_rate": 9.988738792578126e-06,
+      "loss": 0.2402,
+      "step": 186
+    },
+    {
+      "epoch": 0.35904,
+      "grad_norm": 0.7261161804199219,
+      "learning_rate": 9.987975839025091e-06,
+      "loss": 0.2418,
+      "step": 187
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 0.6552152633666992,
+      "learning_rate": 9.987187911381059e-06,
+      "loss": 0.2588,
+      "step": 188
+    },
+    {
+      "epoch": 0.36288,
+      "grad_norm": 0.6563432216644287,
+      "learning_rate": 9.986375013591077e-06,
+      "loss": 0.2396,
+      "step": 189
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.6631369590759277,
+      "learning_rate": 9.985537149725207e-06,
+      "loss": 0.2021,
+      "step": 190
+    },
+    {
+      "epoch": 0.36672,
+      "grad_norm": 0.6302971839904785,
+      "learning_rate": 9.984674323978517e-06,
+      "loss": 0.2169,
+      "step": 191
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 0.5025025606155396,
+      "learning_rate": 9.983786540671052e-06,
+      "loss": 0.2129,
+      "step": 192
+    },
+    {
+      "epoch": 0.37056,
+      "grad_norm": 0.7525653839111328,
+      "learning_rate": 9.982873804247817e-06,
+      "loss": 0.3003,
+      "step": 193
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 0.6977959275245667,
+      "learning_rate": 9.981936119278758e-06,
+      "loss": 0.2613,
+      "step": 194
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 0.6141538023948669,
+      "learning_rate": 9.980973490458728e-06,
+      "loss": 0.2609,
+      "step": 195
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 0.6548112034797668,
+      "learning_rate": 9.979985922607476e-06,
+      "loss": 0.2488,
+      "step": 196
+    },
+    {
+      "epoch": 0.37824,
+      "grad_norm": 0.5984265208244324,
+      "learning_rate": 9.978973420669613e-06,
+      "loss": 0.2117,
+      "step": 197
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 0.6127104759216309,
+      "learning_rate": 9.977935989714594e-06,
+      "loss": 0.3104,
+      "step": 198
+    },
+    {
+      "epoch": 0.38208,
+      "grad_norm": 0.6232851147651672,
+      "learning_rate": 9.97687363493669e-06,
+      "loss": 0.2319,
+      "step": 199
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.5733835101127625,
+      "learning_rate": 9.975786361654959e-06,
+      "loss": 0.2154,
+      "step": 200
+    },
+    {
+      "epoch": 0.38592,
+      "grad_norm": 0.6111357808113098,
+      "learning_rate": 9.974674175313228e-06,
+      "loss": 0.2229,
+      "step": 201
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 0.5716155767440796,
+      "learning_rate": 9.973537081480056e-06,
+      "loss": 0.2301,
+      "step": 202
+    },
+    {
+      "epoch": 0.38976,
+      "grad_norm": 0.5913110971450806,
+      "learning_rate": 9.972375085848712e-06,
+      "loss": 0.2287,
+      "step": 203
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 0.5164219737052917,
+      "learning_rate": 9.971188194237141e-06,
+      "loss": 0.1856,
+      "step": 204
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 0.618375301361084,
+      "learning_rate": 9.969976412587943e-06,
+      "loss": 0.2867,
+      "step": 205
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 0.690893292427063,
+      "learning_rate": 9.96873974696834e-06,
+      "loss": 0.2461,
+      "step": 206
+    },
+    {
+      "epoch": 0.39744,
+      "grad_norm": 0.5912200212478638,
+      "learning_rate": 9.967478203570143e-06,
+      "loss": 0.2656,
+      "step": 207
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 0.6609474420547485,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 0.2487,
+      "step": 208
+    },
+    {
+      "epoch": 0.40128,
+      "grad_norm": 0.6051667928695679,
+      "learning_rate": 9.96488050882796e-06,
+      "loss": 0.2085,
+      "step": 209
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.6379647850990295,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.2608,
+      "step": 210
+    },
+    {
+      "epoch": 0.40512,
+      "grad_norm": 0.617935836315155,
+      "learning_rate": 9.9621833803865e-06,
+      "loss": 0.2438,
+      "step": 211
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 0.6639268398284912,
+      "learning_rate": 9.960797545330936e-06,
+      "loss": 0.2788,
+      "step": 212
+    },
+    {
+      "epoch": 0.40896,
+      "grad_norm": 0.6133950352668762,
+      "learning_rate": 9.95938687226226e-06,
+      "loss": 0.2642,
+      "step": 213
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 0.5786393284797668,
+      "learning_rate": 9.95795136824351e-06,
+      "loss": 0.2463,
+      "step": 214
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 0.6375519633293152,
+      "learning_rate": 9.956491040462051e-06,
+      "loss": 0.2721,
+      "step": 215
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 0.6497066020965576,
+      "learning_rate": 9.955005896229543e-06,
+      "loss": 0.3134,
+      "step": 216
+    },
+    {
+      "epoch": 0.41664,
+      "grad_norm": 0.5720874071121216,
+      "learning_rate": 9.95349594298189e-06,
+      "loss": 0.2955,
+      "step": 217
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 0.6487167477607727,
+      "learning_rate": 9.951961188279216e-06,
+      "loss": 0.2684,
+      "step": 218
+    },
+    {
+      "epoch": 0.42048,
+      "grad_norm": 0.5410241484642029,
+      "learning_rate": 9.950401639805822e-06,
+      "loss": 0.2682,
+      "step": 219
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.6091211438179016,
+      "learning_rate": 9.948817305370145e-06,
+      "loss": 0.2926,
+      "step": 220
+    },
+    {
+      "epoch": 0.42432,
+      "grad_norm": 0.6725223064422607,
+      "learning_rate": 9.947208192904722e-06,
+      "loss": 0.2757,
+      "step": 221
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 0.5879321098327637,
+      "learning_rate": 9.94557431046616e-06,
+      "loss": 0.2266,
+      "step": 222
+    },
+    {
+      "epoch": 0.42816,
+      "grad_norm": 0.6407105922698975,
+      "learning_rate": 9.943915666235068e-06,
+      "loss": 0.2525,
+      "step": 223
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 0.6570138931274414,
+      "learning_rate": 9.942232268516051e-06,
+      "loss": 0.3127,
+      "step": 224
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 0.6066915988922119,
+      "learning_rate": 9.940524125737641e-06,
+      "loss": 0.219,
+      "step": 225
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 0.5652486085891724,
+      "learning_rate": 9.938791246452267e-06,
+      "loss": 0.2066,
+      "step": 226
+    },
+    {
+      "epoch": 0.43584,
+      "grad_norm": 0.5479260683059692,
+      "learning_rate": 9.937033639336212e-06,
+      "loss": 0.2142,
+      "step": 227
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 0.5939159393310547,
+      "learning_rate": 9.935251313189564e-06,
+      "loss": 0.2202,
+      "step": 228
+    },
+    {
+      "epoch": 0.43968,
+      "grad_norm": 0.619520366191864,
+      "learning_rate": 9.933444276936185e-06,
+      "loss": 0.2642,
+      "step": 229
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.6813686490058899,
+      "learning_rate": 9.931612539623643e-06,
+      "loss": 0.3228,
+      "step": 230
+    },
+    {
+      "epoch": 0.44352,
+      "grad_norm": 0.5384846329689026,
+      "learning_rate": 9.929756110423195e-06,
+      "loss": 0.2109,
+      "step": 231
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 0.5991112589836121,
+      "learning_rate": 9.927874998629714e-06,
+      "loss": 0.3153,
+      "step": 232
+    },
+    {
+      "epoch": 0.44736,
+      "grad_norm": 0.6019455194473267,
+      "learning_rate": 9.925969213661664e-06,
+      "loss": 0.2277,
+      "step": 233
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 0.5704585909843445,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 0.2711,
+      "step": 234
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 0.6854246258735657,
+      "learning_rate": 9.92208366249333e-06,
+      "loss": 0.329,
+      "step": 235
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 0.5871822834014893,
+      "learning_rate": 9.920103915747452e-06,
+      "loss": 0.2506,
+      "step": 236
+    },
+    {
+      "epoch": 0.45504,
+      "grad_norm": 0.616062343120575,
+      "learning_rate": 9.91809953473572e-06,
+      "loss": 0.2404,
+      "step": 237
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 0.6078352928161621,
+      "learning_rate": 9.916070529493785e-06,
+      "loss": 0.2438,
+      "step": 238
+    },
+    {
+      "epoch": 0.45888,
+      "grad_norm": 0.642365574836731,
+      "learning_rate": 9.914016910180593e-06,
+      "loss": 0.2678,
+      "step": 239
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.56107497215271,
+      "learning_rate": 9.911938687078324e-06,
+      "loss": 0.2353,
+      "step": 240
+    },
+    {
+      "epoch": 0.46272,
+      "grad_norm": 0.6760509014129639,
+      "learning_rate": 9.90983587059235e-06,
+      "loss": 0.2483,
+      "step": 241
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 0.6076192259788513,
+      "learning_rate": 9.907708471251173e-06,
+      "loss": 0.2358,
+      "step": 242
+    },
+    {
+      "epoch": 0.46656,
+      "grad_norm": 0.5566860437393188,
+      "learning_rate": 9.905556499706384e-06,
+      "loss": 0.2316,
+      "step": 243
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 0.6088006496429443,
+      "learning_rate": 9.9033799667326e-06,
+      "loss": 0.2781,
+      "step": 244
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 0.6450731754302979,
+      "learning_rate": 9.901178883227413e-06,
+      "loss": 0.2393,
+      "step": 245
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 0.5425491333007812,
+      "learning_rate": 9.89895326021134e-06,
+      "loss": 0.2374,
+      "step": 246
+    },
+    {
+      "epoch": 0.47424,
+      "grad_norm": 0.7007735967636108,
+      "learning_rate": 9.896703108827758e-06,
+      "loss": 0.3059,
+      "step": 247
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 0.5942116379737854,
+      "learning_rate": 9.89442844034286e-06,
+      "loss": 0.2663,
+      "step": 248
+    },
+    {
+      "epoch": 0.47808,
+      "grad_norm": 0.5787867903709412,
+      "learning_rate": 9.89212926614559e-06,
+      "loss": 0.2775,
+      "step": 249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6353012323379517,
+      "learning_rate": 9.889805597747588e-06,
+      "loss": 0.2131,
+      "step": 250
+    },
+    {
+      "epoch": 0.48192,
+      "grad_norm": 0.6993714570999146,
+      "learning_rate": 9.887457446783133e-06,
+      "loss": 0.2532,
+      "step": 251
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 0.6654945611953735,
+      "learning_rate": 9.885084825009085e-06,
+      "loss": 0.263,
+      "step": 252
+    },
+    {
+      "epoch": 0.48576,
+      "grad_norm": 0.5533124804496765,
+      "learning_rate": 9.88268774430483e-06,
+      "loss": 0.2543,
+      "step": 253
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 0.6826618909835815,
+      "learning_rate": 9.88026621667221e-06,
+      "loss": 0.2572,
+      "step": 254
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 0.5849668979644775,
+      "learning_rate": 9.87782025423547e-06,
+      "loss": 0.2467,
+      "step": 255
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 0.6387521624565125,
+      "learning_rate": 9.875349869241202e-06,
+      "loss": 0.231,
+      "step": 256
+    },
+    {
+      "epoch": 0.49344,
+      "grad_norm": 0.6104084253311157,
+      "learning_rate": 9.872855074058274e-06,
+      "loss": 0.2807,
+      "step": 257
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 0.5684860944747925,
+      "learning_rate": 9.870335881177774e-06,
+      "loss": 0.2478,
+      "step": 258
+    },
+    {
+      "epoch": 0.49728,
+      "grad_norm": 0.637405276298523,
+      "learning_rate": 9.867792303212944e-06,
+      "loss": 0.2977,
+      "step": 259
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.5618866682052612,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 0.2454,
+      "step": 260
+    },
+    {
+      "epoch": 0.50112,
+      "grad_norm": 0.5791952610015869,
+      "learning_rate": 9.862632043093667e-06,
+      "loss": 0.2201,
+      "step": 261
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 0.6754571795463562,
+      "learning_rate": 9.860015386775915e-06,
+      "loss": 0.2765,
+      "step": 262
+    },
+    {
+      "epoch": 0.50496,
+      "grad_norm": 0.611716628074646,
+      "learning_rate": 9.857374397047093e-06,
+      "loss": 0.2618,
+      "step": 263
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 0.5850534439086914,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.2449,
+      "step": 264
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 0.55945885181427,
+      "learning_rate": 9.852019470370254e-06,
+      "loss": 0.2252,
+      "step": 265
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 0.5426106452941895,
+      "learning_rate": 9.849305560233598e-06,
+      "loss": 0.2446,
+      "step": 266
+    },
+    {
+      "epoch": 0.51264,
+      "grad_norm": 0.5385035276412964,
+      "learning_rate": 9.846567370308462e-06,
+      "loss": 0.2047,
+      "step": 267
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 0.6069697737693787,
+      "learning_rate": 9.843804914304578e-06,
+      "loss": 0.2869,
+      "step": 268
+    },
+    {
+      "epoch": 0.51648,
+      "grad_norm": 0.5724849104881287,
+      "learning_rate": 9.841018206053167e-06,
+      "loss": 0.24,
+      "step": 269
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.5020802617073059,
+      "learning_rate": 9.838207259506891e-06,
+      "loss": 0.2289,
+      "step": 270
+    },
+    {
+      "epoch": 0.52032,
+      "grad_norm": 0.5671609044075012,
+      "learning_rate": 9.835372088739758e-06,
+      "loss": 0.2737,
+      "step": 271
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 0.5900096297264099,
+      "learning_rate": 9.83251270794707e-06,
+      "loss": 0.2401,
+      "step": 272
+    },
+    {
+      "epoch": 0.52416,
+      "grad_norm": 0.6300389766693115,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.259,
+      "step": 273
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 0.6236871480941772,
+      "learning_rate": 9.826721373672235e-06,
+      "loss": 0.2832,
+      "step": 274
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 0.5602126717567444,
+      "learning_rate": 9.82378944918648e-06,
+      "loss": 0.232,
+      "step": 275
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 0.6304101943969727,
+      "learning_rate": 9.820833372667813e-06,
+      "loss": 0.2507,
+      "step": 276
+    },
+    {
+      "epoch": 0.53184,
+      "grad_norm": 0.5833680629730225,
+      "learning_rate": 9.817853158916889e-06,
+      "loss": 0.2564,
+      "step": 277
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 0.543664276599884,
+      "learning_rate": 9.814848822855216e-06,
+      "loss": 0.2451,
+      "step": 278
+    },
+    {
+      "epoch": 0.53568,
+      "grad_norm": 0.6190241575241089,
+      "learning_rate": 9.811820379525085e-06,
+      "loss": 0.2482,
+      "step": 279
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.6057313680648804,
+      "learning_rate": 9.80876784408948e-06,
+      "loss": 0.2532,
+      "step": 280
+    },
+    {
+      "epoch": 0.53952,
+      "grad_norm": 0.5597018599510193,
+      "learning_rate": 9.805691231832018e-06,
+      "loss": 0.227,
+      "step": 281
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 0.5189545154571533,
+      "learning_rate": 9.802590558156863e-06,
+      "loss": 0.2132,
+      "step": 282
+    },
+    {
+      "epoch": 0.54336,
+      "grad_norm": 0.586489737033844,
+      "learning_rate": 9.799465838588646e-06,
+      "loss": 0.2228,
+      "step": 283
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 0.6584839224815369,
+      "learning_rate": 9.796317088772402e-06,
+      "loss": 0.3078,
+      "step": 284
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 0.5404163599014282,
+      "learning_rate": 9.793144324473473e-06,
+      "loss": 0.2625,
+      "step": 285
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 0.5509477257728577,
+      "learning_rate": 9.789947561577445e-06,
+      "loss": 0.267,
+      "step": 286
+    },
+    {
+      "epoch": 0.55104,
+      "grad_norm": 0.5755206942558289,
+      "learning_rate": 9.786726816090058e-06,
+      "loss": 0.2634,
+      "step": 287
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 0.5985530614852905,
+      "learning_rate": 9.783482104137127e-06,
+      "loss": 0.2592,
+      "step": 288
+    },
+    {
+      "epoch": 0.55488,
+      "grad_norm": 0.5863509178161621,
+      "learning_rate": 9.78021344196447e-06,
+      "loss": 0.2691,
+      "step": 289
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.5964622497558594,
+      "learning_rate": 9.776920845937816e-06,
+      "loss": 0.2382,
+      "step": 290
+    },
+    {
+      "epoch": 0.55872,
+      "grad_norm": 0.581659197807312,
+      "learning_rate": 9.77360433254273e-06,
+      "loss": 0.2586,
+      "step": 291
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 0.5457330346107483,
+      "learning_rate": 9.770263918384523e-06,
+      "loss": 0.229,
+      "step": 292
+    },
+    {
+      "epoch": 0.56256,
+      "grad_norm": 0.6449571251869202,
+      "learning_rate": 9.766899620188181e-06,
+      "loss": 0.3025,
+      "step": 293
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 0.48156464099884033,
+      "learning_rate": 9.763511454798268e-06,
+      "loss": 0.2369,
+      "step": 294
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 0.583936870098114,
+      "learning_rate": 9.760099439178852e-06,
+      "loss": 0.3104,
+      "step": 295
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 0.5008465647697449,
+      "learning_rate": 9.75666359041341e-06,
+      "loss": 0.2098,
+      "step": 296
+    },
+    {
+      "epoch": 0.57024,
+      "grad_norm": 0.5655997395515442,
+      "learning_rate": 9.753203925704756e-06,
+      "loss": 0.2525,
+      "step": 297
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 0.4871576130390167,
+      "learning_rate": 9.749720462374939e-06,
+      "loss": 0.2153,
+      "step": 298
+    },
+    {
+      "epoch": 0.57408,
+      "grad_norm": 0.5949690341949463,
+      "learning_rate": 9.74621321786517e-06,
+      "loss": 0.2685,
+      "step": 299
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.5801219940185547,
+      "learning_rate": 9.742682209735727e-06,
+      "loss": 0.2324,
+      "step": 300
+    },
+    {
+      "epoch": 0.57792,
+      "grad_norm": 0.522911787033081,
+      "learning_rate": 9.739127455665872e-06,
+      "loss": 0.2474,
+      "step": 301
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 0.6006777286529541,
+      "learning_rate": 9.735548973453753e-06,
+      "loss": 0.2634,
+      "step": 302
+    },
+    {
+      "epoch": 0.58176,
+      "grad_norm": 0.6128909587860107,
+      "learning_rate": 9.731946781016328e-06,
+      "loss": 0.2708,
+      "step": 303
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 0.6076428890228271,
+      "learning_rate": 9.728320896389263e-06,
+      "loss": 0.234,
+      "step": 304
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 0.5654532313346863,
+      "learning_rate": 9.724671337726855e-06,
+      "loss": 0.2443,
+      "step": 305
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 0.5384954214096069,
+      "learning_rate": 9.720998123301924e-06,
+      "loss": 0.2274,
+      "step": 306
+    },
+    {
+      "epoch": 0.58944,
+      "grad_norm": 0.555728018283844,
+      "learning_rate": 9.717301271505739e-06,
+      "loss": 0.2662,
+      "step": 307
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 0.5359607934951782,
+      "learning_rate": 9.713580800847917e-06,
+      "loss": 0.2759,
+      "step": 308
+    },
+    {
+      "epoch": 0.59328,
+      "grad_norm": 0.6365648508071899,
+      "learning_rate": 9.709836729956326e-06,
+      "loss": 0.2825,
+      "step": 309
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.5268747806549072,
+      "learning_rate": 9.706069077577e-06,
+      "loss": 0.187,
+      "step": 310
+    },
+    {
+      "epoch": 0.59712,
+      "grad_norm": 0.5515536069869995,
+      "learning_rate": 9.702277862574048e-06,
+      "loss": 0.2461,
+      "step": 311
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 0.5644161105155945,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.2962,
+      "step": 312
+    },
+    {
+      "epoch": 0.60096,
+      "grad_norm": 0.5509657859802246,
+      "learning_rate": 9.694624820743446e-06,
+      "loss": 0.2364,
+      "step": 313
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 0.49658122658729553,
+      "learning_rate": 9.690763032233498e-06,
+      "loss": 0.2071,
+      "step": 314
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 0.5542978644371033,
+      "learning_rate": 9.686877757735126e-06,
+      "loss": 0.2162,
+      "step": 315
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 0.5309951901435852,
+      "learning_rate": 9.682969016701357e-06,
+      "loss": 0.2353,
+      "step": 316
+    },
+    {
+      "epoch": 0.60864,
+      "grad_norm": 0.566458523273468,
+      "learning_rate": 9.679036828702704e-06,
+      "loss": 0.2603,
+      "step": 317
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 0.603590726852417,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 0.2977,
+      "step": 318
+    },
+    {
+      "epoch": 0.61248,
+      "grad_norm": 0.6277260780334473,
+      "learning_rate": 9.671102190679678e-06,
+      "loss": 0.2437,
+      "step": 319
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.6408616900444031,
+      "learning_rate": 9.66709978038292e-06,
+      "loss": 0.3418,
+      "step": 320
+    },
+    {
+      "epoch": 0.61632,
+      "grad_norm": 0.5598763227462769,
+      "learning_rate": 9.663074002576303e-06,
+      "loss": 0.2355,
+      "step": 321
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 0.5852317214012146,
+      "learning_rate": 9.659024877416328e-06,
+      "loss": 0.2992,
+      "step": 322
+    },
+    {
+      "epoch": 0.62016,
+      "grad_norm": 0.5498847961425781,
+      "learning_rate": 9.654952425176392e-06,
+      "loss": 0.2228,
+      "step": 323
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 0.5655949711799622,
+      "learning_rate": 9.650856666246693e-06,
+      "loss": 0.2411,
+      "step": 324
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 0.5796497464179993,
+      "learning_rate": 9.646737621134112e-06,
+      "loss": 0.2739,
+      "step": 325
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 0.5519629120826721,
+      "learning_rate": 9.642595310462133e-06,
+      "loss": 0.2345,
+      "step": 326
+    },
+    {
+      "epoch": 0.62784,
+      "grad_norm": 0.502500593662262,
+      "learning_rate": 9.638429754970715e-06,
+      "loss": 0.2388,
+      "step": 327
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 0.5340890884399414,
+      "learning_rate": 9.63424097551621e-06,
+      "loss": 0.2207,
+      "step": 328
+    },
+    {
+      "epoch": 0.63168,
+      "grad_norm": 0.5414576530456543,
+      "learning_rate": 9.630028993071244e-06,
+      "loss": 0.2702,
+      "step": 329
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.5914018750190735,
+      "learning_rate": 9.62579382872462e-06,
+      "loss": 0.2273,
+      "step": 330
+    },
+    {
+      "epoch": 0.63552,
+      "grad_norm": 0.6208275556564331,
+      "learning_rate": 9.621535503681205e-06,
+      "loss": 0.2386,
+      "step": 331
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 0.5301160216331482,
+      "learning_rate": 9.617254039261835e-06,
+      "loss": 0.2184,
+      "step": 332
+    },
+    {
+      "epoch": 0.63936,
+      "grad_norm": 0.478123277425766,
+      "learning_rate": 9.612949456903195e-06,
+      "loss": 0.2079,
+      "step": 333
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 0.5615213513374329,
+      "learning_rate": 9.608621778157722e-06,
+      "loss": 0.2433,
+      "step": 334
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 0.5730128288269043,
+      "learning_rate": 9.604271024693495e-06,
+      "loss": 0.2219,
+      "step": 335
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 0.5428596138954163,
+      "learning_rate": 9.599897218294122e-06,
+      "loss": 0.23,
+      "step": 336
+    },
+    {
+      "epoch": 0.64704,
+      "grad_norm": 0.5735312104225159,
+      "learning_rate": 9.595500380858632e-06,
+      "loss": 0.2779,
+      "step": 337
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 0.5011629462242126,
+      "learning_rate": 9.591080534401371e-06,
+      "loss": 0.2338,
+      "step": 338
+    },
+    {
+      "epoch": 0.65088,
+      "grad_norm": 0.5241754651069641,
+      "learning_rate": 9.586637701051886e-06,
+      "loss": 0.2695,
+      "step": 339
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.5298614501953125,
+      "learning_rate": 9.582171903054815e-06,
+      "loss": 0.2579,
+      "step": 340
+    },
+    {
+      "epoch": 0.65472,
+      "grad_norm": 0.4902747869491577,
+      "learning_rate": 9.577683162769781e-06,
+      "loss": 0.2276,
+      "step": 341
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 0.587215006351471,
+      "learning_rate": 9.573171502671273e-06,
+      "loss": 0.2544,
+      "step": 342
+    },
+    {
+      "epoch": 0.65856,
+      "grad_norm": 0.5700992345809937,
+      "learning_rate": 9.568636945348534e-06,
+      "loss": 0.272,
+      "step": 343
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 0.5466131567955017,
+      "learning_rate": 9.564079513505455e-06,
+      "loss": 0.2599,
+      "step": 344
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 0.5170016884803772,
+      "learning_rate": 9.55949922996045e-06,
+      "loss": 0.2152,
+      "step": 345
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 0.5629295110702515,
+      "learning_rate": 9.554896117646357e-06,
+      "loss": 0.2811,
+      "step": 346
+    },
+    {
+      "epoch": 0.66624,
+      "grad_norm": 0.5906686186790466,
+      "learning_rate": 9.550270199610307e-06,
+      "loss": 0.2272,
+      "step": 347
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 0.5153129696846008,
+      "learning_rate": 9.54562149901362e-06,
+      "loss": 0.2291,
+      "step": 348
+    },
+    {
+      "epoch": 0.67008,
+      "grad_norm": 0.5753116607666016,
+      "learning_rate": 9.54095003913168e-06,
+      "loss": 0.2448,
+      "step": 349
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.565992534160614,
+      "learning_rate": 9.536255843353832e-06,
+      "loss": 0.2286,
+      "step": 350
+    },
+    {
+      "epoch": 0.67392,
+      "grad_norm": 0.5491060614585876,
+      "learning_rate": 9.531538935183252e-06,
+      "loss": 0.2127,
+      "step": 351
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 0.47297415137290955,
+      "learning_rate": 9.526799338236828e-06,
+      "loss": 0.2203,
+      "step": 352
+    },
+    {
+      "epoch": 0.67776,
+      "grad_norm": 0.576661229133606,
+      "learning_rate": 9.522037076245057e-06,
+      "loss": 0.2186,
+      "step": 353
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 0.5790955424308777,
+      "learning_rate": 9.517252173051912e-06,
+      "loss": 0.2277,
+      "step": 354
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 0.6210295557975769,
+      "learning_rate": 9.512444652614728e-06,
+      "loss": 0.2633,
+      "step": 355
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 0.5863310098648071,
+      "learning_rate": 9.507614539004082e-06,
+      "loss": 0.2331,
+      "step": 356
+    },
+    {
+      "epoch": 0.68544,
+      "grad_norm": 0.5566739439964294,
+      "learning_rate": 9.502761856403668e-06,
+      "loss": 0.2154,
+      "step": 357
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 0.5596410632133484,
+      "learning_rate": 9.497886629110187e-06,
+      "loss": 0.2413,
+      "step": 358
+    },
+    {
+      "epoch": 0.68928,
+      "grad_norm": 0.5672785043716431,
+      "learning_rate": 9.49298888153321e-06,
+      "loss": 0.2609,
+      "step": 359
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.5325866341590881,
+      "learning_rate": 9.488068638195072e-06,
+      "loss": 0.2489,
+      "step": 360
+    },
+    {
+      "epoch": 0.69312,
+      "grad_norm": 0.5587996244430542,
+      "learning_rate": 9.483125923730731e-06,
+      "loss": 0.2371,
+      "step": 361
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 0.5696333646774292,
+      "learning_rate": 9.478160762887668e-06,
+      "loss": 0.2693,
+      "step": 362
+    },
+    {
+      "epoch": 0.69696,
+      "grad_norm": 0.5916242599487305,
+      "learning_rate": 9.473173180525737e-06,
+      "loss": 0.2894,
+      "step": 363
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 0.4755476415157318,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 0.1996,
+      "step": 364
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 0.5198381543159485,
+      "learning_rate": 9.463130851245899e-06,
+      "loss": 0.2431,
+      "step": 365
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 0.5410839915275574,
+      "learning_rate": 9.458076154608515e-06,
+      "loss": 0.2791,
+      "step": 366
+    },
+    {
+      "epoch": 0.70464,
+      "grad_norm": 0.5169705748558044,
+      "learning_rate": 9.452999137013064e-06,
+      "loss": 0.2147,
+      "step": 367
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 0.5489817261695862,
+      "learning_rate": 9.447899823879456e-06,
+      "loss": 0.2337,
+      "step": 368
+    },
+    {
+      "epoch": 0.70848,
+      "grad_norm": 0.5449849963188171,
+      "learning_rate": 9.442778240739234e-06,
+      "loss": 0.2626,
+      "step": 369
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.5482586026191711,
+      "learning_rate": 9.437634413235437e-06,
+      "loss": 0.2668,
+      "step": 370
+    },
+    {
+      "epoch": 0.71232,
+      "grad_norm": 0.5569681525230408,
+      "learning_rate": 9.43246836712249e-06,
+      "loss": 0.2261,
+      "step": 371
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 0.5599213242530823,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.2481,
+      "step": 372
+    },
+    {
+      "epoch": 0.71616,
+      "grad_norm": 0.5262666344642639,
+      "learning_rate": 9.4220697226429e-06,
+      "loss": 0.2454,
+      "step": 373
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 0.5451787710189819,
+      "learning_rate": 9.416837176340807e-06,
+      "loss": 0.2586,
+      "step": 374
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.540029764175415,
+      "learning_rate": 9.411582515558391e-06,
+      "loss": 0.2808,
+      "step": 375
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 0.5353924036026001,
+      "learning_rate": 9.406305766604996e-06,
+      "loss": 0.2344,
+      "step": 376
+    },
+    {
+      "epoch": 0.72384,
+      "grad_norm": 0.5113332271575928,
+      "learning_rate": 9.401006955900555e-06,
+      "loss": 0.2239,
+      "step": 377
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 0.5374947190284729,
+      "learning_rate": 9.395686109975475e-06,
+      "loss": 0.2578,
+      "step": 378
+    },
+    {
+      "epoch": 0.72768,
+      "grad_norm": 0.5157745480537415,
+      "learning_rate": 9.390343255470471e-06,
+      "loss": 0.2652,
+      "step": 379
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.4885575771331787,
+      "learning_rate": 9.384978419136469e-06,
+      "loss": 0.2274,
+      "step": 380
+    },
+    {
+      "epoch": 0.73152,
+      "grad_norm": 0.5641977787017822,
+      "learning_rate": 9.37959162783444e-06,
+      "loss": 0.2769,
+      "step": 381
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 0.5881760716438293,
+      "learning_rate": 9.374182908535293e-06,
+      "loss": 0.2246,
+      "step": 382
+    },
+    {
+      "epoch": 0.73536,
+      "grad_norm": 0.5225879549980164,
+      "learning_rate": 9.368752288319722e-06,
+      "loss": 0.2339,
+      "step": 383
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 0.5258158445358276,
+      "learning_rate": 9.363299794378072e-06,
+      "loss": 0.2307,
+      "step": 384
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 0.5229877233505249,
+      "learning_rate": 9.357825454010214e-06,
+      "loss": 0.245,
+      "step": 385
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 0.6318989992141724,
+      "learning_rate": 9.352329294625397e-06,
+      "loss": 0.267,
+      "step": 386
+    },
+    {
+      "epoch": 0.74304,
+      "grad_norm": 0.5190733671188354,
+      "learning_rate": 9.346811343742115e-06,
+      "loss": 0.2342,
+      "step": 387
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 0.5874361395835876,
+      "learning_rate": 9.34127162898797e-06,
+      "loss": 0.2405,
+      "step": 388
+    },
+    {
+      "epoch": 0.74688,
+      "grad_norm": 0.535131573677063,
+      "learning_rate": 9.335710178099528e-06,
+      "loss": 0.2047,
+      "step": 389
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.5369254946708679,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.2278,
+      "step": 390
+    },
+    {
+      "epoch": 0.75072,
+      "grad_norm": 0.5374722480773926,
+      "learning_rate": 9.324522179410054e-06,
+      "loss": 0.2394,
+      "step": 391
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 0.5450063347816467,
+      "learning_rate": 9.318895687625752e-06,
+      "loss": 0.2602,
+      "step": 392
+    },
+    {
+      "epoch": 0.75456,
+      "grad_norm": 0.4665793478488922,
+      "learning_rate": 9.313247571740336e-06,
+      "loss": 0.2327,
+      "step": 393
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 0.6042125821113586,
+      "learning_rate": 9.307577860033123e-06,
+      "loss": 0.3077,
+      "step": 394
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 0.4995546340942383,
+      "learning_rate": 9.301886580891563e-06,
+      "loss": 0.22,
+      "step": 395
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 0.5512630939483643,
+      "learning_rate": 9.296173762811084e-06,
+      "loss": 0.2392,
+      "step": 396
+    },
+    {
+      "epoch": 0.76224,
+      "grad_norm": 0.5573067665100098,
+      "learning_rate": 9.290439434394965e-06,
+      "loss": 0.2732,
+      "step": 397
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 0.5003942251205444,
+      "learning_rate": 9.284683624354172e-06,
+      "loss": 0.2372,
+      "step": 398
+    },
+    {
+      "epoch": 0.76608,
+      "grad_norm": 0.5498794317245483,
+      "learning_rate": 9.278906361507238e-06,
+      "loss": 0.2265,
+      "step": 399
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.6044484376907349,
+      "learning_rate": 9.273107674780102e-06,
+      "loss": 0.2037,
+      "step": 400
+    },
+    {
+      "epoch": 0.76992,
+      "grad_norm": 0.552961528301239,
+      "learning_rate": 9.26728759320597e-06,
+      "loss": 0.2637,
+      "step": 401
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 0.5439772009849548,
+      "learning_rate": 9.261446145925167e-06,
+      "loss": 0.2869,
+      "step": 402
+    },
+    {
+      "epoch": 0.77376,
+      "grad_norm": 0.5917429327964783,
+      "learning_rate": 9.255583362184998e-06,
+      "loss": 0.322,
+      "step": 403
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 0.5769372582435608,
+      "learning_rate": 9.249699271339594e-06,
+      "loss": 0.2302,
+      "step": 404
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 0.5873336791992188,
+      "learning_rate": 9.243793902849764e-06,
+      "loss": 0.2705,
+      "step": 405
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 0.5380690097808838,
+      "learning_rate": 9.237867286282855e-06,
+      "loss": 0.2287,
+      "step": 406
+    },
+    {
+      "epoch": 0.78144,
+      "grad_norm": 0.5661457180976868,
+      "learning_rate": 9.231919451312603e-06,
+      "loss": 0.2882,
+      "step": 407
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 0.5237562656402588,
+      "learning_rate": 9.225950427718974e-06,
+      "loss": 0.2119,
+      "step": 408
+    },
+    {
+      "epoch": 0.78528,
+      "grad_norm": 0.5161927342414856,
+      "learning_rate": 9.21996024538803e-06,
+      "loss": 0.2188,
+      "step": 409
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.5826548337936401,
+      "learning_rate": 9.213948934311767e-06,
+      "loss": 0.2727,
+      "step": 410
+    },
+    {
+      "epoch": 0.78912,
+      "grad_norm": 0.525320291519165,
+      "learning_rate": 9.207916524587971e-06,
+      "loss": 0.2552,
+      "step": 411
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 0.5418238043785095,
+      "learning_rate": 9.201863046420065e-06,
+      "loss": 0.207,
+      "step": 412
+    },
+    {
+      "epoch": 0.79296,
+      "grad_norm": 0.5078104734420776,
+      "learning_rate": 9.195788530116962e-06,
+      "loss": 0.2239,
+      "step": 413
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 0.509878396987915,
+      "learning_rate": 9.189693006092907e-06,
+      "loss": 0.2013,
+      "step": 414
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 0.5485806465148926,
+      "learning_rate": 9.183576504867328e-06,
+      "loss": 0.2246,
+      "step": 415
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 0.5393770933151245,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 0.2432,
+      "step": 416
+    },
+    {
+      "epoch": 0.80064,
+      "grad_norm": 0.5778350830078125,
+      "learning_rate": 9.171280693414307e-06,
+      "loss": 0.2401,
+      "step": 417
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 0.5869356393814087,
+      "learning_rate": 9.165101444750259e-06,
+      "loss": 0.2778,
+      "step": 418
+    },
+    {
+      "epoch": 0.80448,
+      "grad_norm": 0.49087539315223694,
+      "learning_rate": 9.158901342011163e-06,
+      "loss": 0.2437,
+      "step": 419
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.549502968788147,
+      "learning_rate": 9.152680416240059e-06,
+      "loss": 0.2411,
+      "step": 420
+    },
+    {
+      "epoch": 0.80832,
+      "grad_norm": 0.5855428576469421,
+      "learning_rate": 9.146438698584245e-06,
+      "loss": 0.2771,
+      "step": 421
+    },
+    {
+      "epoch": 0.81024,
+      "grad_norm": 0.5593034625053406,
+      "learning_rate": 9.140176220295118e-06,
+      "loss": 0.273,
+      "step": 422
+    },
+    {
+      "epoch": 0.81216,
+      "grad_norm": 0.5278566479682922,
+      "learning_rate": 9.133893012728027e-06,
+      "loss": 0.2371,
+      "step": 423
+    },
+    {
+      "epoch": 0.81408,
+      "grad_norm": 0.5541067719459534,
+      "learning_rate": 9.1275891073421e-06,
+      "loss": 0.2589,
+      "step": 424
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 0.49912944436073303,
+      "learning_rate": 9.121264535700107e-06,
+      "loss": 0.1931,
+      "step": 425
+    },
+    {
+      "epoch": 0.81792,
+      "grad_norm": 0.5119869709014893,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.2583,
+      "step": 426
+    },
+    {
+      "epoch": 0.81984,
+      "grad_norm": 0.529334545135498,
+      "learning_rate": 9.10855352041618e-06,
+      "loss": 0.2112,
+      "step": 427
+    },
+    {
+      "epoch": 0.82176,
+      "grad_norm": 0.6273512244224548,
+      "learning_rate": 9.102167140416503e-06,
+      "loss": 0.308,
+      "step": 428
+    },
+    {
+      "epoch": 0.82368,
+      "grad_norm": 0.5162720680236816,
+      "learning_rate": 9.09576022144496e-06,
+      "loss": 0.2577,
+      "step": 429
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.5145787596702576,
+      "learning_rate": 9.089332795580085e-06,
+      "loss": 0.2715,
+      "step": 430
+    },
+    {
+      "epoch": 0.82752,
+      "grad_norm": 0.5328844785690308,
+      "learning_rate": 9.082884895003098e-06,
+      "loss": 0.2401,
+      "step": 431
+    },
+    {
+      "epoch": 0.82944,
+      "grad_norm": 0.5591639280319214,
+      "learning_rate": 9.076416551997721e-06,
+      "loss": 0.2509,
+      "step": 432
+    },
+    {
+      "epoch": 0.83136,
+      "grad_norm": 0.5044178366661072,
+      "learning_rate": 9.06992779895004e-06,
+      "loss": 0.2181,
+      "step": 433
+    },
+    {
+      "epoch": 0.83328,
+      "grad_norm": 0.6215513348579407,
+      "learning_rate": 9.063418668348323e-06,
+      "loss": 0.276,
+      "step": 434
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 0.5775128602981567,
+      "learning_rate": 9.056889192782865e-06,
+      "loss": 0.2321,
+      "step": 435
+    },
+    {
+      "epoch": 0.83712,
+      "grad_norm": 0.4803057312965393,
+      "learning_rate": 9.050339404945834e-06,
+      "loss": 0.2403,
+      "step": 436
+    },
+    {
+      "epoch": 0.83904,
+      "grad_norm": 0.5041178464889526,
+      "learning_rate": 9.04376933763109e-06,
+      "loss": 0.2147,
+      "step": 437
+    },
+    {
+      "epoch": 0.84096,
+      "grad_norm": 0.5535988807678223,
+      "learning_rate": 9.037179023734036e-06,
+      "loss": 0.2683,
+      "step": 438
+    },
+    {
+      "epoch": 0.84288,
+      "grad_norm": 0.5130117535591125,
+      "learning_rate": 9.030568496251437e-06,
+      "loss": 0.2338,
+      "step": 439
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.5483830571174622,
+      "learning_rate": 9.023937788281278e-06,
+      "loss": 0.2691,
+      "step": 440
+    },
+    {
+      "epoch": 0.84672,
+      "grad_norm": 0.561305582523346,
+      "learning_rate": 9.017286933022573e-06,
+      "loss": 0.2653,
+      "step": 441
+    },
+    {
+      "epoch": 0.84864,
+      "grad_norm": 0.5333814024925232,
+      "learning_rate": 9.01061596377522e-06,
+      "loss": 0.2334,
+      "step": 442
+    },
+    {
+      "epoch": 0.85056,
+      "grad_norm": 0.592914879322052,
+      "learning_rate": 9.003924913939816e-06,
+      "loss": 0.3047,
+      "step": 443
+    },
+    {
+      "epoch": 0.85248,
+      "grad_norm": 0.5908307433128357,
+      "learning_rate": 8.997213817017508e-06,
+      "loss": 0.2504,
+      "step": 444
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 0.5451295971870422,
+      "learning_rate": 8.990482706609805e-06,
+      "loss": 0.2769,
+      "step": 445
+    },
+    {
+      "epoch": 0.85632,
+      "grad_norm": 0.5059984922409058,
+      "learning_rate": 8.98373161641843e-06,
+      "loss": 0.2186,
+      "step": 446
+    },
+    {
+      "epoch": 0.85824,
+      "grad_norm": 0.5239985585212708,
+      "learning_rate": 8.97696058024514e-06,
+      "loss": 0.2401,
+      "step": 447
+    },
+    {
+      "epoch": 0.86016,
+      "grad_norm": 0.47070619463920593,
+      "learning_rate": 8.970169631991556e-06,
+      "loss": 0.2109,
+      "step": 448
+    },
+    {
+      "epoch": 0.86208,
+      "grad_norm": 0.4998605251312256,
+      "learning_rate": 8.963358805658998e-06,
+      "loss": 0.2049,
+      "step": 449
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.5802593231201172,
+      "learning_rate": 8.95652813534831e-06,
+      "loss": 0.2398,
+      "step": 450
+    },
+    {
+      "epoch": 0.86592,
+      "grad_norm": 0.5537770390510559,
+      "learning_rate": 8.949677655259696e-06,
+      "loss": 0.2639,
+      "step": 451
+    },
+    {
+      "epoch": 0.86784,
+      "grad_norm": 0.5472473502159119,
+      "learning_rate": 8.942807399692543e-06,
+      "loss": 0.2547,
+      "step": 452
+    },
+    {
+      "epoch": 0.86976,
+      "grad_norm": 0.5411022305488586,
+      "learning_rate": 8.935917403045251e-06,
+      "loss": 0.2519,
+      "step": 453
+    },
+    {
+      "epoch": 0.87168,
+      "grad_norm": 0.5475094318389893,
+      "learning_rate": 8.92900769981506e-06,
+      "loss": 0.248,
+      "step": 454
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 0.5562335848808289,
+      "learning_rate": 8.92207832459788e-06,
+      "loss": 0.2423,
+      "step": 455
+    },
+    {
+      "epoch": 0.87552,
+      "grad_norm": 0.526985228061676,
+      "learning_rate": 8.915129312088112e-06,
+      "loss": 0.257,
+      "step": 456
+    },
+    {
+      "epoch": 0.87744,
+      "grad_norm": 0.5491049885749817,
+      "learning_rate": 8.908160697078482e-06,
+      "loss": 0.264,
+      "step": 457
+    },
+    {
+      "epoch": 0.87936,
+      "grad_norm": 0.5826250910758972,
+      "learning_rate": 8.901172514459864e-06,
+      "loss": 0.2499,
+      "step": 458
+    },
+    {
+      "epoch": 0.88128,
+      "grad_norm": 0.47528281807899475,
+      "learning_rate": 8.894164799221099e-06,
+      "loss": 0.2294,
+      "step": 459
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.5538657307624817,
+      "learning_rate": 8.88713758644883e-06,
+      "loss": 0.2477,
+      "step": 460
+    },
+    {
+      "epoch": 0.88512,
+      "grad_norm": 0.5086276531219482,
+      "learning_rate": 8.88009091132732e-06,
+      "loss": 0.2327,
+      "step": 461
+    },
+    {
+      "epoch": 0.88704,
+      "grad_norm": 0.5816463232040405,
+      "learning_rate": 8.873024809138272e-06,
+      "loss": 0.2594,
+      "step": 462
+    },
+    {
+      "epoch": 0.88896,
+      "grad_norm": 0.5248664617538452,
+      "learning_rate": 8.86593931526067e-06,
+      "loss": 0.2306,
+      "step": 463
+    },
+    {
+      "epoch": 0.89088,
+      "grad_norm": 0.5257532596588135,
+      "learning_rate": 8.858834465170576e-06,
+      "loss": 0.2541,
+      "step": 464
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 0.47922995686531067,
+      "learning_rate": 8.851710294440974e-06,
+      "loss": 0.2468,
+      "step": 465
+    },
+    {
+      "epoch": 0.89472,
+      "grad_norm": 0.5384715795516968,
+      "learning_rate": 8.84456683874158e-06,
+      "loss": 0.2246,
+      "step": 466
+    },
+    {
+      "epoch": 0.89664,
+      "grad_norm": 0.6066480875015259,
+      "learning_rate": 8.837404133838667e-06,
+      "loss": 0.2259,
+      "step": 467
+    },
+    {
+      "epoch": 0.89856,
+      "grad_norm": 0.5413900017738342,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.2354,
+      "step": 468
+    },
+    {
+      "epoch": 0.90048,
+      "grad_norm": 0.561143696308136,
+      "learning_rate": 8.823021119969102e-06,
+      "loss": 0.2746,
+      "step": 469
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.5409194827079773,
+      "learning_rate": 8.815800883016168e-06,
+      "loss": 0.2464,
+      "step": 470
+    },
+    {
+      "epoch": 0.90432,
+      "grad_norm": 0.5433072447776794,
+      "learning_rate": 8.808561540886796e-06,
+      "loss": 0.3011,
+      "step": 471
+    },
+    {
+      "epoch": 0.90624,
+      "grad_norm": 0.4825468063354492,
+      "learning_rate": 8.801303129827352e-06,
+      "loss": 0.227,
+      "step": 472
+    },
+    {
+      "epoch": 0.90816,
+      "grad_norm": 0.5159386396408081,
+      "learning_rate": 8.794025686179677e-06,
+      "loss": 0.2256,
+      "step": 473
+    },
+    {
+      "epoch": 0.91008,
+      "grad_norm": 0.5148577094078064,
+      "learning_rate": 8.786729246380901e-06,
+      "loss": 0.2225,
+      "step": 474
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 0.49355366826057434,
+      "learning_rate": 8.779413846963267e-06,
+      "loss": 0.2245,
+      "step": 475
+    },
+    {
+      "epoch": 0.91392,
+      "grad_norm": 0.43090417981147766,
+      "learning_rate": 8.772079524553951e-06,
+      "loss": 0.2016,
+      "step": 476
+    },
+    {
+      "epoch": 0.91584,
+      "grad_norm": 0.5442637801170349,
+      "learning_rate": 8.764726315874872e-06,
+      "loss": 0.2695,
+      "step": 477
+    },
+    {
+      "epoch": 0.91776,
+      "grad_norm": 0.4973200261592865,
+      "learning_rate": 8.757354257742501e-06,
+      "loss": 0.244,
+      "step": 478
+    },
+    {
+      "epoch": 0.91968,
+      "grad_norm": 0.5265247821807861,
+      "learning_rate": 8.749963387067697e-06,
+      "loss": 0.2717,
+      "step": 479
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.5752131938934326,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.2505,
+      "step": 480
+    },
+    {
+      "epoch": 0.92352,
+      "grad_norm": 0.48869258165359497,
+      "learning_rate": 8.735125356204982e-06,
+      "loss": 0.234,
+      "step": 481
+    },
+    {
+      "epoch": 0.92544,
+      "grad_norm": 0.5698112845420837,
+      "learning_rate": 8.727678270308994e-06,
+      "loss": 0.3257,
+      "step": 482
+    },
+    {
+      "epoch": 0.92736,
+      "grad_norm": 0.5946613550186157,
+      "learning_rate": 8.720212520454054e-06,
+      "loss": 0.2591,
+      "step": 483
+    },
+    {
+      "epoch": 0.92928,
+      "grad_norm": 0.581592857837677,
+      "learning_rate": 8.712728144020118e-06,
+      "loss": 0.2401,
+      "step": 484
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 0.5188082456588745,
+      "learning_rate": 8.705225178480397e-06,
+      "loss": 0.2422,
+      "step": 485
+    },
+    {
+      "epoch": 0.93312,
+      "grad_norm": 0.48649245500564575,
+      "learning_rate": 8.697703661401187e-06,
+      "loss": 0.2216,
+      "step": 486
+    },
+    {
+      "epoch": 0.93504,
+      "grad_norm": 0.5194732546806335,
+      "learning_rate": 8.69016363044166e-06,
+      "loss": 0.2647,
+      "step": 487
+    },
+    {
+      "epoch": 0.93696,
+      "grad_norm": 0.46351325511932373,
+      "learning_rate": 8.682605123353685e-06,
+      "loss": 0.2257,
+      "step": 488
+    },
+    {
+      "epoch": 0.93888,
+      "grad_norm": 0.5205086469650269,
+      "learning_rate": 8.675028177981643e-06,
+      "loss": 0.2175,
+      "step": 489
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.5495590567588806,
+      "learning_rate": 8.66743283226223e-06,
+      "loss": 0.225,
+      "step": 490
+    },
+    {
+      "epoch": 0.94272,
+      "grad_norm": 0.5487601161003113,
+      "learning_rate": 8.659819124224266e-06,
+      "loss": 0.2179,
+      "step": 491
+    },
+    {
+      "epoch": 0.94464,
+      "grad_norm": 0.6367857456207275,
+      "learning_rate": 8.652187091988516e-06,
+      "loss": 0.2556,
+      "step": 492
+    },
+    {
+      "epoch": 0.94656,
+      "grad_norm": 0.5416897535324097,
+      "learning_rate": 8.644536773767488e-06,
+      "loss": 0.2508,
+      "step": 493
+    },
+    {
+      "epoch": 0.94848,
+      "grad_norm": 0.5474621653556824,
+      "learning_rate": 8.636868207865244e-06,
+      "loss": 0.2241,
+      "step": 494
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 0.6248787045478821,
+      "learning_rate": 8.629181432677213e-06,
+      "loss": 0.2353,
+      "step": 495
+    },
+    {
+      "epoch": 0.95232,
+      "grad_norm": 0.48837146162986755,
+      "learning_rate": 8.621476486689991e-06,
+      "loss": 0.2552,
+      "step": 496
+    },
+    {
+      "epoch": 0.95424,
+      "grad_norm": 0.5383952260017395,
+      "learning_rate": 8.613753408481158e-06,
+      "loss": 0.3039,
+      "step": 497
+    },
+    {
+      "epoch": 0.95616,
+      "grad_norm": 0.47824954986572266,
+      "learning_rate": 8.606012236719073e-06,
+      "loss": 0.2456,
+      "step": 498
+    },
+    {
+      "epoch": 0.95808,
+      "grad_norm": 0.551753044128418,
+      "learning_rate": 8.598253010162693e-06,
+      "loss": 0.2224,
+      "step": 499
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.587669849395752,
+      "learning_rate": 8.590475767661371e-06,
+      "loss": 0.2396,
+      "step": 500
+    },
+    {
+      "epoch": 0.96192,
+      "grad_norm": 0.509161651134491,
+      "learning_rate": 8.58268054815466e-06,
+      "loss": 0.2371,
+      "step": 501
+    },
+    {
+      "epoch": 0.96384,
+      "grad_norm": 0.4888247847557068,
+      "learning_rate": 8.574867390672124e-06,
+      "loss": 0.253,
+      "step": 502
+    },
+    {
+      "epoch": 0.96576,
+      "grad_norm": 0.5514271259307861,
+      "learning_rate": 8.567036334333142e-06,
+      "loss": 0.3291,
+      "step": 503
+    },
+    {
+      "epoch": 0.96768,
+      "grad_norm": 0.6007022857666016,
+      "learning_rate": 8.559187418346703e-06,
+      "loss": 0.2468,
+      "step": 504
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 0.5416086316108704,
+      "learning_rate": 8.551320682011227e-06,
+      "loss": 0.2499,
+      "step": 505
+    },
+    {
+      "epoch": 0.97152,
+      "grad_norm": 0.48627182841300964,
+      "learning_rate": 8.543436164714351e-06,
+      "loss": 0.2367,
+      "step": 506
+    },
+    {
+      "epoch": 0.97344,
+      "grad_norm": 0.5021848678588867,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.2502,
+      "step": 507
+    },
+    {
+      "epoch": 0.97536,
+      "grad_norm": 0.4916617274284363,
+      "learning_rate": 8.527613945231886e-06,
+      "loss": 0.2552,
+      "step": 508
+    },
+    {
+      "epoch": 0.97728,
+      "grad_norm": 0.5162515640258789,
+      "learning_rate": 8.519676322265914e-06,
+      "loss": 0.2094,
+      "step": 509
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.4565170407295227,
+      "learning_rate": 8.511721076777388e-06,
+      "loss": 0.2023,
+      "step": 510
+    },
+    {
+      "epoch": 0.98112,
+      "grad_norm": 0.5166844725608826,
+      "learning_rate": 8.503748248597095e-06,
+      "loss": 0.2355,
+      "step": 511
+    },
+    {
+      "epoch": 0.98304,
+      "grad_norm": 0.5048564076423645,
+      "learning_rate": 8.495757877643857e-06,
+      "loss": 0.2603,
+      "step": 512
+    },
+    {
+      "epoch": 0.98496,
+      "grad_norm": 0.48669204115867615,
+      "learning_rate": 8.487750003924341e-06,
+      "loss": 0.2243,
+      "step": 513
+    },
+    {
+      "epoch": 0.98688,
+      "grad_norm": 0.5194842219352722,
+      "learning_rate": 8.479724667532836e-06,
+      "loss": 0.2454,
+      "step": 514
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 0.4637187719345093,
+      "learning_rate": 8.471681908651067e-06,
+      "loss": 0.2731,
+      "step": 515
+    },
+    {
+      "epoch": 0.99072,
+      "grad_norm": 0.5181217789649963,
+      "learning_rate": 8.463621767547998e-06,
+      "loss": 0.2466,
+      "step": 516
+    },
+    {
+      "epoch": 0.99264,
+      "grad_norm": 0.542394757270813,
+      "learning_rate": 8.455544284579614e-06,
+      "loss": 0.2964,
+      "step": 517
+    },
+    {
+      "epoch": 0.99456,
+      "grad_norm": 0.588638961315155,
+      "learning_rate": 8.447449500188731e-06,
+      "loss": 0.2284,
+      "step": 518
+    },
+    {
+      "epoch": 0.99648,
+      "grad_norm": 0.5160729289054871,
+      "learning_rate": 8.439337454904794e-06,
+      "loss": 0.2533,
+      "step": 519
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.554336667060852,
+      "learning_rate": 8.43120818934367e-06,
+      "loss": 0.2624,
+      "step": 520
+    },
+    {
+      "epoch": 1.00128,
+      "grad_norm": 0.9654662609100342,
+      "learning_rate": 8.42306174420744e-06,
+      "loss": 0.3766,
+      "step": 521
+    },
+    {
+      "epoch": 1.0032,
+      "grad_norm": 0.4895855784416199,
+      "learning_rate": 8.414898160284208e-06,
+      "loss": 0.2299,
+      "step": 522
+    },
+    {
+      "epoch": 1.00512,
+      "grad_norm": 0.49350422620773315,
+      "learning_rate": 8.406717478447889e-06,
+      "loss": 0.2069,
+      "step": 523
+    },
+    {
+      "epoch": 1.00704,
+      "grad_norm": 0.5192241072654724,
+      "learning_rate": 8.398519739657997e-06,
+      "loss": 0.2519,
+      "step": 524
+    },
+    {
+      "epoch": 1.00896,
+      "grad_norm": 0.4978283643722534,
+      "learning_rate": 8.390304984959455e-06,
+      "loss": 0.1837,
+      "step": 525
+    },
+    {
+      "epoch": 1.01088,
+      "grad_norm": 0.49985983967781067,
+      "learning_rate": 8.382073255482381e-06,
+      "loss": 0.2254,
+      "step": 526
+    },
+    {
+      "epoch": 1.0128,
+      "grad_norm": 0.43738505244255066,
+      "learning_rate": 8.373824592441884e-06,
+      "loss": 0.1789,
+      "step": 527
+    },
+    {
+      "epoch": 1.01472,
+      "grad_norm": 0.5051630735397339,
+      "learning_rate": 8.36555903713785e-06,
+      "loss": 0.1988,
+      "step": 528
+    },
+    {
+      "epoch": 1.01664,
+      "grad_norm": 0.5134169459342957,
+      "learning_rate": 8.357276630954756e-06,
+      "loss": 0.2206,
+      "step": 529
+    },
+    {
+      "epoch": 1.01856,
+      "grad_norm": 0.5272481441497803,
+      "learning_rate": 8.348977415361435e-06,
+      "loss": 0.2044,
+      "step": 530
+    },
+    {
+      "epoch": 1.02048,
+      "grad_norm": 0.45038509368896484,
+      "learning_rate": 8.340661431910888e-06,
+      "loss": 0.1792,
+      "step": 531
+    },
+    {
+      "epoch": 1.0224,
+      "grad_norm": 0.5271627902984619,
+      "learning_rate": 8.332328722240072e-06,
+      "loss": 0.2211,
+      "step": 532
+    },
+    {
+      "epoch": 1.02432,
+      "grad_norm": 0.5080485343933105,
+      "learning_rate": 8.323979328069689e-06,
+      "loss": 0.2055,
+      "step": 533
+    },
+    {
+      "epoch": 1.02624,
+      "grad_norm": 0.45411771535873413,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 0.1872,
+      "step": 534
+    },
+    {
+      "epoch": 1.02816,
+      "grad_norm": 0.5783286690711975,
+      "learning_rate": 8.307230653530501e-06,
+      "loss": 0.2539,
+      "step": 535
+    },
+    {
+      "epoch": 1.03008,
+      "grad_norm": 0.6029261946678162,
+      "learning_rate": 8.298831457019943e-06,
+      "loss": 0.279,
+      "step": 536
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 0.4312973618507385,
+      "learning_rate": 8.290415743725894e-06,
+      "loss": 0.198,
+      "step": 537
+    },
+    {
+      "epoch": 1.03392,
+      "grad_norm": 0.4706169068813324,
+      "learning_rate": 8.28198355578465e-06,
+      "loss": 0.1898,
+      "step": 538
+    },
+    {
+      "epoch": 1.03584,
+      "grad_norm": 0.47474080324172974,
+      "learning_rate": 8.273534935414975e-06,
+      "loss": 0.1899,
+      "step": 539
+    },
+    {
+      "epoch": 1.03776,
+      "grad_norm": 0.5079405903816223,
+      "learning_rate": 8.265069924917925e-06,
+      "loss": 0.2382,
+      "step": 540
+    },
+    {
+      "epoch": 1.03968,
+      "grad_norm": 0.5403974652290344,
+      "learning_rate": 8.256588566676616e-06,
+      "loss": 0.237,
+      "step": 541
+    },
+    {
+      "epoch": 1.0416,
+      "grad_norm": 0.5268903970718384,
+      "learning_rate": 8.248090903156003e-06,
+      "loss": 0.2136,
+      "step": 542
+    },
+    {
+      "epoch": 1.04352,
+      "grad_norm": 0.49097713828086853,
+      "learning_rate": 8.239576976902694e-06,
+      "loss": 0.189,
+      "step": 543
+    },
+    {
+      "epoch": 1.04544,
+      "grad_norm": 0.4670626223087311,
+      "learning_rate": 8.231046830544716e-06,
+      "loss": 0.195,
+      "step": 544
+    },
+    {
+      "epoch": 1.04736,
+      "grad_norm": 0.455961674451828,
+      "learning_rate": 8.222500506791305e-06,
+      "loss": 0.2217,
+      "step": 545
+    },
+    {
+      "epoch": 1.04928,
+      "grad_norm": 0.4674827456474304,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.1884,
+      "step": 546
+    },
+    {
+      "epoch": 1.0512,
+      "grad_norm": 0.4885239005088806,
+      "learning_rate": 8.205359498339917e-06,
+      "loss": 0.2123,
+      "step": 547
+    },
+    {
+      "epoch": 1.05312,
+      "grad_norm": 0.4880794286727905,
+      "learning_rate": 8.196764899464552e-06,
+      "loss": 0.1991,
+      "step": 548
+    },
+    {
+      "epoch": 1.05504,
+      "grad_norm": 0.49276429414749146,
+      "learning_rate": 8.188154294838542e-06,
+      "loss": 0.2262,
+      "step": 549
+    },
+    {
+      "epoch": 1.05696,
+      "grad_norm": 0.5290385484695435,
+      "learning_rate": 8.179527727573975e-06,
+      "loss": 0.2298,
+      "step": 550
+    },
+    {
+      "epoch": 1.05888,
+      "grad_norm": 0.5113620758056641,
+      "learning_rate": 8.170885240862854e-06,
+      "loss": 0.2189,
+      "step": 551
+    },
+    {
+      "epoch": 1.0608,
+      "grad_norm": 0.4960134029388428,
+      "learning_rate": 8.162226877976886e-06,
+      "loss": 0.1991,
+      "step": 552
+    },
+    {
+      "epoch": 1.06272,
+      "grad_norm": 0.5470779538154602,
+      "learning_rate": 8.153552682267278e-06,
+      "loss": 0.2621,
+      "step": 553
+    },
+    {
+      "epoch": 1.06464,
+      "grad_norm": 0.44238218665122986,
+      "learning_rate": 8.144862697164499e-06,
+      "loss": 0.1746,
+      "step": 554
+    },
+    {
+      "epoch": 1.06656,
+      "grad_norm": 0.4669162631034851,
+      "learning_rate": 8.136156966178082e-06,
+      "loss": 0.2101,
+      "step": 555
+    },
+    {
+      "epoch": 1.06848,
+      "grad_norm": 0.5901086926460266,
+      "learning_rate": 8.127435532896388e-06,
+      "loss": 0.2788,
+      "step": 556
+    },
+    {
+      "epoch": 1.0704,
+      "grad_norm": 0.5696573853492737,
+      "learning_rate": 8.118698440986405e-06,
+      "loss": 0.2037,
+      "step": 557
+    },
+    {
+      "epoch": 1.07232,
+      "grad_norm": 0.5250774025917053,
+      "learning_rate": 8.10994573419352e-06,
+      "loss": 0.2203,
+      "step": 558
+    },
+    {
+      "epoch": 1.07424,
+      "grad_norm": 0.4473564922809601,
+      "learning_rate": 8.101177456341301e-06,
+      "loss": 0.1808,
+      "step": 559
+    },
+    {
+      "epoch": 1.07616,
+      "grad_norm": 0.5792028307914734,
+      "learning_rate": 8.092393651331275e-06,
+      "loss": 0.2583,
+      "step": 560
+    },
+    {
+      "epoch": 1.07808,
+      "grad_norm": 0.4896470606327057,
+      "learning_rate": 8.083594363142717e-06,
+      "loss": 0.2008,
+      "step": 561
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5601317286491394,
+      "learning_rate": 8.074779635832417e-06,
+      "loss": 0.217,
+      "step": 562
+    },
+    {
+      "epoch": 1.08192,
+      "grad_norm": 0.547504186630249,
+      "learning_rate": 8.065949513534474e-06,
+      "loss": 0.2279,
+      "step": 563
+    },
+    {
+      "epoch": 1.08384,
+      "grad_norm": 0.5030434727668762,
+      "learning_rate": 8.057104040460062e-06,
+      "loss": 0.209,
+      "step": 564
+    },
+    {
+      "epoch": 1.08576,
+      "grad_norm": 0.5093943476676941,
+      "learning_rate": 8.048243260897216e-06,
+      "loss": 0.204,
+      "step": 565
+    },
+    {
+      "epoch": 1.08768,
+      "grad_norm": 0.4664313793182373,
+      "learning_rate": 8.03936721921061e-06,
+      "loss": 0.1756,
+      "step": 566
+    },
+    {
+      "epoch": 1.0896,
+      "grad_norm": 0.4682534337043762,
+      "learning_rate": 8.030475959841333e-06,
+      "loss": 0.1811,
+      "step": 567
+    },
+    {
+      "epoch": 1.09152,
+      "grad_norm": 0.5074787735939026,
+      "learning_rate": 8.021569527306663e-06,
+      "loss": 0.2104,
+      "step": 568
+    },
+    {
+      "epoch": 1.09344,
+      "grad_norm": 0.5077551007270813,
+      "learning_rate": 8.012647966199851e-06,
+      "loss": 0.2706,
+      "step": 569
+    },
+    {
+      "epoch": 1.09536,
+      "grad_norm": 0.4922370910644531,
+      "learning_rate": 8.003711321189895e-06,
+      "loss": 0.1908,
+      "step": 570
+    },
+    {
+      "epoch": 1.09728,
+      "grad_norm": 0.48884233832359314,
+      "learning_rate": 7.994759637021316e-06,
+      "loss": 0.187,
+      "step": 571
+    },
+    {
+      "epoch": 1.0992,
+      "grad_norm": 0.49399569630622864,
+      "learning_rate": 7.985792958513932e-06,
+      "loss": 0.2224,
+      "step": 572
+    },
+    {
+      "epoch": 1.10112,
+      "grad_norm": 0.535283088684082,
+      "learning_rate": 7.976811330562637e-06,
+      "loss": 0.2344,
+      "step": 573
+    },
+    {
+      "epoch": 1.10304,
+      "grad_norm": 0.48860377073287964,
+      "learning_rate": 7.967814798137173e-06,
+      "loss": 0.2028,
+      "step": 574
+    },
+    {
+      "epoch": 1.10496,
+      "grad_norm": 0.4875934422016144,
+      "learning_rate": 7.95880340628191e-06,
+      "loss": 0.1986,
+      "step": 575
+    },
+    {
+      "epoch": 1.10688,
+      "grad_norm": 0.47198566794395447,
+      "learning_rate": 7.949777200115617e-06,
+      "loss": 0.1832,
+      "step": 576
+    },
+    {
+      "epoch": 1.1088,
+      "grad_norm": 0.5035191774368286,
+      "learning_rate": 7.940736224831227e-06,
+      "loss": 0.2068,
+      "step": 577
+    },
+    {
+      "epoch": 1.11072,
+      "grad_norm": 0.4756562411785126,
+      "learning_rate": 7.931680525695634e-06,
+      "loss": 0.1864,
+      "step": 578
+    },
+    {
+      "epoch": 1.11264,
+      "grad_norm": 0.5011635422706604,
+      "learning_rate": 7.922610148049445e-06,
+      "loss": 0.2686,
+      "step": 579
+    },
+    {
+      "epoch": 1.11456,
+      "grad_norm": 0.48313355445861816,
+      "learning_rate": 7.913525137306756e-06,
+      "loss": 0.1976,
+      "step": 580
+    },
+    {
+      "epoch": 1.11648,
+      "grad_norm": 0.46709144115448,
+      "learning_rate": 7.90442553895494e-06,
+      "loss": 0.1752,
+      "step": 581
+    },
+    {
+      "epoch": 1.1184,
+      "grad_norm": 0.48863455653190613,
+      "learning_rate": 7.895311398554395e-06,
+      "loss": 0.1754,
+      "step": 582
+    },
+    {
+      "epoch": 1.12032,
+      "grad_norm": 0.47324907779693604,
+      "learning_rate": 7.886182761738339e-06,
+      "loss": 0.1863,
+      "step": 583
+    },
+    {
+      "epoch": 1.12224,
+      "grad_norm": 0.5840582847595215,
+      "learning_rate": 7.877039674212569e-06,
+      "loss": 0.2956,
+      "step": 584
+    },
+    {
+      "epoch": 1.12416,
+      "grad_norm": 0.4763057231903076,
+      "learning_rate": 7.86788218175523e-06,
+      "loss": 0.2057,
+      "step": 585
+    },
+    {
+      "epoch": 1.12608,
+      "grad_norm": 0.5128911137580872,
+      "learning_rate": 7.8587103302166e-06,
+      "loss": 0.2347,
+      "step": 586
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 0.4510782063007355,
+      "learning_rate": 7.849524165518838e-06,
+      "loss": 0.1936,
+      "step": 587
+    },
+    {
+      "epoch": 1.12992,
+      "grad_norm": 0.5071703195571899,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.2296,
+      "step": 588
+    },
+    {
+      "epoch": 1.13184,
+      "grad_norm": 0.5359281897544861,
+      "learning_rate": 7.831109080692687e-06,
+      "loss": 0.2006,
+      "step": 589
+    },
+    {
+      "epoch": 1.13376,
+      "grad_norm": 0.5238724946975708,
+      "learning_rate": 7.821880252766025e-06,
+      "loss": 0.1901,
+      "step": 590
+    },
+    {
+      "epoch": 1.13568,
+      "grad_norm": 0.5808878540992737,
+      "learning_rate": 7.812637296083235e-06,
+      "loss": 0.2201,
+      "step": 591
+    },
+    {
+      "epoch": 1.1376,
+      "grad_norm": 0.5040211081504822,
+      "learning_rate": 7.803380256922495e-06,
+      "loss": 0.2011,
+      "step": 592
+    },
+    {
+      "epoch": 1.13952,
+      "grad_norm": 0.5709764361381531,
+      "learning_rate": 7.794109181632494e-06,
+      "loss": 0.2278,
+      "step": 593
+    },
+    {
+      "epoch": 1.14144,
+      "grad_norm": 0.5645810961723328,
+      "learning_rate": 7.784824116632198e-06,
+      "loss": 0.2544,
+      "step": 594
+    },
+    {
+      "epoch": 1.14336,
+      "grad_norm": 0.4434546232223511,
+      "learning_rate": 7.775525108410616e-06,
+      "loss": 0.1591,
+      "step": 595
+    },
+    {
+      "epoch": 1.14528,
+      "grad_norm": 0.5892537236213684,
+      "learning_rate": 7.76621220352657e-06,
+      "loss": 0.282,
+      "step": 596
+    },
+    {
+      "epoch": 1.1472,
+      "grad_norm": 0.465692937374115,
+      "learning_rate": 7.75688544860846e-06,
+      "loss": 0.2215,
+      "step": 597
+    },
+    {
+      "epoch": 1.14912,
+      "grad_norm": 0.46853163838386536,
+      "learning_rate": 7.747544890354031e-06,
+      "loss": 0.1798,
+      "step": 598
+    },
+    {
+      "epoch": 1.15104,
+      "grad_norm": 0.5773459672927856,
+      "learning_rate": 7.73819057553014e-06,
+      "loss": 0.2441,
+      "step": 599
+    },
+    {
+      "epoch": 1.15296,
+      "grad_norm": 0.5152280926704407,
+      "learning_rate": 7.728822550972523e-06,
+      "loss": 0.2015,
+      "step": 600
+    },
+    {
+      "epoch": 1.15488,
+      "grad_norm": 0.5448644757270813,
+      "learning_rate": 7.719440863585555e-06,
+      "loss": 0.2246,
+      "step": 601
+    },
+    {
+      "epoch": 1.1568,
+      "grad_norm": 0.5351273417472839,
+      "learning_rate": 7.710045560342021e-06,
+      "loss": 0.2242,
+      "step": 602
+    },
+    {
+      "epoch": 1.15872,
+      "grad_norm": 0.4712618887424469,
+      "learning_rate": 7.700636688282878e-06,
+      "loss": 0.1831,
+      "step": 603
+    },
+    {
+      "epoch": 1.16064,
+      "grad_norm": 0.5316839814186096,
+      "learning_rate": 7.69121429451702e-06,
+      "loss": 0.2339,
+      "step": 604
+    },
+    {
+      "epoch": 1.16256,
+      "grad_norm": 0.5002859830856323,
+      "learning_rate": 7.681778426221043e-06,
+      "loss": 0.215,
+      "step": 605
+    },
+    {
+      "epoch": 1.16448,
+      "grad_norm": 0.4991815984249115,
+      "learning_rate": 7.672329130639007e-06,
+      "loss": 0.212,
+      "step": 606
+    },
+    {
+      "epoch": 1.1663999999999999,
+      "grad_norm": 0.591431736946106,
+      "learning_rate": 7.662866455082201e-06,
+      "loss": 0.2858,
+      "step": 607
+    },
+    {
+      "epoch": 1.16832,
+      "grad_norm": 0.5187620520591736,
+      "learning_rate": 7.65339044692891e-06,
+      "loss": 0.2037,
+      "step": 608
+    },
+    {
+      "epoch": 1.17024,
+      "grad_norm": 0.5149185061454773,
+      "learning_rate": 7.643901153624163e-06,
+      "loss": 0.1978,
+      "step": 609
+    },
+    {
+      "epoch": 1.17216,
+      "grad_norm": 0.5005959272384644,
+      "learning_rate": 7.634398622679517e-06,
+      "loss": 0.2141,
+      "step": 610
+    },
+    {
+      "epoch": 1.17408,
+      "grad_norm": 0.4626937210559845,
+      "learning_rate": 7.624882901672801e-06,
+      "loss": 0.2045,
+      "step": 611
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 0.547808825969696,
+      "learning_rate": 7.615354038247889e-06,
+      "loss": 0.224,
+      "step": 612
+    },
+    {
+      "epoch": 1.17792,
+      "grad_norm": 0.538033127784729,
+      "learning_rate": 7.605812080114453e-06,
+      "loss": 0.2219,
+      "step": 613
+    },
+    {
+      "epoch": 1.17984,
+      "grad_norm": 0.42939358949661255,
+      "learning_rate": 7.59625707504773e-06,
+      "loss": 0.1815,
+      "step": 614
+    },
+    {
+      "epoch": 1.18176,
+      "grad_norm": 0.5268927216529846,
+      "learning_rate": 7.586689070888284e-06,
+      "loss": 0.2253,
+      "step": 615
+    },
+    {
+      "epoch": 1.18368,
+      "grad_norm": 0.5008851289749146,
+      "learning_rate": 7.577108115541761e-06,
+      "loss": 0.1933,
+      "step": 616
+    },
+    {
+      "epoch": 1.1856,
+      "grad_norm": 0.5203779339790344,
+      "learning_rate": 7.567514256978652e-06,
+      "loss": 0.226,
+      "step": 617
+    },
+    {
+      "epoch": 1.18752,
+      "grad_norm": 0.4966503381729126,
+      "learning_rate": 7.557907543234051e-06,
+      "loss": 0.1928,
+      "step": 618
+    },
+    {
+      "epoch": 1.18944,
+      "grad_norm": 0.5570352077484131,
+      "learning_rate": 7.54828802240742e-06,
+      "loss": 0.2004,
+      "step": 619
+    },
+    {
+      "epoch": 1.19136,
+      "grad_norm": 0.43986955285072327,
+      "learning_rate": 7.53865574266234e-06,
+      "loss": 0.2009,
+      "step": 620
+    },
+    {
+      "epoch": 1.1932800000000001,
+      "grad_norm": 0.5081701874732971,
+      "learning_rate": 7.529010752226277e-06,
+      "loss": 0.2101,
+      "step": 621
+    },
+    {
+      "epoch": 1.1952,
+      "grad_norm": 0.5489816069602966,
+      "learning_rate": 7.519353099390336e-06,
+      "loss": 0.2233,
+      "step": 622
+    },
+    {
+      "epoch": 1.19712,
+      "grad_norm": 0.4732995629310608,
+      "learning_rate": 7.509682832509024e-06,
+      "loss": 0.203,
+      "step": 623
+    },
+    {
+      "epoch": 1.19904,
+      "grad_norm": 0.5080229640007019,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2354,
+      "step": 624
+    },
+    {
+      "epoch": 1.20096,
+      "grad_norm": 0.4721149802207947,
+      "learning_rate": 7.490304650343842e-06,
+      "loss": 0.1818,
+      "step": 625
+    },
+    {
+      "epoch": 1.20288,
+      "grad_norm": 0.5268194079399109,
+      "learning_rate": 7.480596832083795e-06,
+      "loss": 0.2275,
+      "step": 626
+    },
+    {
+      "epoch": 1.2048,
+      "grad_norm": 0.4850090742111206,
+      "learning_rate": 7.470876593825535e-06,
+      "loss": 0.2039,
+      "step": 627
+    },
+    {
+      "epoch": 1.20672,
+      "grad_norm": 0.4863957166671753,
+      "learning_rate": 7.461143984236925e-06,
+      "loss": 0.2008,
+      "step": 628
+    },
+    {
+      "epoch": 1.20864,
+      "grad_norm": 0.4906001687049866,
+      "learning_rate": 7.451399052047764e-06,
+      "loss": 0.2291,
+      "step": 629
+    },
+    {
+      "epoch": 1.21056,
+      "grad_norm": 0.45427659153938293,
+      "learning_rate": 7.441641846049557e-06,
+      "loss": 0.2143,
+      "step": 630
+    },
+    {
+      "epoch": 1.21248,
+      "grad_norm": 0.47992178797721863,
+      "learning_rate": 7.431872415095252e-06,
+      "loss": 0.2113,
+      "step": 631
+    },
+    {
+      "epoch": 1.2144,
+      "grad_norm": 0.49418720602989197,
+      "learning_rate": 7.422090808099014e-06,
+      "loss": 0.2247,
+      "step": 632
+    },
+    {
+      "epoch": 1.21632,
+      "grad_norm": 0.45458993315696716,
+      "learning_rate": 7.412297074035968e-06,
+      "loss": 0.1938,
+      "step": 633
+    },
+    {
+      "epoch": 1.21824,
+      "grad_norm": 0.5675350427627563,
+      "learning_rate": 7.402491261941958e-06,
+      "loss": 0.2761,
+      "step": 634
+    },
+    {
+      "epoch": 1.22016,
+      "grad_norm": 0.47387605905532837,
+      "learning_rate": 7.3926734209133e-06,
+      "loss": 0.2162,
+      "step": 635
+    },
+    {
+      "epoch": 1.22208,
+      "grad_norm": 0.535197377204895,
+      "learning_rate": 7.382843600106539e-06,
+      "loss": 0.2645,
+      "step": 636
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 0.434085488319397,
+      "learning_rate": 7.373001848738203e-06,
+      "loss": 0.173,
+      "step": 637
+    },
+    {
+      "epoch": 1.22592,
+      "grad_norm": 0.5035478472709656,
+      "learning_rate": 7.363148216084548e-06,
+      "loss": 0.1955,
+      "step": 638
+    },
+    {
+      "epoch": 1.22784,
+      "grad_norm": 0.5237541198730469,
+      "learning_rate": 7.353282751481324e-06,
+      "loss": 0.2246,
+      "step": 639
+    },
+    {
+      "epoch": 1.22976,
+      "grad_norm": 0.5198637247085571,
+      "learning_rate": 7.343405504323519e-06,
+      "loss": 0.2323,
+      "step": 640
+    },
+    {
+      "epoch": 1.2316799999999999,
+      "grad_norm": 0.4697301983833313,
+      "learning_rate": 7.333516524065116e-06,
+      "loss": 0.1962,
+      "step": 641
+    },
+    {
+      "epoch": 1.2336,
+      "grad_norm": 0.5421521067619324,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.2349,
+      "step": 642
+    },
+    {
+      "epoch": 1.23552,
+      "grad_norm": 0.47649693489074707,
+      "learning_rate": 7.3137035623559274e-06,
+      "loss": 0.2001,
+      "step": 643
+    },
+    {
+      "epoch": 1.23744,
+      "grad_norm": 0.5248587131500244,
+      "learning_rate": 7.303779680105844e-06,
+      "loss": 0.2127,
+      "step": 644
+    },
+    {
+      "epoch": 1.23936,
+      "grad_norm": 0.492305189371109,
+      "learning_rate": 7.2938442631560714e-06,
+      "loss": 0.204,
+      "step": 645
+    },
+    {
+      "epoch": 1.24128,
+      "grad_norm": 0.4846341013908386,
+      "learning_rate": 7.28389736125184e-06,
+      "loss": 0.2174,
+      "step": 646
+    },
+    {
+      "epoch": 1.2432,
+      "grad_norm": 0.523556113243103,
+      "learning_rate": 7.273939024195881e-06,
+      "loss": 0.2029,
+      "step": 647
+    },
+    {
+      "epoch": 1.24512,
+      "grad_norm": 0.5016922950744629,
+      "learning_rate": 7.263969301848188e-06,
+      "loss": 0.1964,
+      "step": 648
+    },
+    {
+      "epoch": 1.24704,
+      "grad_norm": 0.5252081751823425,
+      "learning_rate": 7.2539882441257484e-06,
+      "loss": 0.2091,
+      "step": 649
+    },
+    {
+      "epoch": 1.24896,
+      "grad_norm": 0.5014604926109314,
+      "learning_rate": 7.243995901002312e-06,
+      "loss": 0.2005,
+      "step": 650
+    },
+    {
+      "epoch": 1.25088,
+      "grad_norm": 0.46460098028182983,
+      "learning_rate": 7.2339923225081296e-06,
+      "loss": 0.1929,
+      "step": 651
+    },
+    {
+      "epoch": 1.2528000000000001,
+      "grad_norm": 0.5700141191482544,
+      "learning_rate": 7.223977558729707e-06,
+      "loss": 0.2456,
+      "step": 652
+    },
+    {
+      "epoch": 1.25472,
+      "grad_norm": 0.5112634301185608,
+      "learning_rate": 7.213951659809553e-06,
+      "loss": 0.1907,
+      "step": 653
+    },
+    {
+      "epoch": 1.25664,
+      "grad_norm": 0.5293146967887878,
+      "learning_rate": 7.203914675945929e-06,
+      "loss": 0.2469,
+      "step": 654
+    },
+    {
+      "epoch": 1.2585600000000001,
+      "grad_norm": 0.5058817267417908,
+      "learning_rate": 7.193866657392597e-06,
+      "loss": 0.2424,
+      "step": 655
+    },
+    {
+      "epoch": 1.26048,
+      "grad_norm": 0.4793136715888977,
+      "learning_rate": 7.183807654458565e-06,
+      "loss": 0.1807,
+      "step": 656
+    },
+    {
+      "epoch": 1.2624,
+      "grad_norm": 0.5089419484138489,
+      "learning_rate": 7.173737717507843e-06,
+      "loss": 0.2257,
+      "step": 657
+    },
+    {
+      "epoch": 1.26432,
+      "grad_norm": 0.44122233986854553,
+      "learning_rate": 7.163656896959181e-06,
+      "loss": 0.1767,
+      "step": 658
+    },
+    {
+      "epoch": 1.26624,
+      "grad_norm": 0.5154277682304382,
+      "learning_rate": 7.153565243285829e-06,
+      "loss": 0.2037,
+      "step": 659
+    },
+    {
+      "epoch": 1.26816,
+      "grad_norm": 0.47802266478538513,
+      "learning_rate": 7.143462807015271e-06,
+      "loss": 0.2116,
+      "step": 660
+    },
+    {
+      "epoch": 1.27008,
+      "grad_norm": 0.5024216175079346,
+      "learning_rate": 7.133349638728979e-06,
+      "loss": 0.2488,
+      "step": 661
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 0.5174927115440369,
+      "learning_rate": 7.1232257890621605e-06,
+      "loss": 0.2548,
+      "step": 662
+    },
+    {
+      "epoch": 1.27392,
+      "grad_norm": 0.4883042871952057,
+      "learning_rate": 7.113091308703498e-06,
+      "loss": 0.1895,
+      "step": 663
+    },
+    {
+      "epoch": 1.27584,
+      "grad_norm": 0.490737646818161,
+      "learning_rate": 7.102946248394908e-06,
+      "loss": 0.204,
+      "step": 664
+    },
+    {
+      "epoch": 1.27776,
+      "grad_norm": 0.5285569429397583,
+      "learning_rate": 7.092790658931274e-06,
+      "loss": 0.1934,
+      "step": 665
+    },
+    {
+      "epoch": 1.27968,
+      "grad_norm": 0.5232217311859131,
+      "learning_rate": 7.082624591160201e-06,
+      "loss": 0.2317,
+      "step": 666
+    },
+    {
+      "epoch": 1.2816,
+      "grad_norm": 0.5300791263580322,
+      "learning_rate": 7.072448095981751e-06,
+      "loss": 0.2405,
+      "step": 667
+    },
+    {
+      "epoch": 1.28352,
+      "grad_norm": 0.5607843399047852,
+      "learning_rate": 7.0622612243482035e-06,
+      "loss": 0.2122,
+      "step": 668
+    },
+    {
+      "epoch": 1.28544,
+      "grad_norm": 0.4322912096977234,
+      "learning_rate": 7.052064027263785e-06,
+      "loss": 0.1812,
+      "step": 669
+    },
+    {
+      "epoch": 1.28736,
+      "grad_norm": 0.6258375644683838,
+      "learning_rate": 7.041856555784421e-06,
+      "loss": 0.2603,
+      "step": 670
+    },
+    {
+      "epoch": 1.28928,
+      "grad_norm": 0.5203710198402405,
+      "learning_rate": 7.031638861017484e-06,
+      "loss": 0.2291,
+      "step": 671
+    },
+    {
+      "epoch": 1.2912,
+      "grad_norm": 0.5006266236305237,
+      "learning_rate": 7.021410994121525e-06,
+      "loss": 0.1884,
+      "step": 672
+    },
+    {
+      "epoch": 1.29312,
+      "grad_norm": 0.5167715549468994,
+      "learning_rate": 7.011173006306034e-06,
+      "loss": 0.1807,
+      "step": 673
+    },
+    {
+      "epoch": 1.29504,
+      "grad_norm": 0.5537475943565369,
+      "learning_rate": 7.0009249488311685e-06,
+      "loss": 0.2245,
+      "step": 674
+    },
+    {
+      "epoch": 1.29696,
+      "grad_norm": 0.4899687170982361,
+      "learning_rate": 6.990666873007506e-06,
+      "loss": 0.2052,
+      "step": 675
+    },
+    {
+      "epoch": 1.29888,
+      "grad_norm": 0.4782716929912567,
+      "learning_rate": 6.980398830195785e-06,
+      "loss": 0.2162,
+      "step": 676
+    },
+    {
+      "epoch": 1.3008,
+      "grad_norm": 0.5198002457618713,
+      "learning_rate": 6.970120871806647e-06,
+      "loss": 0.228,
+      "step": 677
+    },
+    {
+      "epoch": 1.3027199999999999,
+      "grad_norm": 0.5188785791397095,
+      "learning_rate": 6.959833049300376e-06,
+      "loss": 0.2253,
+      "step": 678
+    },
+    {
+      "epoch": 1.30464,
+      "grad_norm": 0.5002376437187195,
+      "learning_rate": 6.949535414186651e-06,
+      "loss": 0.2097,
+      "step": 679
+    },
+    {
+      "epoch": 1.30656,
+      "grad_norm": 0.4720045030117035,
+      "learning_rate": 6.939228018024275e-06,
+      "loss": 0.217,
+      "step": 680
+    },
+    {
+      "epoch": 1.3084799999999999,
+      "grad_norm": 0.5313807129859924,
+      "learning_rate": 6.9289109124209276e-06,
+      "loss": 0.2259,
+      "step": 681
+    },
+    {
+      "epoch": 1.3104,
+      "grad_norm": 0.5444822907447815,
+      "learning_rate": 6.918584149032899e-06,
+      "loss": 0.2219,
+      "step": 682
+    },
+    {
+      "epoch": 1.31232,
+      "grad_norm": 0.5565977096557617,
+      "learning_rate": 6.908247779564836e-06,
+      "loss": 0.2477,
+      "step": 683
+    },
+    {
+      "epoch": 1.31424,
+      "grad_norm": 0.5039553642272949,
+      "learning_rate": 6.897901855769483e-06,
+      "loss": 0.1955,
+      "step": 684
+    },
+    {
+      "epoch": 1.31616,
+      "grad_norm": 0.5093384981155396,
+      "learning_rate": 6.887546429447418e-06,
+      "loss": 0.2227,
+      "step": 685
+    },
+    {
+      "epoch": 1.31808,
+      "grad_norm": 0.5090653300285339,
+      "learning_rate": 6.8771815524468e-06,
+      "loss": 0.2007,
+      "step": 686
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.504021167755127,
+      "learning_rate": 6.8668072766631054e-06,
+      "loss": 0.2623,
+      "step": 687
+    },
+    {
+      "epoch": 1.32192,
+      "grad_norm": 0.44561517238616943,
+      "learning_rate": 6.856423654038868e-06,
+      "loss": 0.1497,
+      "step": 688
+    },
+    {
+      "epoch": 1.3238400000000001,
+      "grad_norm": 0.5367900133132935,
+      "learning_rate": 6.8460307365634225e-06,
+      "loss": 0.2624,
+      "step": 689
+    },
+    {
+      "epoch": 1.32576,
+      "grad_norm": 0.516869068145752,
+      "learning_rate": 6.8356285762726385e-06,
+      "loss": 0.2123,
+      "step": 690
+    },
+    {
+      "epoch": 1.32768,
+      "grad_norm": 0.4735253155231476,
+      "learning_rate": 6.825217225248664e-06,
+      "loss": 0.2158,
+      "step": 691
+    },
+    {
+      "epoch": 1.3296000000000001,
+      "grad_norm": 0.5434602499008179,
+      "learning_rate": 6.814796735619664e-06,
+      "loss": 0.2308,
+      "step": 692
+    },
+    {
+      "epoch": 1.33152,
+      "grad_norm": 0.45334556698799133,
+      "learning_rate": 6.804367159559561e-06,
+      "loss": 0.2064,
+      "step": 693
+    },
+    {
+      "epoch": 1.33344,
+      "grad_norm": 0.4540878236293793,
+      "learning_rate": 6.793928549287767e-06,
+      "loss": 0.1923,
+      "step": 694
+    },
+    {
+      "epoch": 1.33536,
+      "grad_norm": 0.4330199062824249,
+      "learning_rate": 6.783480957068934e-06,
+      "loss": 0.1841,
+      "step": 695
+    },
+    {
+      "epoch": 1.33728,
+      "grad_norm": 0.4893876314163208,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.2047,
+      "step": 696
+    },
+    {
+      "epoch": 1.3392,
+      "grad_norm": 0.5151920914649963,
+      "learning_rate": 6.762559036073331e-06,
+      "loss": 0.225,
+      "step": 697
+    },
+    {
+      "epoch": 1.34112,
+      "grad_norm": 0.5023897290229797,
+      "learning_rate": 6.75208481204967e-06,
+      "loss": 0.1892,
+      "step": 698
+    },
+    {
+      "epoch": 1.34304,
+      "grad_norm": 0.5250627398490906,
+      "learning_rate": 6.74160181558465e-06,
+      "loss": 0.2499,
+      "step": 699
+    },
+    {
+      "epoch": 1.34496,
+      "grad_norm": 0.5386905074119568,
+      "learning_rate": 6.731110099165165e-06,
+      "loss": 0.1994,
+      "step": 700
+    },
+    {
+      "epoch": 1.34688,
+      "grad_norm": 0.5110102891921997,
+      "learning_rate": 6.7206097153217474e-06,
+      "loss": 0.1807,
+      "step": 701
+    },
+    {
+      "epoch": 1.3488,
+      "grad_norm": 0.5553632378578186,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.2064,
+      "step": 702
+    },
+    {
+      "epoch": 1.35072,
+      "grad_norm": 0.5036051869392395,
+      "learning_rate": 6.699583155702025e-06,
+      "loss": 0.2046,
+      "step": 703
+    },
+    {
+      "epoch": 1.35264,
+      "grad_norm": 0.5022258758544922,
+      "learning_rate": 6.689057085202737e-06,
+      "loss": 0.1924,
+      "step": 704
+    },
+    {
+      "epoch": 1.35456,
+      "grad_norm": 0.5286658406257629,
+      "learning_rate": 6.678522557833025e-06,
+      "loss": 0.2453,
+      "step": 705
+    },
+    {
+      "epoch": 1.35648,
+      "grad_norm": 0.5339585542678833,
+      "learning_rate": 6.667979626337783e-06,
+      "loss": 0.2637,
+      "step": 706
+    },
+    {
+      "epoch": 1.3584,
+      "grad_norm": 0.5099043250083923,
+      "learning_rate": 6.6574283435039795e-06,
+      "loss": 0.2273,
+      "step": 707
+    },
+    {
+      "epoch": 1.36032,
+      "grad_norm": 0.4986157715320587,
+      "learning_rate": 6.646868762160399e-06,
+      "loss": 0.1986,
+      "step": 708
+    },
+    {
+      "epoch": 1.36224,
+      "grad_norm": 0.533613920211792,
+      "learning_rate": 6.6363009351773755e-06,
+      "loss": 0.2479,
+      "step": 709
+    },
+    {
+      "epoch": 1.36416,
+      "grad_norm": 0.4837549924850464,
+      "learning_rate": 6.625724915466526e-06,
+      "loss": 0.2033,
+      "step": 710
+    },
+    {
+      "epoch": 1.36608,
+      "grad_norm": 0.5586143136024475,
+      "learning_rate": 6.615140755980491e-06,
+      "loss": 0.2384,
+      "step": 711
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 0.4786967635154724,
+      "learning_rate": 6.6045485097126585e-06,
+      "loss": 0.1937,
+      "step": 712
+    },
+    {
+      "epoch": 1.36992,
+      "grad_norm": 0.48212671279907227,
+      "learning_rate": 6.593948229696915e-06,
+      "loss": 0.1803,
+      "step": 713
+    },
+    {
+      "epoch": 1.37184,
+      "grad_norm": 0.499746173620224,
+      "learning_rate": 6.583339969007364e-06,
+      "loss": 0.2306,
+      "step": 714
+    },
+    {
+      "epoch": 1.3737599999999999,
+      "grad_norm": 0.5464677810668945,
+      "learning_rate": 6.572723780758069e-06,
+      "loss": 0.2444,
+      "step": 715
+    },
+    {
+      "epoch": 1.37568,
+      "grad_norm": 0.5543858408927917,
+      "learning_rate": 6.562099718102788e-06,
+      "loss": 0.2709,
+      "step": 716
+    },
+    {
+      "epoch": 1.3776,
+      "grad_norm": 0.5112447738647461,
+      "learning_rate": 6.551467834234702e-06,
+      "loss": 0.2331,
+      "step": 717
+    },
+    {
+      "epoch": 1.37952,
+      "grad_norm": 0.5704824924468994,
+      "learning_rate": 6.540828182386154e-06,
+      "loss": 0.2661,
+      "step": 718
+    },
+    {
+      "epoch": 1.38144,
+      "grad_norm": 0.5187730193138123,
+      "learning_rate": 6.530180815828377e-06,
+      "loss": 0.2424,
+      "step": 719
+    },
+    {
+      "epoch": 1.38336,
+      "grad_norm": 0.4762147068977356,
+      "learning_rate": 6.519525787871235e-06,
+      "loss": 0.2065,
+      "step": 720
+    },
+    {
+      "epoch": 1.38528,
+      "grad_norm": 0.5206378698348999,
+      "learning_rate": 6.508863151862947e-06,
+      "loss": 0.1909,
+      "step": 721
+    },
+    {
+      "epoch": 1.3872,
+      "grad_norm": 0.5740947127342224,
+      "learning_rate": 6.498192961189829e-06,
+      "loss": 0.2337,
+      "step": 722
+    },
+    {
+      "epoch": 1.3891200000000001,
+      "grad_norm": 0.4958505928516388,
+      "learning_rate": 6.487515269276015e-06,
+      "loss": 0.1977,
+      "step": 723
+    },
+    {
+      "epoch": 1.39104,
+      "grad_norm": 0.5559083819389343,
+      "learning_rate": 6.476830129583207e-06,
+      "loss": 0.2396,
+      "step": 724
+    },
+    {
+      "epoch": 1.39296,
+      "grad_norm": 0.4593888223171234,
+      "learning_rate": 6.466137595610388e-06,
+      "loss": 0.1605,
+      "step": 725
+    },
+    {
+      "epoch": 1.3948800000000001,
+      "grad_norm": 0.5182091593742371,
+      "learning_rate": 6.455437720893565e-06,
+      "loss": 0.2248,
+      "step": 726
+    },
+    {
+      "epoch": 1.3968,
+      "grad_norm": 0.48563438653945923,
+      "learning_rate": 6.444730559005504e-06,
+      "loss": 0.2207,
+      "step": 727
+    },
+    {
+      "epoch": 1.39872,
+      "grad_norm": 0.49449673295021057,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 0.208,
+      "step": 728
+    },
+    {
+      "epoch": 1.40064,
+      "grad_norm": 0.5345655679702759,
+      "learning_rate": 6.423294588188874e-06,
+      "loss": 0.2647,
+      "step": 729
+    },
+    {
+      "epoch": 1.40256,
+      "grad_norm": 0.4944145083427429,
+      "learning_rate": 6.412565886587186e-06,
+      "loss": 0.2075,
+      "step": 730
+    },
+    {
+      "epoch": 1.40448,
+      "grad_norm": 0.4769936203956604,
+      "learning_rate": 6.401830112467479e-06,
+      "loss": 0.2013,
+      "step": 731
+    },
+    {
+      "epoch": 1.4064,
+      "grad_norm": 0.41314756870269775,
+      "learning_rate": 6.391087319582264e-06,
+      "loss": 0.1583,
+      "step": 732
+    },
+    {
+      "epoch": 1.40832,
+      "grad_norm": 0.5208045244216919,
+      "learning_rate": 6.380337561719184e-06,
+      "loss": 0.2128,
+      "step": 733
+    },
+    {
+      "epoch": 1.41024,
+      "grad_norm": 0.5039541125297546,
+      "learning_rate": 6.369580892700763e-06,
+      "loss": 0.2295,
+      "step": 734
+    },
+    {
+      "epoch": 1.41216,
+      "grad_norm": 0.490587443113327,
+      "learning_rate": 6.358817366384122e-06,
+      "loss": 0.1838,
+      "step": 735
+    },
+    {
+      "epoch": 1.41408,
+      "grad_norm": 0.5714592933654785,
+      "learning_rate": 6.34804703666072e-06,
+      "loss": 0.2463,
+      "step": 736
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 0.559694766998291,
+      "learning_rate": 6.337269957456074e-06,
+      "loss": 0.2724,
+      "step": 737
+    },
+    {
+      "epoch": 1.41792,
+      "grad_norm": 0.5104078054428101,
+      "learning_rate": 6.326486182729504e-06,
+      "loss": 0.2218,
+      "step": 738
+    },
+    {
+      "epoch": 1.41984,
+      "grad_norm": 0.5027304291725159,
+      "learning_rate": 6.315695766473844e-06,
+      "loss": 0.1988,
+      "step": 739
+    },
+    {
+      "epoch": 1.42176,
+      "grad_norm": 0.4671670198440552,
+      "learning_rate": 6.304898762715187e-06,
+      "loss": 0.1868,
+      "step": 740
+    },
+    {
+      "epoch": 1.42368,
+      "grad_norm": 0.4799177348613739,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.1843,
+      "step": 741
+    },
+    {
+      "epoch": 1.4256,
+      "grad_norm": 0.5107128620147705,
+      "learning_rate": 6.283285208957884e-06,
+      "loss": 0.223,
+      "step": 742
+    },
+    {
+      "epoch": 1.42752,
+      "grad_norm": 0.5050696134567261,
+      "learning_rate": 6.272468767175255e-06,
+      "loss": 0.2361,
+      "step": 743
+    },
+    {
+      "epoch": 1.42944,
+      "grad_norm": 0.4882921576499939,
+      "learning_rate": 6.261645954321109e-06,
+      "loss": 0.2009,
+      "step": 744
+    },
+    {
+      "epoch": 1.43136,
+      "grad_norm": 0.4591202735900879,
+      "learning_rate": 6.2508168245837476e-06,
+      "loss": 0.1877,
+      "step": 745
+    },
+    {
+      "epoch": 1.4332799999999999,
+      "grad_norm": 0.5136224627494812,
+      "learning_rate": 6.239981432183093e-06,
+      "loss": 0.2582,
+      "step": 746
+    },
+    {
+      "epoch": 1.4352,
+      "grad_norm": 0.52320396900177,
+      "learning_rate": 6.2291398313704265e-06,
+      "loss": 0.2614,
+      "step": 747
+    },
+    {
+      "epoch": 1.43712,
+      "grad_norm": 0.5025537014007568,
+      "learning_rate": 6.21829207642811e-06,
+      "loss": 0.2443,
+      "step": 748
+    },
+    {
+      "epoch": 1.4390399999999999,
+      "grad_norm": 0.4769178330898285,
+      "learning_rate": 6.207438221669325e-06,
+      "loss": 0.189,
+      "step": 749
+    },
+    {
+      "epoch": 1.44096,
+      "grad_norm": 0.4796161949634552,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.1896,
+      "step": 750
+    },
+    {
+      "epoch": 1.44288,
+      "grad_norm": 0.5188882946968079,
+      "learning_rate": 6.185712430107489e-06,
+      "loss": 0.2003,
+      "step": 751
+    },
+    {
+      "epoch": 1.4447999999999999,
+      "grad_norm": 0.506746768951416,
+      "learning_rate": 6.1748406020824115e-06,
+      "loss": 0.2195,
+      "step": 752
+    },
+    {
+      "epoch": 1.44672,
+      "grad_norm": 0.5100806355476379,
+      "learning_rate": 6.163962891796261e-06,
+      "loss": 0.2307,
+      "step": 753
+    },
+    {
+      "epoch": 1.44864,
+      "grad_norm": 0.520645022392273,
+      "learning_rate": 6.153079353712201e-06,
+      "loss": 0.2067,
+      "step": 754
+    },
+    {
+      "epoch": 1.45056,
+      "grad_norm": 0.5536027550697327,
+      "learning_rate": 6.142190042322569e-06,
+      "loss": 0.2551,
+      "step": 755
+    },
+    {
+      "epoch": 1.45248,
+      "grad_norm": 0.5364271998405457,
+      "learning_rate": 6.131295012148613e-06,
+      "loss": 0.2458,
+      "step": 756
+    },
+    {
+      "epoch": 1.4544000000000001,
+      "grad_norm": 0.5415770411491394,
+      "learning_rate": 6.120394317740205e-06,
+      "loss": 0.2181,
+      "step": 757
+    },
+    {
+      "epoch": 1.45632,
+      "grad_norm": 0.5311259031295776,
+      "learning_rate": 6.1094880136755886e-06,
+      "loss": 0.2294,
+      "step": 758
+    },
+    {
+      "epoch": 1.45824,
+      "grad_norm": 0.4757941663265228,
+      "learning_rate": 6.0985761545610865e-06,
+      "loss": 0.208,
+      "step": 759
+    },
+    {
+      "epoch": 1.4601600000000001,
+      "grad_norm": 0.5190828442573547,
+      "learning_rate": 6.087658795030838e-06,
+      "loss": 0.221,
+      "step": 760
+    },
+    {
+      "epoch": 1.46208,
+      "grad_norm": 0.5155487656593323,
+      "learning_rate": 6.076735989746522e-06,
+      "loss": 0.2188,
+      "step": 761
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 0.5949140787124634,
+      "learning_rate": 6.065807793397081e-06,
+      "loss": 0.2728,
+      "step": 762
+    },
+    {
+      "epoch": 1.4659200000000001,
+      "grad_norm": 0.5376107096672058,
+      "learning_rate": 6.0548742606984545e-06,
+      "loss": 0.2242,
+      "step": 763
+    },
+    {
+      "epoch": 1.46784,
+      "grad_norm": 0.4967236816883087,
+      "learning_rate": 6.043935446393294e-06,
+      "loss": 0.1974,
+      "step": 764
+    },
+    {
+      "epoch": 1.46976,
+      "grad_norm": 0.5123416781425476,
+      "learning_rate": 6.032991405250702e-06,
+      "loss": 0.2005,
+      "step": 765
+    },
+    {
+      "epoch": 1.47168,
+      "grad_norm": 0.47030726075172424,
+      "learning_rate": 6.022042192065946e-06,
+      "loss": 0.1964,
+      "step": 766
+    },
+    {
+      "epoch": 1.4736,
+      "grad_norm": 0.5985661149024963,
+      "learning_rate": 6.011087861660191e-06,
+      "loss": 0.2513,
+      "step": 767
+    },
+    {
+      "epoch": 1.47552,
+      "grad_norm": 0.537276566028595,
+      "learning_rate": 6.000128468880223e-06,
+      "loss": 0.232,
+      "step": 768
+    },
+    {
+      "epoch": 1.47744,
+      "grad_norm": 0.48686933517456055,
+      "learning_rate": 5.989164068598175e-06,
+      "loss": 0.1777,
+      "step": 769
+    },
+    {
+      "epoch": 1.47936,
+      "grad_norm": 0.47734299302101135,
+      "learning_rate": 5.978194715711254e-06,
+      "loss": 0.1899,
+      "step": 770
+    },
+    {
+      "epoch": 1.48128,
+      "grad_norm": 0.49828436970710754,
+      "learning_rate": 5.9672204651414564e-06,
+      "loss": 0.2223,
+      "step": 771
+    },
+    {
+      "epoch": 1.4832,
+      "grad_norm": 0.5040861964225769,
+      "learning_rate": 5.956241371835312e-06,
+      "loss": 0.2069,
+      "step": 772
+    },
+    {
+      "epoch": 1.48512,
+      "grad_norm": 0.47698572278022766,
+      "learning_rate": 5.945257490763588e-06,
+      "loss": 0.1909,
+      "step": 773
+    },
+    {
+      "epoch": 1.48704,
+      "grad_norm": 0.48523035645484924,
+      "learning_rate": 5.934268876921026e-06,
+      "loss": 0.2095,
+      "step": 774
+    },
+    {
+      "epoch": 1.48896,
+      "grad_norm": 0.5115752220153809,
+      "learning_rate": 5.923275585326064e-06,
+      "loss": 0.2485,
+      "step": 775
+    },
+    {
+      "epoch": 1.49088,
+      "grad_norm": 0.48026034235954285,
+      "learning_rate": 5.912277671020564e-06,
+      "loss": 0.1918,
+      "step": 776
+    },
+    {
+      "epoch": 1.4928,
+      "grad_norm": 0.5084309577941895,
+      "learning_rate": 5.90127518906953e-06,
+      "loss": 0.2218,
+      "step": 777
+    },
+    {
+      "epoch": 1.49472,
+      "grad_norm": 0.5267378091812134,
+      "learning_rate": 5.890268194560834e-06,
+      "loss": 0.2318,
+      "step": 778
+    },
+    {
+      "epoch": 1.49664,
+      "grad_norm": 0.5183507204055786,
+      "learning_rate": 5.8792567426049465e-06,
+      "loss": 0.2295,
+      "step": 779
+    },
+    {
+      "epoch": 1.49856,
+      "grad_norm": 0.4647071361541748,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.2116,
+      "step": 780
+    },
+    {
+      "epoch": 1.50048,
+      "grad_norm": 0.4547733664512634,
+      "learning_rate": 5.857220686904779e-06,
+      "loss": 0.1644,
+      "step": 781
+    },
+    {
+      "epoch": 1.5024,
+      "grad_norm": 0.5043164491653442,
+      "learning_rate": 5.846196193491919e-06,
+      "loss": 0.2171,
+      "step": 782
+    },
+    {
+      "epoch": 1.5043199999999999,
+      "grad_norm": 0.4805786907672882,
+      "learning_rate": 5.835167463294155e-06,
+      "loss": 0.202,
+      "step": 783
+    },
+    {
+      "epoch": 1.50624,
+      "grad_norm": 0.5299478769302368,
+      "learning_rate": 5.824134551530783e-06,
+      "loss": 0.2282,
+      "step": 784
+    },
+    {
+      "epoch": 1.50816,
+      "grad_norm": 0.5086731314659119,
+      "learning_rate": 5.813097513442035e-06,
+      "loss": 0.2121,
+      "step": 785
+    },
+    {
+      "epoch": 1.5100799999999999,
+      "grad_norm": 0.5129131078720093,
+      "learning_rate": 5.8020564042888015e-06,
+      "loss": 0.2303,
+      "step": 786
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 0.45654988288879395,
+      "learning_rate": 5.79101127935236e-06,
+      "loss": 0.2047,
+      "step": 787
+    },
+    {
+      "epoch": 1.5139200000000002,
+      "grad_norm": 0.5027675032615662,
+      "learning_rate": 5.77996219393409e-06,
+      "loss": 0.2206,
+      "step": 788
+    },
+    {
+      "epoch": 1.5158399999999999,
+      "grad_norm": 0.4700727164745331,
+      "learning_rate": 5.768909203355203e-06,
+      "loss": 0.1799,
+      "step": 789
+    },
+    {
+      "epoch": 1.51776,
+      "grad_norm": 0.6292695999145508,
+      "learning_rate": 5.757852362956463e-06,
+      "loss": 0.2955,
+      "step": 790
+    },
+    {
+      "epoch": 1.5196800000000001,
+      "grad_norm": 0.4593273103237152,
+      "learning_rate": 5.7467917280979105e-06,
+      "loss": 0.1928,
+      "step": 791
+    },
+    {
+      "epoch": 1.5215999999999998,
+      "grad_norm": 0.5011586546897888,
+      "learning_rate": 5.735727354158581e-06,
+      "loss": 0.2179,
+      "step": 792
+    },
+    {
+      "epoch": 1.52352,
+      "grad_norm": 0.4886990785598755,
+      "learning_rate": 5.724659296536234e-06,
+      "loss": 0.2008,
+      "step": 793
+    },
+    {
+      "epoch": 1.5254400000000001,
+      "grad_norm": 0.4429033696651459,
+      "learning_rate": 5.713587610647073e-06,
+      "loss": 0.1859,
+      "step": 794
+    },
+    {
+      "epoch": 1.52736,
+      "grad_norm": 0.4847736656665802,
+      "learning_rate": 5.7025123519254644e-06,
+      "loss": 0.2014,
+      "step": 795
+    },
+    {
+      "epoch": 1.52928,
+      "grad_norm": 0.5115492343902588,
+      "learning_rate": 5.6914335758236665e-06,
+      "loss": 0.2326,
+      "step": 796
+    },
+    {
+      "epoch": 1.5312000000000001,
+      "grad_norm": 0.4668499827384949,
+      "learning_rate": 5.680351337811547e-06,
+      "loss": 0.2056,
+      "step": 797
+    },
+    {
+      "epoch": 1.53312,
+      "grad_norm": 0.512101411819458,
+      "learning_rate": 5.669265693376309e-06,
+      "loss": 0.2491,
+      "step": 798
+    },
+    {
+      "epoch": 1.53504,
+      "grad_norm": 0.48129650950431824,
+      "learning_rate": 5.658176698022208e-06,
+      "loss": 0.208,
+      "step": 799
+    },
+    {
+      "epoch": 1.53696,
+      "grad_norm": 0.5290973782539368,
+      "learning_rate": 5.647084407270277e-06,
+      "loss": 0.2221,
+      "step": 800
+    },
+    {
+      "epoch": 1.53888,
+      "grad_norm": 0.5432623028755188,
+      "learning_rate": 5.6359888766580555e-06,
+      "loss": 0.2234,
+      "step": 801
+    },
+    {
+      "epoch": 1.5408,
+      "grad_norm": 0.4756374657154083,
+      "learning_rate": 5.624890161739292e-06,
+      "loss": 0.1834,
+      "step": 802
+    },
+    {
+      "epoch": 1.54272,
+      "grad_norm": 0.4619845449924469,
+      "learning_rate": 5.6137883180836925e-06,
+      "loss": 0.2062,
+      "step": 803
+    },
+    {
+      "epoch": 1.54464,
+      "grad_norm": 0.6713989973068237,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.2048,
+      "step": 804
+    },
+    {
+      "epoch": 1.54656,
+      "grad_norm": 0.5586594343185425,
+      "learning_rate": 5.591575466918816e-06,
+      "loss": 0.2253,
+      "step": 805
+    },
+    {
+      "epoch": 1.54848,
+      "grad_norm": 0.5152693390846252,
+      "learning_rate": 5.5804645706261515e-06,
+      "loss": 0.2095,
+      "step": 806
+    },
+    {
+      "epoch": 1.5504,
+      "grad_norm": 0.5025952458381653,
+      "learning_rate": 5.569350768029312e-06,
+      "loss": 0.2192,
+      "step": 807
+    },
+    {
+      "epoch": 1.55232,
+      "grad_norm": 0.49874231219291687,
+      "learning_rate": 5.5582341147735396e-06,
+      "loss": 0.207,
+      "step": 808
+    },
+    {
+      "epoch": 1.55424,
+      "grad_norm": 0.5105839967727661,
+      "learning_rate": 5.5471146665183455e-06,
+      "loss": 0.2615,
+      "step": 809
+    },
+    {
+      "epoch": 1.55616,
+      "grad_norm": 0.48718419671058655,
+      "learning_rate": 5.53599247893724e-06,
+      "loss": 0.1986,
+      "step": 810
+    },
+    {
+      "epoch": 1.55808,
+      "grad_norm": 0.512748658657074,
+      "learning_rate": 5.524867607717445e-06,
+      "loss": 0.2333,
+      "step": 811
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.48858407139778137,
+      "learning_rate": 5.5137401085596224e-06,
+      "loss": 0.221,
+      "step": 812
+    },
+    {
+      "epoch": 1.56192,
+      "grad_norm": 0.43382564187049866,
+      "learning_rate": 5.502610037177586e-06,
+      "loss": 0.1737,
+      "step": 813
+    },
+    {
+      "epoch": 1.56384,
+      "grad_norm": 0.46070921421051025,
+      "learning_rate": 5.491477449298036e-06,
+      "loss": 0.2,
+      "step": 814
+    },
+    {
+      "epoch": 1.56576,
+      "grad_norm": 0.42012134194374084,
+      "learning_rate": 5.480342400660268e-06,
+      "loss": 0.1705,
+      "step": 815
+    },
+    {
+      "epoch": 1.56768,
+      "grad_norm": 0.5009737014770508,
+      "learning_rate": 5.469204947015897e-06,
+      "loss": 0.2184,
+      "step": 816
+    },
+    {
+      "epoch": 1.5695999999999999,
+      "grad_norm": 0.4881472587585449,
+      "learning_rate": 5.458065144128584e-06,
+      "loss": 0.2166,
+      "step": 817
+    },
+    {
+      "epoch": 1.57152,
+      "grad_norm": 0.4771557152271271,
+      "learning_rate": 5.4469230477737466e-06,
+      "loss": 0.1974,
+      "step": 818
+    },
+    {
+      "epoch": 1.57344,
+      "grad_norm": 0.4701177775859833,
+      "learning_rate": 5.435778713738292e-06,
+      "loss": 0.1951,
+      "step": 819
+    },
+    {
+      "epoch": 1.5753599999999999,
+      "grad_norm": 0.494834840297699,
+      "learning_rate": 5.424632197820325e-06,
+      "loss": 0.1984,
+      "step": 820
+    },
+    {
+      "epoch": 1.57728,
+      "grad_norm": 0.4886401295661926,
+      "learning_rate": 5.413483555828879e-06,
+      "loss": 0.1866,
+      "step": 821
+    },
+    {
+      "epoch": 1.5792000000000002,
+      "grad_norm": 0.5240247845649719,
+      "learning_rate": 5.402332843583631e-06,
+      "loss": 0.2207,
+      "step": 822
+    },
+    {
+      "epoch": 1.5811199999999999,
+      "grad_norm": 0.5362030863761902,
+      "learning_rate": 5.391180116914621e-06,
+      "loss": 0.2235,
+      "step": 823
+    },
+    {
+      "epoch": 1.58304,
+      "grad_norm": 0.47925835847854614,
+      "learning_rate": 5.380025431661981e-06,
+      "loss": 0.2052,
+      "step": 824
+    },
+    {
+      "epoch": 1.5849600000000001,
+      "grad_norm": 0.5050221085548401,
+      "learning_rate": 5.368868843675642e-06,
+      "loss": 0.2441,
+      "step": 825
+    },
+    {
+      "epoch": 1.5868799999999998,
+      "grad_norm": 0.495376318693161,
+      "learning_rate": 5.3577104088150685e-06,
+      "loss": 0.2028,
+      "step": 826
+    },
+    {
+      "epoch": 1.5888,
+      "grad_norm": 0.46564653515815735,
+      "learning_rate": 5.346550182948966e-06,
+      "loss": 0.1966,
+      "step": 827
+    },
+    {
+      "epoch": 1.5907200000000001,
+      "grad_norm": 0.48003560304641724,
+      "learning_rate": 5.335388221955012e-06,
+      "loss": 0.2132,
+      "step": 828
+    },
+    {
+      "epoch": 1.5926399999999998,
+      "grad_norm": 0.5184383988380432,
+      "learning_rate": 5.3242245817195705e-06,
+      "loss": 0.2135,
+      "step": 829
+    },
+    {
+      "epoch": 1.59456,
+      "grad_norm": 0.47808152437210083,
+      "learning_rate": 5.31305931813741e-06,
+      "loss": 0.1912,
+      "step": 830
+    },
+    {
+      "epoch": 1.5964800000000001,
+      "grad_norm": 0.4853704571723938,
+      "learning_rate": 5.301892487111431e-06,
+      "loss": 0.184,
+      "step": 831
+    },
+    {
+      "epoch": 1.5984,
+      "grad_norm": 0.5342936515808105,
+      "learning_rate": 5.290724144552379e-06,
+      "loss": 0.2107,
+      "step": 832
+    },
+    {
+      "epoch": 1.60032,
+      "grad_norm": 0.49120697379112244,
+      "learning_rate": 5.279554346378572e-06,
+      "loss": 0.213,
+      "step": 833
+    },
+    {
+      "epoch": 1.60224,
+      "grad_norm": 0.5054184198379517,
+      "learning_rate": 5.268383148515608e-06,
+      "loss": 0.2449,
+      "step": 834
+    },
+    {
+      "epoch": 1.60416,
+      "grad_norm": 0.4645128846168518,
+      "learning_rate": 5.257210606896102e-06,
+      "loss": 0.1951,
+      "step": 835
+    },
+    {
+      "epoch": 1.60608,
+      "grad_norm": 0.4538768231868744,
+      "learning_rate": 5.246036777459391e-06,
+      "loss": 0.1809,
+      "step": 836
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 0.48349300026893616,
+      "learning_rate": 5.234861716151264e-06,
+      "loss": 0.2251,
+      "step": 837
+    },
+    {
+      "epoch": 1.60992,
+      "grad_norm": 0.5083431005477905,
+      "learning_rate": 5.223685478923671e-06,
+      "loss": 0.2191,
+      "step": 838
+    },
+    {
+      "epoch": 1.61184,
+      "grad_norm": 0.5216970443725586,
+      "learning_rate": 5.2125081217344595e-06,
+      "loss": 0.2156,
+      "step": 839
+    },
+    {
+      "epoch": 1.61376,
+      "grad_norm": 0.4621428847312927,
+      "learning_rate": 5.201329700547077e-06,
+      "loss": 0.1818,
+      "step": 840
+    },
+    {
+      "epoch": 1.61568,
+      "grad_norm": 0.5295419692993164,
+      "learning_rate": 5.190150271330298e-06,
+      "loss": 0.2638,
+      "step": 841
+    },
+    {
+      "epoch": 1.6176,
+      "grad_norm": 0.569073498249054,
+      "learning_rate": 5.178969890057953e-06,
+      "loss": 0.2606,
+      "step": 842
+    },
+    {
+      "epoch": 1.61952,
+      "grad_norm": 0.44174954295158386,
+      "learning_rate": 5.167788612708627e-06,
+      "loss": 0.1842,
+      "step": 843
+    },
+    {
+      "epoch": 1.62144,
+      "grad_norm": 0.4608680009841919,
+      "learning_rate": 5.156606495265402e-06,
+      "loss": 0.2135,
+      "step": 844
+    },
+    {
+      "epoch": 1.62336,
+      "grad_norm": 0.515315592288971,
+      "learning_rate": 5.145423593715558e-06,
+      "loss": 0.2212,
+      "step": 845
+    },
+    {
+      "epoch": 1.62528,
+      "grad_norm": 0.4829460680484772,
+      "learning_rate": 5.1342399640503074e-06,
+      "loss": 0.189,
+      "step": 846
+    },
+    {
+      "epoch": 1.6272,
+      "grad_norm": 0.5139299035072327,
+      "learning_rate": 5.1230556622645026e-06,
+      "loss": 0.2061,
+      "step": 847
+    },
+    {
+      "epoch": 1.62912,
+      "grad_norm": 0.5167098045349121,
+      "learning_rate": 5.111870744356366e-06,
+      "loss": 0.2195,
+      "step": 848
+    },
+    {
+      "epoch": 1.63104,
+      "grad_norm": 0.42873433232307434,
+      "learning_rate": 5.100685266327202e-06,
+      "loss": 0.1882,
+      "step": 849
+    },
+    {
+      "epoch": 1.63296,
+      "grad_norm": 0.454526424407959,
+      "learning_rate": 5.0894992841811216e-06,
+      "loss": 0.2018,
+      "step": 850
+    },
+    {
+      "epoch": 1.6348799999999999,
+      "grad_norm": 0.47184333205223083,
+      "learning_rate": 5.0783128539247585e-06,
+      "loss": 0.1809,
+      "step": 851
+    },
+    {
+      "epoch": 1.6368,
+      "grad_norm": 0.5137799382209778,
+      "learning_rate": 5.067126031566988e-06,
+      "loss": 0.227,
+      "step": 852
+    },
+    {
+      "epoch": 1.63872,
+      "grad_norm": 0.44022536277770996,
+      "learning_rate": 5.055938873118653e-06,
+      "loss": 0.1601,
+      "step": 853
+    },
+    {
+      "epoch": 1.6406399999999999,
+      "grad_norm": 0.5104033946990967,
+      "learning_rate": 5.044751434592274e-06,
+      "loss": 0.1967,
+      "step": 854
+    },
+    {
+      "epoch": 1.64256,
+      "grad_norm": 0.4647650420665741,
+      "learning_rate": 5.033563772001782e-06,
+      "loss": 0.1981,
+      "step": 855
+    },
+    {
+      "epoch": 1.6444800000000002,
+      "grad_norm": 0.533186137676239,
+      "learning_rate": 5.022375941362218e-06,
+      "loss": 0.2346,
+      "step": 856
+    },
+    {
+      "epoch": 1.6463999999999999,
+      "grad_norm": 0.5240260362625122,
+      "learning_rate": 5.011187998689474e-06,
+      "loss": 0.2305,
+      "step": 857
+    },
+    {
+      "epoch": 1.64832,
+      "grad_norm": 0.4535600543022156,
+      "learning_rate": 5e-06,
+      "loss": 0.197,
+      "step": 858
+    },
+    {
+      "epoch": 1.6502400000000002,
+      "grad_norm": 0.4647574722766876,
+      "learning_rate": 4.988812001310528e-06,
+      "loss": 0.2282,
+      "step": 859
+    },
+    {
+      "epoch": 1.6521599999999999,
+      "grad_norm": 0.45036518573760986,
+      "learning_rate": 4.977624058637783e-06,
+      "loss": 0.1959,
+      "step": 860
+    },
+    {
+      "epoch": 1.65408,
+      "grad_norm": 0.49932432174682617,
+      "learning_rate": 4.9664362279982205e-06,
+      "loss": 0.2066,
+      "step": 861
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 0.5592161417007446,
+      "learning_rate": 4.955248565407727e-06,
+      "loss": 0.2432,
+      "step": 862
+    },
+    {
+      "epoch": 1.6579199999999998,
+      "grad_norm": 0.4920986294746399,
+      "learning_rate": 4.944061126881348e-06,
+      "loss": 0.2119,
+      "step": 863
+    },
+    {
+      "epoch": 1.65984,
+      "grad_norm": 0.45344191789627075,
+      "learning_rate": 4.932873968433014e-06,
+      "loss": 0.164,
+      "step": 864
+    },
+    {
+      "epoch": 1.6617600000000001,
+      "grad_norm": 0.5338271856307983,
+      "learning_rate": 4.921687146075244e-06,
+      "loss": 0.2195,
+      "step": 865
+    },
+    {
+      "epoch": 1.66368,
+      "grad_norm": 0.4213380217552185,
+      "learning_rate": 4.910500715818879e-06,
+      "loss": 0.1751,
+      "step": 866
+    },
+    {
+      "epoch": 1.6656,
+      "grad_norm": 0.5420383214950562,
+      "learning_rate": 4.8993147336728e-06,
+      "loss": 0.2587,
+      "step": 867
+    },
+    {
+      "epoch": 1.6675200000000001,
+      "grad_norm": 0.5241522192955017,
+      "learning_rate": 4.8881292556436355e-06,
+      "loss": 0.2454,
+      "step": 868
+    },
+    {
+      "epoch": 1.66944,
+      "grad_norm": 0.5246589183807373,
+      "learning_rate": 4.876944337735499e-06,
+      "loss": 0.217,
+      "step": 869
+    },
+    {
+      "epoch": 1.67136,
+      "grad_norm": 0.5251557230949402,
+      "learning_rate": 4.865760035949695e-06,
+      "loss": 0.1943,
+      "step": 870
+    },
+    {
+      "epoch": 1.67328,
+      "grad_norm": 0.5081140995025635,
+      "learning_rate": 4.854576406284443e-06,
+      "loss": 0.2124,
+      "step": 871
+    },
+    {
+      "epoch": 1.6752,
+      "grad_norm": 0.4514849781990051,
+      "learning_rate": 4.8433935047346e-06,
+      "loss": 0.1772,
+      "step": 872
+    },
+    {
+      "epoch": 1.67712,
+      "grad_norm": 0.49481576681137085,
+      "learning_rate": 4.832211387291374e-06,
+      "loss": 0.2121,
+      "step": 873
+    },
+    {
+      "epoch": 1.67904,
+      "grad_norm": 0.5115485191345215,
+      "learning_rate": 4.821030109942048e-06,
+      "loss": 0.2313,
+      "step": 874
+    },
+    {
+      "epoch": 1.68096,
+      "grad_norm": 0.5210694074630737,
+      "learning_rate": 4.8098497286697024e-06,
+      "loss": 0.2014,
+      "step": 875
+    },
+    {
+      "epoch": 1.68288,
+      "grad_norm": 0.5102753043174744,
+      "learning_rate": 4.798670299452926e-06,
+      "loss": 0.2047,
+      "step": 876
+    },
+    {
+      "epoch": 1.6848,
+      "grad_norm": 0.44327253103256226,
+      "learning_rate": 4.787491878265542e-06,
+      "loss": 0.2119,
+      "step": 877
+    },
+    {
+      "epoch": 1.68672,
+      "grad_norm": 0.47956904768943787,
+      "learning_rate": 4.77631452107633e-06,
+      "loss": 0.2117,
+      "step": 878
+    },
+    {
+      "epoch": 1.68864,
+      "grad_norm": 0.5541418790817261,
+      "learning_rate": 4.765138283848739e-06,
+      "loss": 0.248,
+      "step": 879
+    },
+    {
+      "epoch": 1.69056,
+      "grad_norm": 0.4431523084640503,
+      "learning_rate": 4.75396322254061e-06,
+      "loss": 0.1863,
+      "step": 880
+    },
+    {
+      "epoch": 1.69248,
+      "grad_norm": 0.528711199760437,
+      "learning_rate": 4.742789393103899e-06,
+      "loss": 0.2416,
+      "step": 881
+    },
+    {
+      "epoch": 1.6944,
+      "grad_norm": 0.469567209482193,
+      "learning_rate": 4.731616851484392e-06,
+      "loss": 0.2092,
+      "step": 882
+    },
+    {
+      "epoch": 1.69632,
+      "grad_norm": 0.45672881603240967,
+      "learning_rate": 4.720445653621429e-06,
+      "loss": 0.2034,
+      "step": 883
+    },
+    {
+      "epoch": 1.69824,
+      "grad_norm": 0.43577560782432556,
+      "learning_rate": 4.7092758554476215e-06,
+      "loss": 0.1792,
+      "step": 884
+    },
+    {
+      "epoch": 1.70016,
+      "grad_norm": 0.4950227737426758,
+      "learning_rate": 4.69810751288857e-06,
+      "loss": 0.2098,
+      "step": 885
+    },
+    {
+      "epoch": 1.70208,
+      "grad_norm": 0.4854544699192047,
+      "learning_rate": 4.686940681862591e-06,
+      "loss": 0.198,
+      "step": 886
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 0.5014539957046509,
+      "learning_rate": 4.675775418280432e-06,
+      "loss": 0.2196,
+      "step": 887
+    },
+    {
+      "epoch": 1.7059199999999999,
+      "grad_norm": 0.5592138171195984,
+      "learning_rate": 4.664611778044988e-06,
+      "loss": 0.2254,
+      "step": 888
+    },
+    {
+      "epoch": 1.70784,
+      "grad_norm": 0.47507989406585693,
+      "learning_rate": 4.653449817051035e-06,
+      "loss": 0.1973,
+      "step": 889
+    },
+    {
+      "epoch": 1.70976,
+      "grad_norm": 0.5694088935852051,
+      "learning_rate": 4.642289591184934e-06,
+      "loss": 0.2337,
+      "step": 890
+    },
+    {
+      "epoch": 1.7116799999999999,
+      "grad_norm": 0.5275542140007019,
+      "learning_rate": 4.631131156324359e-06,
+      "loss": 0.2096,
+      "step": 891
+    },
+    {
+      "epoch": 1.7136,
+      "grad_norm": 0.4893650412559509,
+      "learning_rate": 4.619974568338021e-06,
+      "loss": 0.2052,
+      "step": 892
+    },
+    {
+      "epoch": 1.7155200000000002,
+      "grad_norm": 0.5181678533554077,
+      "learning_rate": 4.60881988308538e-06,
+      "loss": 0.2189,
+      "step": 893
+    },
+    {
+      "epoch": 1.7174399999999999,
+      "grad_norm": 0.47760409116744995,
+      "learning_rate": 4.597667156416371e-06,
+      "loss": 0.2318,
+      "step": 894
+    },
+    {
+      "epoch": 1.71936,
+      "grad_norm": 0.4247019290924072,
+      "learning_rate": 4.586516444171123e-06,
+      "loss": 0.1658,
+      "step": 895
+    },
+    {
+      "epoch": 1.7212800000000001,
+      "grad_norm": 0.4383614957332611,
+      "learning_rate": 4.575367802179675e-06,
+      "loss": 0.1806,
+      "step": 896
+    },
+    {
+      "epoch": 1.7231999999999998,
+      "grad_norm": 0.46814900636672974,
+      "learning_rate": 4.564221286261709e-06,
+      "loss": 0.2067,
+      "step": 897
+    },
+    {
+      "epoch": 1.72512,
+      "grad_norm": 0.5234333872795105,
+      "learning_rate": 4.553076952226255e-06,
+      "loss": 0.259,
+      "step": 898
+    },
+    {
+      "epoch": 1.7270400000000001,
+      "grad_norm": 0.4630354642868042,
+      "learning_rate": 4.541934855871417e-06,
+      "loss": 0.2005,
+      "step": 899
+    },
+    {
+      "epoch": 1.72896,
+      "grad_norm": 0.49992188811302185,
+      "learning_rate": 4.530795052984104e-06,
+      "loss": 0.2408,
+      "step": 900
+    },
+    {
+      "epoch": 1.73088,
+      "grad_norm": 0.522808849811554,
+      "learning_rate": 4.519657599339735e-06,
+      "loss": 0.2226,
+      "step": 901
+    },
+    {
+      "epoch": 1.7328000000000001,
+      "grad_norm": 0.559347927570343,
+      "learning_rate": 4.508522550701965e-06,
+      "loss": 0.2525,
+      "step": 902
+    },
+    {
+      "epoch": 1.73472,
+      "grad_norm": 0.5831448435783386,
+      "learning_rate": 4.497389962822416e-06,
+      "loss": 0.2313,
+      "step": 903
+    },
+    {
+      "epoch": 1.73664,
+      "grad_norm": 0.4703871011734009,
+      "learning_rate": 4.48625989144038e-06,
+      "loss": 0.2026,
+      "step": 904
+    },
+    {
+      "epoch": 1.73856,
+      "grad_norm": 0.54423588514328,
+      "learning_rate": 4.475132392282556e-06,
+      "loss": 0.2422,
+      "step": 905
+    },
+    {
+      "epoch": 1.74048,
+      "grad_norm": 0.519375741481781,
+      "learning_rate": 4.464007521062761e-06,
+      "loss": 0.2271,
+      "step": 906
+    },
+    {
+      "epoch": 1.7424,
+      "grad_norm": 0.4715530276298523,
+      "learning_rate": 4.452885333481657e-06,
+      "loss": 0.2129,
+      "step": 907
+    },
+    {
+      "epoch": 1.74432,
+      "grad_norm": 0.5166469812393188,
+      "learning_rate": 4.441765885226462e-06,
+      "loss": 0.2572,
+      "step": 908
+    },
+    {
+      "epoch": 1.74624,
+      "grad_norm": 0.4522540271282196,
+      "learning_rate": 4.4306492319706895e-06,
+      "loss": 0.226,
+      "step": 909
+    },
+    {
+      "epoch": 1.74816,
+      "grad_norm": 0.4760035276412964,
+      "learning_rate": 4.4195354293738484e-06,
+      "loss": 0.1733,
+      "step": 910
+    },
+    {
+      "epoch": 1.75008,
+      "grad_norm": 0.5232778787612915,
+      "learning_rate": 4.4084245330811855e-06,
+      "loss": 0.2415,
+      "step": 911
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 0.5263859629631042,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.2804,
+      "step": 912
+    },
+    {
+      "epoch": 1.75392,
+      "grad_norm": 0.5020578503608704,
+      "learning_rate": 4.386211681916309e-06,
+      "loss": 0.1861,
+      "step": 913
+    },
+    {
+      "epoch": 1.75584,
+      "grad_norm": 0.5134755373001099,
+      "learning_rate": 4.3751098382607084e-06,
+      "loss": 0.2425,
+      "step": 914
+    },
+    {
+      "epoch": 1.75776,
+      "grad_norm": 0.4655511975288391,
+      "learning_rate": 4.364011123341947e-06,
+      "loss": 0.2145,
+      "step": 915
+    },
+    {
+      "epoch": 1.75968,
+      "grad_norm": 0.5031753182411194,
+      "learning_rate": 4.352915592729723e-06,
+      "loss": 0.2078,
+      "step": 916
+    },
+    {
+      "epoch": 1.7616,
+      "grad_norm": 0.4492008686065674,
+      "learning_rate": 4.341823301977794e-06,
+      "loss": 0.1845,
+      "step": 917
+    },
+    {
+      "epoch": 1.76352,
+      "grad_norm": 0.5471025109291077,
+      "learning_rate": 4.330734306623694e-06,
+      "loss": 0.2328,
+      "step": 918
+    },
+    {
+      "epoch": 1.76544,
+      "grad_norm": 0.4387701451778412,
+      "learning_rate": 4.319648662188453e-06,
+      "loss": 0.1993,
+      "step": 919
+    },
+    {
+      "epoch": 1.76736,
+      "grad_norm": 0.4615384042263031,
+      "learning_rate": 4.308566424176336e-06,
+      "loss": 0.187,
+      "step": 920
+    },
+    {
+      "epoch": 1.76928,
+      "grad_norm": 0.5502280592918396,
+      "learning_rate": 4.297487648074538e-06,
+      "loss": 0.2497,
+      "step": 921
+    },
+    {
+      "epoch": 1.7711999999999999,
+      "grad_norm": 0.5093812346458435,
+      "learning_rate": 4.286412389352929e-06,
+      "loss": 0.2172,
+      "step": 922
+    },
+    {
+      "epoch": 1.77312,
+      "grad_norm": 0.4885612428188324,
+      "learning_rate": 4.275340703463767e-06,
+      "loss": 0.2029,
+      "step": 923
+    },
+    {
+      "epoch": 1.77504,
+      "grad_norm": 0.5554471611976624,
+      "learning_rate": 4.264272645841419e-06,
+      "loss": 0.2346,
+      "step": 924
+    },
+    {
+      "epoch": 1.7769599999999999,
+      "grad_norm": 0.5582790970802307,
+      "learning_rate": 4.253208271902091e-06,
+      "loss": 0.2448,
+      "step": 925
+    },
+    {
+      "epoch": 1.77888,
+      "grad_norm": 0.4595748484134674,
+      "learning_rate": 4.242147637043539e-06,
+      "loss": 0.1794,
+      "step": 926
+    },
+    {
+      "epoch": 1.7808000000000002,
+      "grad_norm": 0.5237067937850952,
+      "learning_rate": 4.231090796644798e-06,
+      "loss": 0.2089,
+      "step": 927
+    },
+    {
+      "epoch": 1.7827199999999999,
+      "grad_norm": 0.4331246018409729,
+      "learning_rate": 4.220037806065911e-06,
+      "loss": 0.1858,
+      "step": 928
+    },
+    {
+      "epoch": 1.78464,
+      "grad_norm": 0.46122682094573975,
+      "learning_rate": 4.208988720647642e-06,
+      "loss": 0.2128,
+      "step": 929
+    },
+    {
+      "epoch": 1.7865600000000001,
+      "grad_norm": 0.4906701147556305,
+      "learning_rate": 4.1979435957111984e-06,
+      "loss": 0.1878,
+      "step": 930
+    },
+    {
+      "epoch": 1.7884799999999998,
+      "grad_norm": 0.49091121554374695,
+      "learning_rate": 4.1869024865579664e-06,
+      "loss": 0.1908,
+      "step": 931
+    },
+    {
+      "epoch": 1.7904,
+      "grad_norm": 0.5190486907958984,
+      "learning_rate": 4.175865448469219e-06,
+      "loss": 0.2295,
+      "step": 932
+    },
+    {
+      "epoch": 1.7923200000000001,
+      "grad_norm": 0.5081413984298706,
+      "learning_rate": 4.164832536705845e-06,
+      "loss": 0.2248,
+      "step": 933
+    },
+    {
+      "epoch": 1.7942399999999998,
+      "grad_norm": 0.44351980090141296,
+      "learning_rate": 4.153803806508083e-06,
+      "loss": 0.205,
+      "step": 934
+    },
+    {
+      "epoch": 1.79616,
+      "grad_norm": 0.4614003896713257,
+      "learning_rate": 4.142779313095223e-06,
+      "loss": 0.1716,
+      "step": 935
+    },
+    {
+      "epoch": 1.7980800000000001,
+      "grad_norm": 0.4662795662879944,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.1913,
+      "step": 936
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.5113293528556824,
+      "learning_rate": 4.120743257395054e-06,
+      "loss": 0.2422,
+      "step": 937
+    },
+    {
+      "epoch": 1.80192,
+      "grad_norm": 0.5030005574226379,
+      "learning_rate": 4.109731805439168e-06,
+      "loss": 0.2514,
+      "step": 938
+    },
+    {
+      "epoch": 1.80384,
+      "grad_norm": 0.4767839312553406,
+      "learning_rate": 4.098724810930472e-06,
+      "loss": 0.1887,
+      "step": 939
+    },
+    {
+      "epoch": 1.80576,
+      "grad_norm": 0.5297487378120422,
+      "learning_rate": 4.087722328979437e-06,
+      "loss": 0.246,
+      "step": 940
+    },
+    {
+      "epoch": 1.80768,
+      "grad_norm": 0.5102206468582153,
+      "learning_rate": 4.076724414673937e-06,
+      "loss": 0.1894,
+      "step": 941
+    },
+    {
+      "epoch": 1.8096,
+      "grad_norm": 0.4563246965408325,
+      "learning_rate": 4.065731123078977e-06,
+      "loss": 0.2163,
+      "step": 942
+    },
+    {
+      "epoch": 1.81152,
+      "grad_norm": 0.4089660346508026,
+      "learning_rate": 4.054742509236416e-06,
+      "loss": 0.157,
+      "step": 943
+    },
+    {
+      "epoch": 1.81344,
+      "grad_norm": 0.532998263835907,
+      "learning_rate": 4.043758628164688e-06,
+      "loss": 0.229,
+      "step": 944
+    },
+    {
+      "epoch": 1.81536,
+      "grad_norm": 0.5334629416465759,
+      "learning_rate": 4.032779534858544e-06,
+      "loss": 0.2254,
+      "step": 945
+    },
+    {
+      "epoch": 1.81728,
+      "grad_norm": 0.45861199498176575,
+      "learning_rate": 4.021805284288749e-06,
+      "loss": 0.1918,
+      "step": 946
+    },
+    {
+      "epoch": 1.8192,
+      "grad_norm": 0.5196312665939331,
+      "learning_rate": 4.0108359314018254e-06,
+      "loss": 0.2138,
+      "step": 947
+    },
+    {
+      "epoch": 1.82112,
+      "grad_norm": 0.563522219657898,
+      "learning_rate": 3.999871531119779e-06,
+      "loss": 0.2546,
+      "step": 948
+    },
+    {
+      "epoch": 1.82304,
+      "grad_norm": 0.4445743262767792,
+      "learning_rate": 3.988912138339812e-06,
+      "loss": 0.194,
+      "step": 949
+    },
+    {
+      "epoch": 1.82496,
+      "grad_norm": 0.4495560824871063,
+      "learning_rate": 3.977957807934055e-06,
+      "loss": 0.2005,
+      "step": 950
+    },
+    {
+      "epoch": 1.82688,
+      "grad_norm": 0.4912163019180298,
+      "learning_rate": 3.9670085947493e-06,
+      "loss": 0.2105,
+      "step": 951
+    },
+    {
+      "epoch": 1.8288,
+      "grad_norm": 0.43792349100112915,
+      "learning_rate": 3.956064553606708e-06,
+      "loss": 0.178,
+      "step": 952
+    },
+    {
+      "epoch": 1.83072,
+      "grad_norm": 0.47214657068252563,
+      "learning_rate": 3.945125739301547e-06,
+      "loss": 0.1966,
+      "step": 953
+    },
+    {
+      "epoch": 1.83264,
+      "grad_norm": 0.5191349387168884,
+      "learning_rate": 3.934192206602921e-06,
+      "loss": 0.2234,
+      "step": 954
+    },
+    {
+      "epoch": 1.83456,
+      "grad_norm": 0.4437025189399719,
+      "learning_rate": 3.923264010253479e-06,
+      "loss": 0.1772,
+      "step": 955
+    },
+    {
+      "epoch": 1.83648,
+      "grad_norm": 0.46581515669822693,
+      "learning_rate": 3.912341204969164e-06,
+      "loss": 0.2251,
+      "step": 956
+    },
+    {
+      "epoch": 1.8384,
+      "grad_norm": 0.5558372735977173,
+      "learning_rate": 3.901423845438916e-06,
+      "loss": 0.2573,
+      "step": 957
+    },
+    {
+      "epoch": 1.84032,
+      "grad_norm": 0.5010911226272583,
+      "learning_rate": 3.890511986324413e-06,
+      "loss": 0.2038,
+      "step": 958
+    },
+    {
+      "epoch": 1.8422399999999999,
+      "grad_norm": 0.5610588788986206,
+      "learning_rate": 3.879605682259797e-06,
+      "loss": 0.2335,
+      "step": 959
+    },
+    {
+      "epoch": 1.84416,
+      "grad_norm": 0.5048314332962036,
+      "learning_rate": 3.86870498785139e-06,
+      "loss": 0.1936,
+      "step": 960
+    },
+    {
+      "epoch": 1.8460800000000002,
+      "grad_norm": 0.4711723327636719,
+      "learning_rate": 3.857809957677432e-06,
+      "loss": 0.1793,
+      "step": 961
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 0.5136181712150574,
+      "learning_rate": 3.8469206462878e-06,
+      "loss": 0.2448,
+      "step": 962
+    },
+    {
+      "epoch": 1.84992,
+      "grad_norm": 0.4977407157421112,
+      "learning_rate": 3.83603710820374e-06,
+      "loss": 0.1704,
+      "step": 963
+    },
+    {
+      "epoch": 1.8518400000000002,
+      "grad_norm": 0.5353711247444153,
+      "learning_rate": 3.825159397917589e-06,
+      "loss": 0.2319,
+      "step": 964
+    },
+    {
+      "epoch": 1.8537599999999999,
+      "grad_norm": 0.44187629222869873,
+      "learning_rate": 3.814287569892512e-06,
+      "loss": 0.1683,
+      "step": 965
+    },
+    {
+      "epoch": 1.85568,
+      "grad_norm": 0.4551312029361725,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.1717,
+      "step": 966
+    },
+    {
+      "epoch": 1.8576000000000001,
+      "grad_norm": 0.506772518157959,
+      "learning_rate": 3.7925617783306757e-06,
+      "loss": 0.2295,
+      "step": 967
+    },
+    {
+      "epoch": 1.8595199999999998,
+      "grad_norm": 0.48513033986091614,
+      "learning_rate": 3.781707923571891e-06,
+      "loss": 0.2423,
+      "step": 968
+    },
+    {
+      "epoch": 1.86144,
+      "grad_norm": 0.4860017001628876,
+      "learning_rate": 3.7708601686295756e-06,
+      "loss": 0.2116,
+      "step": 969
+    },
+    {
+      "epoch": 1.8633600000000001,
+      "grad_norm": 0.46134498715400696,
+      "learning_rate": 3.7600185678169083e-06,
+      "loss": 0.1937,
+      "step": 970
+    },
+    {
+      "epoch": 1.86528,
+      "grad_norm": 0.508679986000061,
+      "learning_rate": 3.7491831754162533e-06,
+      "loss": 0.1999,
+      "step": 971
+    },
+    {
+      "epoch": 1.8672,
+      "grad_norm": 0.5298797488212585,
+      "learning_rate": 3.7383540456788915e-06,
+      "loss": 0.2373,
+      "step": 972
+    },
+    {
+      "epoch": 1.8691200000000001,
+      "grad_norm": 0.48576176166534424,
+      "learning_rate": 3.727531232824747e-06,
+      "loss": 0.2136,
+      "step": 973
+    },
+    {
+      "epoch": 1.87104,
+      "grad_norm": 0.45504137873649597,
+      "learning_rate": 3.7167147910421165e-06,
+      "loss": 0.1901,
+      "step": 974
+    },
+    {
+      "epoch": 1.87296,
+      "grad_norm": 0.4950440526008606,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.215,
+      "step": 975
+    },
+    {
+      "epoch": 1.87488,
+      "grad_norm": 0.4914148151874542,
+      "learning_rate": 3.695101237284815e-06,
+      "loss": 0.2164,
+      "step": 976
+    },
+    {
+      "epoch": 1.8768,
+      "grad_norm": 0.4940747618675232,
+      "learning_rate": 3.6843042335261583e-06,
+      "loss": 0.1939,
+      "step": 977
+    },
+    {
+      "epoch": 1.87872,
+      "grad_norm": 0.5065493583679199,
+      "learning_rate": 3.6735138172704967e-06,
+      "loss": 0.1925,
+      "step": 978
+    },
+    {
+      "epoch": 1.88064,
+      "grad_norm": 0.5148585438728333,
+      "learning_rate": 3.662730042543926e-06,
+      "loss": 0.2042,
+      "step": 979
+    },
+    {
+      "epoch": 1.88256,
+      "grad_norm": 0.4783082604408264,
+      "learning_rate": 3.6519529633392825e-06,
+      "loss": 0.2119,
+      "step": 980
+    },
+    {
+      "epoch": 1.88448,
+      "grad_norm": 0.4991118311882019,
+      "learning_rate": 3.6411826336158785e-06,
+      "loss": 0.2332,
+      "step": 981
+    },
+    {
+      "epoch": 1.8864,
+      "grad_norm": 0.5151258707046509,
+      "learning_rate": 3.6304191072992376e-06,
+      "loss": 0.2429,
+      "step": 982
+    },
+    {
+      "epoch": 1.88832,
+      "grad_norm": 0.51267409324646,
+      "learning_rate": 3.619662438280816e-06,
+      "loss": 0.2165,
+      "step": 983
+    },
+    {
+      "epoch": 1.89024,
+      "grad_norm": 0.4503840506076813,
+      "learning_rate": 3.6089126804177373e-06,
+      "loss": 0.1769,
+      "step": 984
+    },
+    {
+      "epoch": 1.89216,
+      "grad_norm": 0.524727463722229,
+      "learning_rate": 3.5981698875325214e-06,
+      "loss": 0.2251,
+      "step": 985
+    },
+    {
+      "epoch": 1.89408,
+      "grad_norm": 0.543687641620636,
+      "learning_rate": 3.5874341134128156e-06,
+      "loss": 0.2305,
+      "step": 986
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 0.5415152311325073,
+      "learning_rate": 3.5767054118111266e-06,
+      "loss": 0.2416,
+      "step": 987
+    },
+    {
+      "epoch": 1.89792,
+      "grad_norm": 0.5072895884513855,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 0.2292,
+      "step": 988
+    },
+    {
+      "epoch": 1.89984,
+      "grad_norm": 0.49900156259536743,
+      "learning_rate": 3.555269440994496e-06,
+      "loss": 0.2352,
+      "step": 989
+    },
+    {
+      "epoch": 1.90176,
+      "grad_norm": 0.4872468411922455,
+      "learning_rate": 3.544562279106436e-06,
+      "loss": 0.2152,
+      "step": 990
+    },
+    {
+      "epoch": 1.90368,
+      "grad_norm": 0.5773963332176208,
+      "learning_rate": 3.5338624043896154e-06,
+      "loss": 0.2584,
+      "step": 991
+    },
+    {
+      "epoch": 1.9056,
+      "grad_norm": 0.4402815103530884,
+      "learning_rate": 3.523169870416795e-06,
+      "loss": 0.1705,
+      "step": 992
+    },
+    {
+      "epoch": 1.9075199999999999,
+      "grad_norm": 0.4753376245498657,
+      "learning_rate": 3.5124847307239863e-06,
+      "loss": 0.1909,
+      "step": 993
+    },
+    {
+      "epoch": 1.90944,
+      "grad_norm": 0.4779719114303589,
+      "learning_rate": 3.501807038810174e-06,
+      "loss": 0.2244,
+      "step": 994
+    },
+    {
+      "epoch": 1.91136,
+      "grad_norm": 0.48947399854660034,
+      "learning_rate": 3.4911368481370535e-06,
+      "loss": 0.1946,
+      "step": 995
+    },
+    {
+      "epoch": 1.9132799999999999,
+      "grad_norm": 0.4469459652900696,
+      "learning_rate": 3.480474212128766e-06,
+      "loss": 0.1829,
+      "step": 996
+    },
+    {
+      "epoch": 1.9152,
+      "grad_norm": 0.4382728934288025,
+      "learning_rate": 3.469819184171623e-06,
+      "loss": 0.22,
+      "step": 997
+    },
+    {
+      "epoch": 1.9171200000000002,
+      "grad_norm": 0.5322670340538025,
+      "learning_rate": 3.459171817613847e-06,
+      "loss": 0.2343,
+      "step": 998
+    },
+    {
+      "epoch": 1.9190399999999999,
+      "grad_norm": 0.4600752890110016,
+      "learning_rate": 3.4485321657653e-06,
+      "loss": 0.1935,
+      "step": 999
+    },
+    {
+      "epoch": 1.92096,
+      "grad_norm": 0.4460200071334839,
+      "learning_rate": 3.4379002818972122e-06,
+      "loss": 0.1647,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9228800000000001,
+      "grad_norm": 0.42405110597610474,
+      "learning_rate": 3.427276219241933e-06,
+      "loss": 0.2055,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9247999999999998,
+      "grad_norm": 0.4926280677318573,
+      "learning_rate": 3.416660030992639e-06,
+      "loss": 0.2078,
+      "step": 1002
+    },
+    {
+      "epoch": 1.92672,
+      "grad_norm": 0.46914806962013245,
+      "learning_rate": 3.406051770303087e-06,
+      "loss": 0.205,
+      "step": 1003
+    },
+    {
+      "epoch": 1.9286400000000001,
+      "grad_norm": 0.5986695289611816,
+      "learning_rate": 3.3954514902873427e-06,
+      "loss": 0.2361,
+      "step": 1004
+    },
+    {
+      "epoch": 1.93056,
+      "grad_norm": 0.520028829574585,
+      "learning_rate": 3.3848592440195118e-06,
+      "loss": 0.1942,
+      "step": 1005
+    },
+    {
+      "epoch": 1.93248,
+      "grad_norm": 0.4387219548225403,
+      "learning_rate": 3.3742750845334748e-06,
+      "loss": 0.2002,
+      "step": 1006
+    },
+    {
+      "epoch": 1.9344000000000001,
+      "grad_norm": 0.4803704619407654,
+      "learning_rate": 3.3636990648226258e-06,
+      "loss": 0.2124,
+      "step": 1007
+    },
+    {
+      "epoch": 1.93632,
+      "grad_norm": 0.45761480927467346,
+      "learning_rate": 3.3531312378396026e-06,
+      "loss": 0.1913,
+      "step": 1008
+    },
+    {
+      "epoch": 1.93824,
+      "grad_norm": 0.4841192662715912,
+      "learning_rate": 3.342571656496022e-06,
+      "loss": 0.1996,
+      "step": 1009
+    },
+    {
+      "epoch": 1.94016,
+      "grad_norm": 0.5036253333091736,
+      "learning_rate": 3.3320203736622185e-06,
+      "loss": 0.213,
+      "step": 1010
+    },
+    {
+      "epoch": 1.94208,
+      "grad_norm": 0.5175012350082397,
+      "learning_rate": 3.3214774421669777e-06,
+      "loss": 0.2344,
+      "step": 1011
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 0.4751274585723877,
+      "learning_rate": 3.310942914797265e-06,
+      "loss": 0.1938,
+      "step": 1012
+    },
+    {
+      "epoch": 1.94592,
+      "grad_norm": 0.449361115694046,
+      "learning_rate": 3.3004168442979755e-06,
+      "loss": 0.1737,
+      "step": 1013
+    },
+    {
+      "epoch": 1.94784,
+      "grad_norm": 0.5092611908912659,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 0.207,
+      "step": 1014
+    },
+    {
+      "epoch": 1.94976,
+      "grad_norm": 0.5049817562103271,
+      "learning_rate": 3.2793902846782534e-06,
+      "loss": 0.1885,
+      "step": 1015
+    },
+    {
+      "epoch": 1.95168,
+      "grad_norm": 0.5169958472251892,
+      "learning_rate": 3.2688899008348386e-06,
+      "loss": 0.2021,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9536,
+      "grad_norm": 0.45318013429641724,
+      "learning_rate": 3.2583981844153487e-06,
+      "loss": 0.2012,
+      "step": 1017
+    },
+    {
+      "epoch": 1.95552,
+      "grad_norm": 0.4782044589519501,
+      "learning_rate": 3.2479151879503324e-06,
+      "loss": 0.1851,
+      "step": 1018
+    },
+    {
+      "epoch": 1.95744,
+      "grad_norm": 0.4753035306930542,
+      "learning_rate": 3.2374409639266695e-06,
+      "loss": 0.1925,
+      "step": 1019
+    },
+    {
+      "epoch": 1.95936,
+      "grad_norm": 0.5349734425544739,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.2439,
+      "step": 1020
+    },
+    {
+      "epoch": 1.96128,
+      "grad_norm": 0.48619574308395386,
+      "learning_rate": 3.2165190429310674e-06,
+      "loss": 0.2516,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9632,
+      "grad_norm": 0.4598257839679718,
+      "learning_rate": 3.206071450712235e-06,
+      "loss": 0.2158,
+      "step": 1022
+    },
+    {
+      "epoch": 1.96512,
+      "grad_norm": 0.4528428018093109,
+      "learning_rate": 3.1956328404404403e-06,
+      "loss": 0.1784,
+      "step": 1023
+    },
+    {
+      "epoch": 1.96704,
+      "grad_norm": 0.5321189761161804,
+      "learning_rate": 3.1852032643803377e-06,
+      "loss": 0.2577,
+      "step": 1024
+    },
+    {
+      "epoch": 1.96896,
+      "grad_norm": 0.5134246945381165,
+      "learning_rate": 3.174782774751338e-06,
+      "loss": 0.1828,
+      "step": 1025
+    },
+    {
+      "epoch": 1.97088,
+      "grad_norm": 0.44750314950942993,
+      "learning_rate": 3.1643714237273628e-06,
+      "loss": 0.2025,
+      "step": 1026
+    },
+    {
+      "epoch": 1.9727999999999999,
+      "grad_norm": 0.48976466059684753,
+      "learning_rate": 3.1539692634365788e-06,
+      "loss": 0.1957,
+      "step": 1027
+    },
+    {
+      "epoch": 1.97472,
+      "grad_norm": 0.5019981861114502,
+      "learning_rate": 3.143576345961132e-06,
+      "loss": 0.2334,
+      "step": 1028
+    },
+    {
+      "epoch": 1.97664,
+      "grad_norm": 0.5189653038978577,
+      "learning_rate": 3.1331927233368954e-06,
+      "loss": 0.2121,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9785599999999999,
+      "grad_norm": 0.5176576375961304,
+      "learning_rate": 3.1228184475532015e-06,
+      "loss": 0.2238,
+      "step": 1030
+    },
+    {
+      "epoch": 1.98048,
+      "grad_norm": 0.45314568281173706,
+      "learning_rate": 3.112453570552583e-06,
+      "loss": 0.1759,
+      "step": 1031
+    },
+    {
+      "epoch": 1.9824000000000002,
+      "grad_norm": 0.44203290343284607,
+      "learning_rate": 3.1020981442305187e-06,
+      "loss": 0.1852,
+      "step": 1032
+    },
+    {
+      "epoch": 1.9843199999999999,
+      "grad_norm": 0.4552965462207794,
+      "learning_rate": 3.091752220435166e-06,
+      "loss": 0.2217,
+      "step": 1033
+    },
+    {
+      "epoch": 1.98624,
+      "grad_norm": 0.5059014558792114,
+      "learning_rate": 3.0814158509671015e-06,
+      "loss": 0.2376,
+      "step": 1034
+    },
+    {
+      "epoch": 1.9881600000000001,
+      "grad_norm": 0.48215579986572266,
+      "learning_rate": 3.0710890875790745e-06,
+      "loss": 0.1919,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9900799999999998,
+      "grad_norm": 0.5080945491790771,
+      "learning_rate": 3.0607719819757264e-06,
+      "loss": 0.1873,
+      "step": 1036
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 0.5062063336372375,
+      "learning_rate": 3.0504645858133507e-06,
+      "loss": 0.2096,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9939200000000001,
+      "grad_norm": 0.5286766290664673,
+      "learning_rate": 3.040166950699626e-06,
+      "loss": 0.2339,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9958399999999998,
+      "grad_norm": 0.5243569016456604,
+      "learning_rate": 3.029879128193356e-06,
+      "loss": 0.2223,
+      "step": 1039
+    },
+    {
+      "epoch": 1.99776,
+      "grad_norm": 0.5236871838569641,
+      "learning_rate": 3.019601169804216e-06,
+      "loss": 0.255,
+      "step": 1040
+    },
+    {
+      "epoch": 2.00064,
+      "grad_norm": 0.970310389995575,
+      "learning_rate": 3.0093331269924954e-06,
+      "loss": 0.366,
+      "step": 1041
+    },
+    {
+      "epoch": 2.00256,
+      "grad_norm": 0.4161249101161957,
+      "learning_rate": 2.9990750511688323e-06,
+      "loss": 0.1571,
+      "step": 1042
+    },
+    {
+      "epoch": 2.00448,
+      "grad_norm": 0.5139119029045105,
+      "learning_rate": 2.988826993693967e-06,
+      "loss": 0.2178,
+      "step": 1043
+    },
+    {
+      "epoch": 2.0064,
+      "grad_norm": 0.4825448989868164,
+      "learning_rate": 2.978589005878476e-06,
+      "loss": 0.1824,
+      "step": 1044
+    },
+    {
+      "epoch": 2.00832,
+      "grad_norm": 0.501803457736969,
+      "learning_rate": 2.968361138982517e-06,
+      "loss": 0.1802,
+      "step": 1045
+    },
+    {
+      "epoch": 2.01024,
+      "grad_norm": 0.43730831146240234,
+      "learning_rate": 2.9581434442155798e-06,
+      "loss": 0.1875,
+      "step": 1046
+    },
+    {
+      "epoch": 2.01216,
+      "grad_norm": 0.45184391736984253,
+      "learning_rate": 2.947935972736217e-06,
+      "loss": 0.1866,
+      "step": 1047
+    },
+    {
+      "epoch": 2.01408,
+      "grad_norm": 0.47673094272613525,
+      "learning_rate": 2.937738775651798e-06,
+      "loss": 0.19,
+      "step": 1048
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 0.4758540987968445,
+      "learning_rate": 2.9275519040182503e-06,
+      "loss": 0.1743,
+      "step": 1049
+    },
+    {
+      "epoch": 2.01792,
+      "grad_norm": 0.4605776369571686,
+      "learning_rate": 2.917375408839803e-06,
+      "loss": 0.2121,
+      "step": 1050
+    },
+    {
+      "epoch": 2.01984,
+      "grad_norm": 0.4795879125595093,
+      "learning_rate": 2.9072093410687268e-06,
+      "loss": 0.1792,
+      "step": 1051
+    },
+    {
+      "epoch": 2.02176,
+      "grad_norm": 0.5252066254615784,
+      "learning_rate": 2.8970537516050935e-06,
+      "loss": 0.2256,
+      "step": 1052
+    },
+    {
+      "epoch": 2.02368,
+      "grad_norm": 0.4893878400325775,
+      "learning_rate": 2.886908691296504e-06,
+      "loss": 0.1834,
+      "step": 1053
+    },
+    {
+      "epoch": 2.0256,
+      "grad_norm": 0.45864781737327576,
+      "learning_rate": 2.876774210937843e-06,
+      "loss": 0.1659,
+      "step": 1054
+    },
+    {
+      "epoch": 2.02752,
+      "grad_norm": 0.5190344452857971,
+      "learning_rate": 2.866650361271023e-06,
+      "loss": 0.2,
+      "step": 1055
+    },
+    {
+      "epoch": 2.02944,
+      "grad_norm": 0.4104481637477875,
+      "learning_rate": 2.8565371929847286e-06,
+      "loss": 0.1395,
+      "step": 1056
+    },
+    {
+      "epoch": 2.03136,
+      "grad_norm": 0.43606022000312805,
+      "learning_rate": 2.84643475671417e-06,
+      "loss": 0.152,
+      "step": 1057
+    },
+    {
+      "epoch": 2.03328,
+      "grad_norm": 0.41685885190963745,
+      "learning_rate": 2.836343103040819e-06,
+      "loss": 0.1685,
+      "step": 1058
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 0.44999656081199646,
+      "learning_rate": 2.8262622824921593e-06,
+      "loss": 0.1692,
+      "step": 1059
+    },
+    {
+      "epoch": 2.03712,
+      "grad_norm": 0.4724535048007965,
+      "learning_rate": 2.816192345541437e-06,
+      "loss": 0.1657,
+      "step": 1060
+    },
+    {
+      "epoch": 2.03904,
+      "grad_norm": 0.4805181622505188,
+      "learning_rate": 2.8061333426074054e-06,
+      "loss": 0.2042,
+      "step": 1061
+    },
+    {
+      "epoch": 2.04096,
+      "grad_norm": 0.4885428249835968,
+      "learning_rate": 2.7960853240540703e-06,
+      "loss": 0.1747,
+      "step": 1062
+    },
+    {
+      "epoch": 2.04288,
+      "grad_norm": 0.5053113698959351,
+      "learning_rate": 2.7860483401904483e-06,
+      "loss": 0.175,
+      "step": 1063
+    },
+    {
+      "epoch": 2.0448,
+      "grad_norm": 0.5168372392654419,
+      "learning_rate": 2.776022441270295e-06,
+      "loss": 0.1725,
+      "step": 1064
+    },
+    {
+      "epoch": 2.04672,
+      "grad_norm": 0.5295779705047607,
+      "learning_rate": 2.766007677491871e-06,
+      "loss": 0.2107,
+      "step": 1065
+    },
+    {
+      "epoch": 2.04864,
+      "grad_norm": 0.4353882968425751,
+      "learning_rate": 2.7560040989976894e-06,
+      "loss": 0.1681,
+      "step": 1066
+    },
+    {
+      "epoch": 2.05056,
+      "grad_norm": 0.4644116759300232,
+      "learning_rate": 2.7460117558742532e-06,
+      "loss": 0.1706,
+      "step": 1067
+    },
+    {
+      "epoch": 2.05248,
+      "grad_norm": 0.4497801959514618,
+      "learning_rate": 2.736030698151815e-06,
+      "loss": 0.1579,
+      "step": 1068
+    },
+    {
+      "epoch": 2.0544,
+      "grad_norm": 0.5074851512908936,
+      "learning_rate": 2.72606097580412e-06,
+      "loss": 0.1745,
+      "step": 1069
+    },
+    {
+      "epoch": 2.05632,
+      "grad_norm": 0.4704873859882355,
+      "learning_rate": 2.7161026387481636e-06,
+      "loss": 0.1853,
+      "step": 1070
+    },
+    {
+      "epoch": 2.05824,
+      "grad_norm": 0.4752654731273651,
+      "learning_rate": 2.7061557368439294e-06,
+      "loss": 0.1697,
+      "step": 1071
+    },
+    {
+      "epoch": 2.06016,
+      "grad_norm": 0.5229239463806152,
+      "learning_rate": 2.6962203198941587e-06,
+      "loss": 0.2094,
+      "step": 1072
+    },
+    {
+      "epoch": 2.06208,
+      "grad_norm": 0.46261343359947205,
+      "learning_rate": 2.686296437644074e-06,
+      "loss": 0.1661,
+      "step": 1073
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 0.5085206627845764,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.1804,
+      "step": 1074
+    },
+    {
+      "epoch": 2.06592,
+      "grad_norm": 0.4511212706565857,
+      "learning_rate": 2.666483475934885e-06,
+      "loss": 0.153,
+      "step": 1075
+    },
+    {
+      "epoch": 2.06784,
+      "grad_norm": 0.47942277789115906,
+      "learning_rate": 2.656594495676482e-06,
+      "loss": 0.1871,
+      "step": 1076
+    },
+    {
+      "epoch": 2.06976,
+      "grad_norm": 0.448357492685318,
+      "learning_rate": 2.6467172485186775e-06,
+      "loss": 0.1703,
+      "step": 1077
+    },
+    {
+      "epoch": 2.07168,
+      "grad_norm": 0.43812739849090576,
+      "learning_rate": 2.636851783915454e-06,
+      "loss": 0.1498,
+      "step": 1078
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.5352055430412292,
+      "learning_rate": 2.626998151261798e-06,
+      "loss": 0.2051,
+      "step": 1079
+    },
+    {
+      "epoch": 2.07552,
+      "grad_norm": 0.4882618486881256,
+      "learning_rate": 2.6171563998934605e-06,
+      "loss": 0.1706,
+      "step": 1080
+    },
+    {
+      "epoch": 2.07744,
+      "grad_norm": 0.4742896258831024,
+      "learning_rate": 2.607326579086701e-06,
+      "loss": 0.1781,
+      "step": 1081
+    },
+    {
+      "epoch": 2.07936,
+      "grad_norm": 0.4287046194076538,
+      "learning_rate": 2.5975087380580445e-06,
+      "loss": 0.1585,
+      "step": 1082
+    },
+    {
+      "epoch": 2.08128,
+      "grad_norm": 0.47216537594795227,
+      "learning_rate": 2.587702925964034e-06,
+      "loss": 0.159,
+      "step": 1083
+    },
+    {
+      "epoch": 2.0832,
+      "grad_norm": 0.5327614545822144,
+      "learning_rate": 2.577909191900988e-06,
+      "loss": 0.1917,
+      "step": 1084
+    },
+    {
+      "epoch": 2.08512,
+      "grad_norm": 0.4365946054458618,
+      "learning_rate": 2.5681275849047482e-06,
+      "loss": 0.1537,
+      "step": 1085
+    },
+    {
+      "epoch": 2.08704,
+      "grad_norm": 0.4499369263648987,
+      "learning_rate": 2.5583581539504464e-06,
+      "loss": 0.166,
+      "step": 1086
+    },
+    {
+      "epoch": 2.08896,
+      "grad_norm": 0.5650961399078369,
+      "learning_rate": 2.5486009479522355e-06,
+      "loss": 0.2267,
+      "step": 1087
+    },
+    {
+      "epoch": 2.09088,
+      "grad_norm": 0.5321717262268066,
+      "learning_rate": 2.5388560157630765e-06,
+      "loss": 0.194,
+      "step": 1088
+    },
+    {
+      "epoch": 2.0928,
+      "grad_norm": 0.4969466030597687,
+      "learning_rate": 2.5291234061744655e-06,
+      "loss": 0.2013,
+      "step": 1089
+    },
+    {
+      "epoch": 2.09472,
+      "grad_norm": 0.4784854054450989,
+      "learning_rate": 2.519403167916207e-06,
+      "loss": 0.1547,
+      "step": 1090
+    },
+    {
+      "epoch": 2.09664,
+      "grad_norm": 0.49899154901504517,
+      "learning_rate": 2.50969534965616e-06,
+      "loss": 0.1897,
+      "step": 1091
+    },
+    {
+      "epoch": 2.09856,
+      "grad_norm": 0.4661808907985687,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.1562,
+      "step": 1092
+    },
+    {
+      "epoch": 2.10048,
+      "grad_norm": 0.5066959857940674,
+      "learning_rate": 2.490317167490976e-06,
+      "loss": 0.1862,
+      "step": 1093
+    },
+    {
+      "epoch": 2.1024,
+      "grad_norm": 0.4774594306945801,
+      "learning_rate": 2.480646900609664e-06,
+      "loss": 0.1681,
+      "step": 1094
+    },
+    {
+      "epoch": 2.10432,
+      "grad_norm": 0.5425265431404114,
+      "learning_rate": 2.4709892477737263e-06,
+      "loss": 0.2487,
+      "step": 1095
+    },
+    {
+      "epoch": 2.10624,
+      "grad_norm": 0.5033844709396362,
+      "learning_rate": 2.4613442573376625e-06,
+      "loss": 0.2103,
+      "step": 1096
+    },
+    {
+      "epoch": 2.10816,
+      "grad_norm": 0.4853045344352722,
+      "learning_rate": 2.4517119775925824e-06,
+      "loss": 0.1881,
+      "step": 1097
+    },
+    {
+      "epoch": 2.11008,
+      "grad_norm": 0.545462429523468,
+      "learning_rate": 2.4420924567659508e-06,
+      "loss": 0.2005,
+      "step": 1098
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.4619285762310028,
+      "learning_rate": 2.4324857430213504e-06,
+      "loss": 0.1708,
+      "step": 1099
+    },
+    {
+      "epoch": 2.11392,
+      "grad_norm": 0.4951660931110382,
+      "learning_rate": 2.422891884458241e-06,
+      "loss": 0.1723,
+      "step": 1100
+    },
+    {
+      "epoch": 2.11584,
+      "grad_norm": 0.509020209312439,
+      "learning_rate": 2.4133109291117156e-06,
+      "loss": 0.175,
+      "step": 1101
+    },
+    {
+      "epoch": 2.11776,
+      "grad_norm": 0.47044476866722107,
+      "learning_rate": 2.4037429249522702e-06,
+      "loss": 0.1482,
+      "step": 1102
+    },
+    {
+      "epoch": 2.11968,
+      "grad_norm": 0.545199990272522,
+      "learning_rate": 2.394187919885548e-06,
+      "loss": 0.1948,
+      "step": 1103
+    },
+    {
+      "epoch": 2.1216,
+      "grad_norm": 0.49692851305007935,
+      "learning_rate": 2.384645961752113e-06,
+      "loss": 0.1912,
+      "step": 1104
+    },
+    {
+      "epoch": 2.12352,
+      "grad_norm": 0.5112411379814148,
+      "learning_rate": 2.3751170983272e-06,
+      "loss": 0.1742,
+      "step": 1105
+    },
+    {
+      "epoch": 2.12544,
+      "grad_norm": 0.559057891368866,
+      "learning_rate": 2.3656013773204843e-06,
+      "loss": 0.2181,
+      "step": 1106
+    },
+    {
+      "epoch": 2.12736,
+      "grad_norm": 0.4811709523200989,
+      "learning_rate": 2.3560988463758366e-06,
+      "loss": 0.1871,
+      "step": 1107
+    },
+    {
+      "epoch": 2.12928,
+      "grad_norm": 0.47404956817626953,
+      "learning_rate": 2.346609553071093e-06,
+      "loss": 0.1632,
+      "step": 1108
+    },
+    {
+      "epoch": 2.1312,
+      "grad_norm": 0.49403196573257446,
+      "learning_rate": 2.3371335449178006e-06,
+      "loss": 0.1822,
+      "step": 1109
+    },
+    {
+      "epoch": 2.13312,
+      "grad_norm": 0.404538631439209,
+      "learning_rate": 2.3276708693609947e-06,
+      "loss": 0.1319,
+      "step": 1110
+    },
+    {
+      "epoch": 2.13504,
+      "grad_norm": 0.4431174099445343,
+      "learning_rate": 2.3182215737789593e-06,
+      "loss": 0.1526,
+      "step": 1111
+    },
+    {
+      "epoch": 2.13696,
+      "grad_norm": 0.48102709650993347,
+      "learning_rate": 2.308785705482982e-06,
+      "loss": 0.186,
+      "step": 1112
+    },
+    {
+      "epoch": 2.13888,
+      "grad_norm": 0.4914579689502716,
+      "learning_rate": 2.2993633117171243e-06,
+      "loss": 0.1468,
+      "step": 1113
+    },
+    {
+      "epoch": 2.1408,
+      "grad_norm": 0.5130366683006287,
+      "learning_rate": 2.289954439657981e-06,
+      "loss": 0.1699,
+      "step": 1114
+    },
+    {
+      "epoch": 2.14272,
+      "grad_norm": 0.48713797330856323,
+      "learning_rate": 2.2805591364144446e-06,
+      "loss": 0.173,
+      "step": 1115
+    },
+    {
+      "epoch": 2.14464,
+      "grad_norm": 0.4773431122303009,
+      "learning_rate": 2.2711774490274767e-06,
+      "loss": 0.1683,
+      "step": 1116
+    },
+    {
+      "epoch": 2.14656,
+      "grad_norm": 0.5155712962150574,
+      "learning_rate": 2.2618094244698614e-06,
+      "loss": 0.1664,
+      "step": 1117
+    },
+    {
+      "epoch": 2.14848,
+      "grad_norm": 0.5153727531433105,
+      "learning_rate": 2.2524551096459703e-06,
+      "loss": 0.1922,
+      "step": 1118
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.5002241134643555,
+      "learning_rate": 2.243114551391542e-06,
+      "loss": 0.1836,
+      "step": 1119
+    },
+    {
+      "epoch": 2.15232,
+      "grad_norm": 0.47715774178504944,
+      "learning_rate": 2.2337877964734324e-06,
+      "loss": 0.1524,
+      "step": 1120
+    },
+    {
+      "epoch": 2.15424,
+      "grad_norm": 0.5658057332038879,
+      "learning_rate": 2.224474891589386e-06,
+      "loss": 0.1846,
+      "step": 1121
+    },
+    {
+      "epoch": 2.15616,
+      "grad_norm": 0.4742417335510254,
+      "learning_rate": 2.2151758833678044e-06,
+      "loss": 0.1572,
+      "step": 1122
+    },
+    {
+      "epoch": 2.15808,
+      "grad_norm": 0.4582442045211792,
+      "learning_rate": 2.205890818367508e-06,
+      "loss": 0.159,
+      "step": 1123
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.5160391330718994,
+      "learning_rate": 2.1966197430775056e-06,
+      "loss": 0.1818,
+      "step": 1124
+    },
+    {
+      "epoch": 2.16192,
+      "grad_norm": 0.4705256521701813,
+      "learning_rate": 2.187362703916766e-06,
+      "loss": 0.1581,
+      "step": 1125
+    },
+    {
+      "epoch": 2.16384,
+      "grad_norm": 0.4833719730377197,
+      "learning_rate": 2.178119747233976e-06,
+      "loss": 0.1487,
+      "step": 1126
+    },
+    {
+      "epoch": 2.16576,
+      "grad_norm": 0.545116126537323,
+      "learning_rate": 2.168890919307315e-06,
+      "loss": 0.2019,
+      "step": 1127
+    },
+    {
+      "epoch": 2.16768,
+      "grad_norm": 0.5459210872650146,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.2035,
+      "step": 1128
+    },
+    {
+      "epoch": 2.1696,
+      "grad_norm": 0.5004191994667053,
+      "learning_rate": 2.1504758344811615e-06,
+      "loss": 0.2016,
+      "step": 1129
+    },
+    {
+      "epoch": 2.17152,
+      "grad_norm": 0.5192164778709412,
+      "learning_rate": 2.141289669783401e-06,
+      "loss": 0.1917,
+      "step": 1130
+    },
+    {
+      "epoch": 2.17344,
+      "grad_norm": 0.4754578471183777,
+      "learning_rate": 2.132117818244771e-06,
+      "loss": 0.1589,
+      "step": 1131
+    },
+    {
+      "epoch": 2.17536,
+      "grad_norm": 0.5209933519363403,
+      "learning_rate": 2.122960325787432e-06,
+      "loss": 0.1919,
+      "step": 1132
+    },
+    {
+      "epoch": 2.17728,
+      "grad_norm": 0.48300960659980774,
+      "learning_rate": 2.1138172382616612e-06,
+      "loss": 0.1646,
+      "step": 1133
+    },
+    {
+      "epoch": 2.1792,
+      "grad_norm": 0.4942103922367096,
+      "learning_rate": 2.104688601445606e-06,
+      "loss": 0.1754,
+      "step": 1134
+    },
+    {
+      "epoch": 2.18112,
+      "grad_norm": 0.4631735682487488,
+      "learning_rate": 2.0955744610450616e-06,
+      "loss": 0.1617,
+      "step": 1135
+    },
+    {
+      "epoch": 2.18304,
+      "grad_norm": 0.5241621136665344,
+      "learning_rate": 2.086474862693244e-06,
+      "loss": 0.187,
+      "step": 1136
+    },
+    {
+      "epoch": 2.1849600000000002,
+      "grad_norm": 0.4830331802368164,
+      "learning_rate": 2.077389851950557e-06,
+      "loss": 0.172,
+      "step": 1137
+    },
+    {
+      "epoch": 2.18688,
+      "grad_norm": 0.4721030592918396,
+      "learning_rate": 2.068319474304365e-06,
+      "loss": 0.1654,
+      "step": 1138
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 0.4895091950893402,
+      "learning_rate": 2.059263775168773e-06,
+      "loss": 0.1702,
+      "step": 1139
+    },
+    {
+      "epoch": 2.19072,
+      "grad_norm": 0.5315499305725098,
+      "learning_rate": 2.050222799884387e-06,
+      "loss": 0.2147,
+      "step": 1140
+    },
+    {
+      "epoch": 2.19264,
+      "grad_norm": 0.4849056303501129,
+      "learning_rate": 2.0411965937180907e-06,
+      "loss": 0.1798,
+      "step": 1141
+    },
+    {
+      "epoch": 2.19456,
+      "grad_norm": 0.48404166102409363,
+      "learning_rate": 2.0321852018628278e-06,
+      "loss": 0.1699,
+      "step": 1142
+    },
+    {
+      "epoch": 2.19648,
+      "grad_norm": 0.49606049060821533,
+      "learning_rate": 2.0231886694373653e-06,
+      "loss": 0.1741,
+      "step": 1143
+    },
+    {
+      "epoch": 2.1984,
+      "grad_norm": 0.5180538892745972,
+      "learning_rate": 2.0142070414860704e-06,
+      "loss": 0.1655,
+      "step": 1144
+    },
+    {
+      "epoch": 2.20032,
+      "grad_norm": 0.46643316745758057,
+      "learning_rate": 2.005240362978686e-06,
+      "loss": 0.1669,
+      "step": 1145
+    },
+    {
+      "epoch": 2.20224,
+      "grad_norm": 0.49554944038391113,
+      "learning_rate": 1.996288678810105e-06,
+      "loss": 0.1701,
+      "step": 1146
+    },
+    {
+      "epoch": 2.20416,
+      "grad_norm": 0.5687827467918396,
+      "learning_rate": 1.98735203380015e-06,
+      "loss": 0.1817,
+      "step": 1147
+    },
+    {
+      "epoch": 2.20608,
+      "grad_norm": 0.5384619832038879,
+      "learning_rate": 1.9784304726933384e-06,
+      "loss": 0.1741,
+      "step": 1148
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 0.45342355966567993,
+      "learning_rate": 1.9695240401586687e-06,
+      "loss": 0.1563,
+      "step": 1149
+    },
+    {
+      "epoch": 2.20992,
+      "grad_norm": 0.5521218776702881,
+      "learning_rate": 1.9606327807893905e-06,
+      "loss": 0.2083,
+      "step": 1150
+    },
+    {
+      "epoch": 2.21184,
+      "grad_norm": 0.49463796615600586,
+      "learning_rate": 1.9517567391027853e-06,
+      "loss": 0.1601,
+      "step": 1151
+    },
+    {
+      "epoch": 2.21376,
+      "grad_norm": 0.49851158261299133,
+      "learning_rate": 1.942895959539939e-06,
+      "loss": 0.1968,
+      "step": 1152
+    },
+    {
+      "epoch": 2.21568,
+      "grad_norm": 0.4996357262134552,
+      "learning_rate": 1.934050486465529e-06,
+      "loss": 0.1881,
+      "step": 1153
+    },
+    {
+      "epoch": 2.2176,
+      "grad_norm": 0.4992733299732208,
+      "learning_rate": 1.9252203641675854e-06,
+      "loss": 0.2102,
+      "step": 1154
+    },
+    {
+      "epoch": 2.21952,
+      "grad_norm": 0.5199818015098572,
+      "learning_rate": 1.9164056368572847e-06,
+      "loss": 0.1762,
+      "step": 1155
+    },
+    {
+      "epoch": 2.22144,
+      "grad_norm": 0.5557647347450256,
+      "learning_rate": 1.9076063486687256e-06,
+      "loss": 0.1983,
+      "step": 1156
+    },
+    {
+      "epoch": 2.22336,
+      "grad_norm": 0.5060413479804993,
+      "learning_rate": 1.8988225436587005e-06,
+      "loss": 0.2091,
+      "step": 1157
+    },
+    {
+      "epoch": 2.22528,
+      "grad_norm": 0.5298248529434204,
+      "learning_rate": 1.8900542658064807e-06,
+      "loss": 0.2038,
+      "step": 1158
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.478118896484375,
+      "learning_rate": 1.8813015590135963e-06,
+      "loss": 0.1815,
+      "step": 1159
+    },
+    {
+      "epoch": 2.22912,
+      "grad_norm": 0.4457749128341675,
+      "learning_rate": 1.8725644671036125e-06,
+      "loss": 0.1467,
+      "step": 1160
+    },
+    {
+      "epoch": 2.23104,
+      "grad_norm": 0.46739745140075684,
+      "learning_rate": 1.8638430338219199e-06,
+      "loss": 0.1708,
+      "step": 1161
+    },
+    {
+      "epoch": 2.23296,
+      "grad_norm": 0.5334953665733337,
+      "learning_rate": 1.8551373028355013e-06,
+      "loss": 0.2108,
+      "step": 1162
+    },
+    {
+      "epoch": 2.23488,
+      "grad_norm": 0.44612693786621094,
+      "learning_rate": 1.846447317732723e-06,
+      "loss": 0.1477,
+      "step": 1163
+    },
+    {
+      "epoch": 2.2368,
+      "grad_norm": 0.5191937685012817,
+      "learning_rate": 1.8377731220231144e-06,
+      "loss": 0.1838,
+      "step": 1164
+    },
+    {
+      "epoch": 2.23872,
+      "grad_norm": 0.5334745049476624,
+      "learning_rate": 1.8291147591371482e-06,
+      "loss": 0.2023,
+      "step": 1165
+    },
+    {
+      "epoch": 2.24064,
+      "grad_norm": 0.47443562746047974,
+      "learning_rate": 1.8204722724260266e-06,
+      "loss": 0.1574,
+      "step": 1166
+    },
+    {
+      "epoch": 2.24256,
+      "grad_norm": 0.5324453115463257,
+      "learning_rate": 1.8118457051614591e-06,
+      "loss": 0.1889,
+      "step": 1167
+    },
+    {
+      "epoch": 2.24448,
+      "grad_norm": 0.5276749134063721,
+      "learning_rate": 1.803235100535452e-06,
+      "loss": 0.2189,
+      "step": 1168
+    },
+    {
+      "epoch": 2.2464,
+      "grad_norm": 0.5361519455909729,
+      "learning_rate": 1.7946405016600843e-06,
+      "loss": 0.1932,
+      "step": 1169
+    },
+    {
+      "epoch": 2.24832,
+      "grad_norm": 0.5363965630531311,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 0.1954,
+      "step": 1170
+    },
+    {
+      "epoch": 2.25024,
+      "grad_norm": 0.46244028210639954,
+      "learning_rate": 1.7774994932086976e-06,
+      "loss": 0.1427,
+      "step": 1171
+    },
+    {
+      "epoch": 2.25216,
+      "grad_norm": 0.5298753976821899,
+      "learning_rate": 1.7689531694552863e-06,
+      "loss": 0.1918,
+      "step": 1172
+    },
+    {
+      "epoch": 2.25408,
+      "grad_norm": 0.5276806950569153,
+      "learning_rate": 1.7604230230973068e-06,
+      "loss": 0.1919,
+      "step": 1173
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 0.49854713678359985,
+      "learning_rate": 1.7519090968439966e-06,
+      "loss": 0.1586,
+      "step": 1174
+    },
+    {
+      "epoch": 2.25792,
+      "grad_norm": 0.5286722183227539,
+      "learning_rate": 1.7434114333233853e-06,
+      "loss": 0.1854,
+      "step": 1175
+    },
+    {
+      "epoch": 2.25984,
+      "grad_norm": 0.5443100333213806,
+      "learning_rate": 1.7349300750820758e-06,
+      "loss": 0.2069,
+      "step": 1176
+    },
+    {
+      "epoch": 2.2617599999999998,
+      "grad_norm": 0.47174641489982605,
+      "learning_rate": 1.7264650645850256e-06,
+      "loss": 0.1975,
+      "step": 1177
+    },
+    {
+      "epoch": 2.26368,
+      "grad_norm": 0.5506880283355713,
+      "learning_rate": 1.7180164442153529e-06,
+      "loss": 0.2015,
+      "step": 1178
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.4742065966129303,
+      "learning_rate": 1.709584256274106e-06,
+      "loss": 0.1651,
+      "step": 1179
+    },
+    {
+      "epoch": 2.26752,
+      "grad_norm": 0.4847202003002167,
+      "learning_rate": 1.7011685429800596e-06,
+      "loss": 0.1951,
+      "step": 1180
+    },
+    {
+      "epoch": 2.26944,
+      "grad_norm": 0.4227399528026581,
+      "learning_rate": 1.6927693464695022e-06,
+      "loss": 0.1487,
+      "step": 1181
+    },
+    {
+      "epoch": 2.27136,
+      "grad_norm": 0.4895167350769043,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.1682,
+      "step": 1182
+    },
+    {
+      "epoch": 2.27328,
+      "grad_norm": 0.4704173803329468,
+      "learning_rate": 1.6760206719303107e-06,
+      "loss": 0.1844,
+      "step": 1183
+    },
+    {
+      "epoch": 2.2752,
+      "grad_norm": 0.48153454065322876,
+      "learning_rate": 1.6676712777599275e-06,
+      "loss": 0.1746,
+      "step": 1184
+    },
+    {
+      "epoch": 2.27712,
+      "grad_norm": 0.4724951386451721,
+      "learning_rate": 1.6593385680891139e-06,
+      "loss": 0.1789,
+      "step": 1185
+    },
+    {
+      "epoch": 2.27904,
+      "grad_norm": 0.4721076786518097,
+      "learning_rate": 1.6510225846385668e-06,
+      "loss": 0.1578,
+      "step": 1186
+    },
+    {
+      "epoch": 2.28096,
+      "grad_norm": 0.4753687381744385,
+      "learning_rate": 1.6427233690452455e-06,
+      "loss": 0.1795,
+      "step": 1187
+    },
+    {
+      "epoch": 2.28288,
+      "grad_norm": 0.55759596824646,
+      "learning_rate": 1.6344409628621482e-06,
+      "loss": 0.192,
+      "step": 1188
+    },
+    {
+      "epoch": 2.2848,
+      "grad_norm": 0.452008992433548,
+      "learning_rate": 1.6261754075581187e-06,
+      "loss": 0.1432,
+      "step": 1189
+    },
+    {
+      "epoch": 2.28672,
+      "grad_norm": 0.4688599705696106,
+      "learning_rate": 1.6179267445176206e-06,
+      "loss": 0.1736,
+      "step": 1190
+    },
+    {
+      "epoch": 2.28864,
+      "grad_norm": 0.5417355895042419,
+      "learning_rate": 1.6096950150405454e-06,
+      "loss": 0.2014,
+      "step": 1191
+    },
+    {
+      "epoch": 2.29056,
+      "grad_norm": 0.5059545040130615,
+      "learning_rate": 1.6014802603420044e-06,
+      "loss": 0.2058,
+      "step": 1192
+    },
+    {
+      "epoch": 2.29248,
+      "grad_norm": 0.48315078020095825,
+      "learning_rate": 1.593282521552113e-06,
+      "loss": 0.1781,
+      "step": 1193
+    },
+    {
+      "epoch": 2.2944,
+      "grad_norm": 0.476159006357193,
+      "learning_rate": 1.5851018397157918e-06,
+      "loss": 0.1575,
+      "step": 1194
+    },
+    {
+      "epoch": 2.29632,
+      "grad_norm": 0.4625089764595032,
+      "learning_rate": 1.5769382557925612e-06,
+      "loss": 0.1453,
+      "step": 1195
+    },
+    {
+      "epoch": 2.29824,
+      "grad_norm": 0.5066043734550476,
+      "learning_rate": 1.5687918106563326e-06,
+      "loss": 0.1638,
+      "step": 1196
+    },
+    {
+      "epoch": 2.30016,
+      "grad_norm": 0.529657781124115,
+      "learning_rate": 1.5606625450952062e-06,
+      "loss": 0.1784,
+      "step": 1197
+    },
+    {
+      "epoch": 2.30208,
+      "grad_norm": 0.4992857277393341,
+      "learning_rate": 1.5525504998112717e-06,
+      "loss": 0.2021,
+      "step": 1198
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.4696641266345978,
+      "learning_rate": 1.5444557154203892e-06,
+      "loss": 0.1502,
+      "step": 1199
+    },
+    {
+      "epoch": 2.30592,
+      "grad_norm": 0.5318266749382019,
+      "learning_rate": 1.5363782324520033e-06,
+      "loss": 0.1863,
+      "step": 1200
+    },
+    {
+      "epoch": 2.30784,
+      "grad_norm": 0.4877982437610626,
+      "learning_rate": 1.5283180913489326e-06,
+      "loss": 0.1483,
+      "step": 1201
+    },
+    {
+      "epoch": 2.30976,
+      "grad_norm": 0.5429249405860901,
+      "learning_rate": 1.520275332467166e-06,
+      "loss": 0.1794,
+      "step": 1202
+    },
+    {
+      "epoch": 2.31168,
+      "grad_norm": 0.5508140921592712,
+      "learning_rate": 1.5122499960756604e-06,
+      "loss": 0.1971,
+      "step": 1203
+    },
+    {
+      "epoch": 2.3136,
+      "grad_norm": 0.527315080165863,
+      "learning_rate": 1.504242122356143e-06,
+      "loss": 0.2118,
+      "step": 1204
+    },
+    {
+      "epoch": 2.3155200000000002,
+      "grad_norm": 0.5142102837562561,
+      "learning_rate": 1.4962517514029069e-06,
+      "loss": 0.1738,
+      "step": 1205
+    },
+    {
+      "epoch": 2.31744,
+      "grad_norm": 0.5102952718734741,
+      "learning_rate": 1.4882789232226124e-06,
+      "loss": 0.1748,
+      "step": 1206
+    },
+    {
+      "epoch": 2.31936,
+      "grad_norm": 0.48686060309410095,
+      "learning_rate": 1.4803236777340856e-06,
+      "loss": 0.1687,
+      "step": 1207
+    },
+    {
+      "epoch": 2.32128,
+      "grad_norm": 0.5023792386054993,
+      "learning_rate": 1.4723860547681163e-06,
+      "loss": 0.1866,
+      "step": 1208
+    },
+    {
+      "epoch": 2.3232,
+      "grad_norm": 0.5040345788002014,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.1679,
+      "step": 1209
+    },
+    {
+      "epoch": 2.32512,
+      "grad_norm": 0.5475457906723022,
+      "learning_rate": 1.4565638352856504e-06,
+      "loss": 0.2266,
+      "step": 1210
+    },
+    {
+      "epoch": 2.32704,
+      "grad_norm": 0.5026825070381165,
+      "learning_rate": 1.4486793179887738e-06,
+      "loss": 0.1577,
+      "step": 1211
+    },
+    {
+      "epoch": 2.32896,
+      "grad_norm": 0.5351523756980896,
+      "learning_rate": 1.4408125816532981e-06,
+      "loss": 0.1858,
+      "step": 1212
+    },
+    {
+      "epoch": 2.33088,
+      "grad_norm": 0.5303723812103271,
+      "learning_rate": 1.4329636656668617e-06,
+      "loss": 0.1839,
+      "step": 1213
+    },
+    {
+      "epoch": 2.3327999999999998,
+      "grad_norm": 0.45616480708122253,
+      "learning_rate": 1.4251326093278773e-06,
+      "loss": 0.16,
+      "step": 1214
+    },
+    {
+      "epoch": 2.33472,
+      "grad_norm": 0.4990748167037964,
+      "learning_rate": 1.4173194518453415e-06,
+      "loss": 0.2081,
+      "step": 1215
+    },
+    {
+      "epoch": 2.33664,
+      "grad_norm": 0.4640475809574127,
+      "learning_rate": 1.4095242323386305e-06,
+      "loss": 0.1749,
+      "step": 1216
+    },
+    {
+      "epoch": 2.33856,
+      "grad_norm": 0.5655854940414429,
+      "learning_rate": 1.4017469898373077e-06,
+      "loss": 0.1753,
+      "step": 1217
+    },
+    {
+      "epoch": 2.34048,
+      "grad_norm": 0.50837242603302,
+      "learning_rate": 1.3939877632809279e-06,
+      "loss": 0.1606,
+      "step": 1218
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.5742613673210144,
+      "learning_rate": 1.3862465915188427e-06,
+      "loss": 0.1892,
+      "step": 1219
+    },
+    {
+      "epoch": 2.34432,
+      "grad_norm": 0.4863821268081665,
+      "learning_rate": 1.3785235133100088e-06,
+      "loss": 0.1535,
+      "step": 1220
+    },
+    {
+      "epoch": 2.34624,
+      "grad_norm": 0.5114768147468567,
+      "learning_rate": 1.3708185673227896e-06,
+      "loss": 0.1931,
+      "step": 1221
+    },
+    {
+      "epoch": 2.34816,
+      "grad_norm": 0.43559226393699646,
+      "learning_rate": 1.3631317921347564e-06,
+      "loss": 0.1527,
+      "step": 1222
+    },
+    {
+      "epoch": 2.35008,
+      "grad_norm": 0.5442113280296326,
+      "learning_rate": 1.3554632262325129e-06,
+      "loss": 0.2289,
+      "step": 1223
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 0.5286880731582642,
+      "learning_rate": 1.347812908011485e-06,
+      "loss": 0.2013,
+      "step": 1224
+    },
+    {
+      "epoch": 2.35392,
+      "grad_norm": 0.5553655624389648,
+      "learning_rate": 1.340180875775735e-06,
+      "loss": 0.1997,
+      "step": 1225
+    },
+    {
+      "epoch": 2.35584,
+      "grad_norm": 0.5233193635940552,
+      "learning_rate": 1.3325671677377727e-06,
+      "loss": 0.2287,
+      "step": 1226
+    },
+    {
+      "epoch": 2.35776,
+      "grad_norm": 0.483112096786499,
+      "learning_rate": 1.3249718220183583e-06,
+      "loss": 0.1698,
+      "step": 1227
+    },
+    {
+      "epoch": 2.35968,
+      "grad_norm": 0.5229559540748596,
+      "learning_rate": 1.3173948766463146e-06,
+      "loss": 0.1843,
+      "step": 1228
+    },
+    {
+      "epoch": 2.3616,
+      "grad_norm": 0.48514819145202637,
+      "learning_rate": 1.309836369558341e-06,
+      "loss": 0.1748,
+      "step": 1229
+    },
+    {
+      "epoch": 2.36352,
+      "grad_norm": 0.5285651683807373,
+      "learning_rate": 1.3022963385988153e-06,
+      "loss": 0.183,
+      "step": 1230
+    },
+    {
+      "epoch": 2.36544,
+      "grad_norm": 0.47766050696372986,
+      "learning_rate": 1.2947748215196038e-06,
+      "loss": 0.1539,
+      "step": 1231
+    },
+    {
+      "epoch": 2.36736,
+      "grad_norm": 0.5245499014854431,
+      "learning_rate": 1.2872718559798852e-06,
+      "loss": 0.1591,
+      "step": 1232
+    },
+    {
+      "epoch": 2.36928,
+      "grad_norm": 0.5019723176956177,
+      "learning_rate": 1.2797874795459464e-06,
+      "loss": 0.1559,
+      "step": 1233
+    },
+    {
+      "epoch": 2.3712,
+      "grad_norm": 0.5391585230827332,
+      "learning_rate": 1.2723217296910078e-06,
+      "loss": 0.2063,
+      "step": 1234
+    },
+    {
+      "epoch": 2.37312,
+      "grad_norm": 0.4685705304145813,
+      "learning_rate": 1.264874643795021e-06,
+      "loss": 0.179,
+      "step": 1235
+    },
+    {
+      "epoch": 2.37504,
+      "grad_norm": 0.4827197790145874,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.1874,
+      "step": 1236
+    },
+    {
+      "epoch": 2.37696,
+      "grad_norm": 0.5373876094818115,
+      "learning_rate": 1.2500366129323039e-06,
+      "loss": 0.2074,
+      "step": 1237
+    },
+    {
+      "epoch": 2.37888,
+      "grad_norm": 0.4884423315525055,
+      "learning_rate": 1.2426457422575e-06,
+      "loss": 0.1563,
+      "step": 1238
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.46749457716941833,
+      "learning_rate": 1.2352736841251306e-06,
+      "loss": 0.1504,
+      "step": 1239
+    },
+    {
+      "epoch": 2.38272,
+      "grad_norm": 0.49014419317245483,
+      "learning_rate": 1.2279204754460494e-06,
+      "loss": 0.1742,
+      "step": 1240
+    },
+    {
+      "epoch": 2.38464,
+      "grad_norm": 0.49345406889915466,
+      "learning_rate": 1.2205861530367342e-06,
+      "loss": 0.1509,
+      "step": 1241
+    },
+    {
+      "epoch": 2.3865600000000002,
+      "grad_norm": 0.49067360162734985,
+      "learning_rate": 1.2132707536191008e-06,
+      "loss": 0.1721,
+      "step": 1242
+    },
+    {
+      "epoch": 2.38848,
+      "grad_norm": 0.5196910500526428,
+      "learning_rate": 1.2059743138203256e-06,
+      "loss": 0.177,
+      "step": 1243
+    },
+    {
+      "epoch": 2.3904,
+      "grad_norm": 0.4899502396583557,
+      "learning_rate": 1.1986968701726492e-06,
+      "loss": 0.1787,
+      "step": 1244
+    },
+    {
+      "epoch": 2.39232,
+      "grad_norm": 0.4909348487854004,
+      "learning_rate": 1.1914384591132045e-06,
+      "loss": 0.1622,
+      "step": 1245
+    },
+    {
+      "epoch": 2.39424,
+      "grad_norm": 0.4723018407821655,
+      "learning_rate": 1.184199116983834e-06,
+      "loss": 0.1736,
+      "step": 1246
+    },
+    {
+      "epoch": 2.39616,
+      "grad_norm": 0.4905262589454651,
+      "learning_rate": 1.1769788800309001e-06,
+      "loss": 0.1706,
+      "step": 1247
+    },
+    {
+      "epoch": 2.39808,
+      "grad_norm": 0.45472776889801025,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.1701,
+      "step": 1248
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.5217065811157227,
+      "learning_rate": 1.1625958661613345e-06,
+      "loss": 0.1979,
+      "step": 1249
+    },
+    {
+      "epoch": 2.40192,
+      "grad_norm": 0.5163058042526245,
+      "learning_rate": 1.1554331612584218e-06,
+      "loss": 0.1698,
+      "step": 1250
+    },
+    {
+      "epoch": 2.4038399999999998,
+      "grad_norm": 0.5149426460266113,
+      "learning_rate": 1.1482897055590275e-06,
+      "loss": 0.1634,
+      "step": 1251
+    },
+    {
+      "epoch": 2.40576,
+      "grad_norm": 0.49433931708335876,
+      "learning_rate": 1.141165534829425e-06,
+      "loss": 0.1681,
+      "step": 1252
+    },
+    {
+      "epoch": 2.40768,
+      "grad_norm": 0.460930198431015,
+      "learning_rate": 1.134060684739331e-06,
+      "loss": 0.1549,
+      "step": 1253
+    },
+    {
+      "epoch": 2.4096,
+      "grad_norm": 0.5268895030021667,
+      "learning_rate": 1.1269751908617277e-06,
+      "loss": 0.1864,
+      "step": 1254
+    },
+    {
+      "epoch": 2.41152,
+      "grad_norm": 0.4987984299659729,
+      "learning_rate": 1.119909088672682e-06,
+      "loss": 0.1903,
+      "step": 1255
+    },
+    {
+      "epoch": 2.41344,
+      "grad_norm": 0.5444393157958984,
+      "learning_rate": 1.1128624135511712e-06,
+      "loss": 0.1749,
+      "step": 1256
+    },
+    {
+      "epoch": 2.41536,
+      "grad_norm": 0.4734117388725281,
+      "learning_rate": 1.105835200778902e-06,
+      "loss": 0.1718,
+      "step": 1257
+    },
+    {
+      "epoch": 2.41728,
+      "grad_norm": 0.4921054542064667,
+      "learning_rate": 1.0988274855401377e-06,
+      "loss": 0.1859,
+      "step": 1258
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 0.4445892572402954,
+      "learning_rate": 1.091839302921518e-06,
+      "loss": 0.1507,
+      "step": 1259
+    },
+    {
+      "epoch": 2.42112,
+      "grad_norm": 0.4863232374191284,
+      "learning_rate": 1.0848706879118893e-06,
+      "loss": 0.169,
+      "step": 1260
+    },
+    {
+      "epoch": 2.42304,
+      "grad_norm": 0.512374997138977,
+      "learning_rate": 1.0779216754021215e-06,
+      "loss": 0.1782,
+      "step": 1261
+    },
+    {
+      "epoch": 2.42496,
+      "grad_norm": 0.49184733629226685,
+      "learning_rate": 1.070992300184941e-06,
+      "loss": 0.18,
+      "step": 1262
+    },
+    {
+      "epoch": 2.42688,
+      "grad_norm": 0.48721885681152344,
+      "learning_rate": 1.0640825969547498e-06,
+      "loss": 0.171,
+      "step": 1263
+    },
+    {
+      "epoch": 2.4288,
+      "grad_norm": 0.4348336160182953,
+      "learning_rate": 1.057192600307456e-06,
+      "loss": 0.1393,
+      "step": 1264
+    },
+    {
+      "epoch": 2.43072,
+      "grad_norm": 0.48496755957603455,
+      "learning_rate": 1.0503223447403033e-06,
+      "loss": 0.2022,
+      "step": 1265
+    },
+    {
+      "epoch": 2.43264,
+      "grad_norm": 0.5035723447799683,
+      "learning_rate": 1.0434718646516917e-06,
+      "loss": 0.1644,
+      "step": 1266
+    },
+    {
+      "epoch": 2.43456,
+      "grad_norm": 0.5311614274978638,
+      "learning_rate": 1.0366411943410033e-06,
+      "loss": 0.1695,
+      "step": 1267
+    },
+    {
+      "epoch": 2.43648,
+      "grad_norm": 0.5220187306404114,
+      "learning_rate": 1.0298303680084448e-06,
+      "loss": 0.1985,
+      "step": 1268
+    },
+    {
+      "epoch": 2.4384,
+      "grad_norm": 0.5018011927604675,
+      "learning_rate": 1.0230394197548605e-06,
+      "loss": 0.1732,
+      "step": 1269
+    },
+    {
+      "epoch": 2.44032,
+      "grad_norm": 0.45088672637939453,
+      "learning_rate": 1.0162683835815706e-06,
+      "loss": 0.1813,
+      "step": 1270
+    },
+    {
+      "epoch": 2.44224,
+      "grad_norm": 0.4448988735675812,
+      "learning_rate": 1.009517293390197e-06,
+      "loss": 0.1618,
+      "step": 1271
+    },
+    {
+      "epoch": 2.44416,
+      "grad_norm": 0.6034268736839294,
+      "learning_rate": 1.0027861829824953e-06,
+      "loss": 0.1973,
+      "step": 1272
+    },
+    {
+      "epoch": 2.44608,
+      "grad_norm": 0.49048563838005066,
+      "learning_rate": 9.960750860601842e-07,
+      "loss": 0.1623,
+      "step": 1273
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 0.5347757935523987,
+      "learning_rate": 9.893840362247809e-07,
+      "loss": 0.1964,
+      "step": 1274
+    },
+    {
+      "epoch": 2.44992,
+      "grad_norm": 0.5404366850852966,
+      "learning_rate": 9.82713066977427e-07,
+      "loss": 0.2058,
+      "step": 1275
+    },
+    {
+      "epoch": 2.45184,
+      "grad_norm": 0.4700262248516083,
+      "learning_rate": 9.760622117187234e-07,
+      "loss": 0.1683,
+      "step": 1276
+    },
+    {
+      "epoch": 2.45376,
+      "grad_norm": 0.482604444026947,
+      "learning_rate": 9.694315037485635e-07,
+      "loss": 0.1751,
+      "step": 1277
+    },
+    {
+      "epoch": 2.45568,
+      "grad_norm": 0.49182188510894775,
+      "learning_rate": 9.628209762659658e-07,
+      "loss": 0.1944,
+      "step": 1278
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.5031192302703857,
+      "learning_rate": 9.562306623689111e-07,
+      "loss": 0.1633,
+      "step": 1279
+    },
+    {
+      "epoch": 2.45952,
+      "grad_norm": 0.5462546348571777,
+      "learning_rate": 9.496605950541676e-07,
+      "loss": 0.2106,
+      "step": 1280
+    },
+    {
+      "epoch": 2.46144,
+      "grad_norm": 0.4601513147354126,
+      "learning_rate": 9.431108072171346e-07,
+      "loss": 0.1428,
+      "step": 1281
+    },
+    {
+      "epoch": 2.4633599999999998,
+      "grad_norm": 0.48662805557250977,
+      "learning_rate": 9.365813316516787e-07,
+      "loss": 0.176,
+      "step": 1282
+    },
+    {
+      "epoch": 2.46528,
+      "grad_norm": 0.52548748254776,
+      "learning_rate": 9.300722010499608e-07,
+      "loss": 0.1923,
+      "step": 1283
+    },
+    {
+      "epoch": 2.4672,
+      "grad_norm": 0.4850045144557953,
+      "learning_rate": 9.235834480022788e-07,
+      "loss": 0.1636,
+      "step": 1284
+    },
+    {
+      "epoch": 2.46912,
+      "grad_norm": 0.5413723587989807,
+      "learning_rate": 9.17115104996903e-07,
+      "loss": 0.2038,
+      "step": 1285
+    },
+    {
+      "epoch": 2.47104,
+      "grad_norm": 0.4840810298919678,
+      "learning_rate": 9.10667204419915e-07,
+      "loss": 0.1624,
+      "step": 1286
+    },
+    {
+      "epoch": 2.47296,
+      "grad_norm": 0.45478230714797974,
+      "learning_rate": 9.042397785550405e-07,
+      "loss": 0.1381,
+      "step": 1287
+    },
+    {
+      "epoch": 2.47488,
+      "grad_norm": 0.49161583185195923,
+      "learning_rate": 8.978328595834984e-07,
+      "loss": 0.172,
+      "step": 1288
+    },
+    {
+      "epoch": 2.4768,
+      "grad_norm": 0.5194839835166931,
+      "learning_rate": 8.91446479583823e-07,
+      "loss": 0.1733,
+      "step": 1289
+    },
+    {
+      "epoch": 2.47872,
+      "grad_norm": 0.4773405194282532,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.1586,
+      "step": 1290
+    },
+    {
+      "epoch": 2.48064,
+      "grad_norm": 0.5241573452949524,
+      "learning_rate": 8.787354642998936e-07,
+      "loss": 0.2182,
+      "step": 1291
+    },
+    {
+      "epoch": 2.48256,
+      "grad_norm": 0.4749302864074707,
+      "learning_rate": 8.724108926579e-07,
+      "loss": 0.1744,
+      "step": 1292
+    },
+    {
+      "epoch": 2.48448,
+      "grad_norm": 0.5018808245658875,
+      "learning_rate": 8.661069872719746e-07,
+      "loss": 0.1735,
+      "step": 1293
+    },
+    {
+      "epoch": 2.4864,
+      "grad_norm": 0.4519702196121216,
+      "learning_rate": 8.598237797048825e-07,
+      "loss": 0.1682,
+      "step": 1294
+    },
+    {
+      "epoch": 2.48832,
+      "grad_norm": 0.46126750111579895,
+      "learning_rate": 8.535613014157556e-07,
+      "loss": 0.1546,
+      "step": 1295
+    },
+    {
+      "epoch": 2.49024,
+      "grad_norm": 0.4593077003955841,
+      "learning_rate": 8.473195837599419e-07,
+      "loss": 0.1697,
+      "step": 1296
+    },
+    {
+      "epoch": 2.49216,
+      "grad_norm": 0.49476540088653564,
+      "learning_rate": 8.410986579888381e-07,
+      "loss": 0.1807,
+      "step": 1297
+    },
+    {
+      "epoch": 2.49408,
+      "grad_norm": 0.5562830567359924,
+      "learning_rate": 8.348985552497424e-07,
+      "loss": 0.2005,
+      "step": 1298
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.5080939531326294,
+      "learning_rate": 8.287193065856936e-07,
+      "loss": 0.1784,
+      "step": 1299
+    },
+    {
+      "epoch": 2.49792,
+      "grad_norm": 0.45330992341041565,
+      "learning_rate": 8.225609429353187e-07,
+      "loss": 0.1523,
+      "step": 1300
+    },
+    {
+      "epoch": 2.49984,
+      "grad_norm": 0.4720747172832489,
+      "learning_rate": 8.164234951326727e-07,
+      "loss": 0.161,
+      "step": 1301
+    },
+    {
+      "epoch": 2.50176,
+      "grad_norm": 0.47904059290885925,
+      "learning_rate": 8.103069939070945e-07,
+      "loss": 0.1666,
+      "step": 1302
+    },
+    {
+      "epoch": 2.50368,
+      "grad_norm": 0.5755215287208557,
+      "learning_rate": 8.042114698830394e-07,
+      "loss": 0.1914,
+      "step": 1303
+    },
+    {
+      "epoch": 2.5056000000000003,
+      "grad_norm": 0.4466748535633087,
+      "learning_rate": 7.981369535799354e-07,
+      "loss": 0.1803,
+      "step": 1304
+    },
+    {
+      "epoch": 2.50752,
+      "grad_norm": 0.5399540662765503,
+      "learning_rate": 7.920834754120305e-07,
+      "loss": 0.1926,
+      "step": 1305
+    },
+    {
+      "epoch": 2.50944,
+      "grad_norm": 0.5035436749458313,
+      "learning_rate": 7.860510656882342e-07,
+      "loss": 0.1782,
+      "step": 1306
+    },
+    {
+      "epoch": 2.51136,
+      "grad_norm": 0.5267941355705261,
+      "learning_rate": 7.800397546119709e-07,
+      "loss": 0.205,
+      "step": 1307
+    },
+    {
+      "epoch": 2.51328,
+      "grad_norm": 0.45300984382629395,
+      "learning_rate": 7.740495722810271e-07,
+      "loss": 0.1539,
+      "step": 1308
+    },
+    {
+      "epoch": 2.5152,
+      "grad_norm": 0.5562019348144531,
+      "learning_rate": 7.680805486873977e-07,
+      "loss": 0.2204,
+      "step": 1309
+    },
+    {
+      "epoch": 2.5171200000000002,
+      "grad_norm": 0.5053987503051758,
+      "learning_rate": 7.621327137171447e-07,
+      "loss": 0.1736,
+      "step": 1310
+    },
+    {
+      "epoch": 2.51904,
+      "grad_norm": 0.4817841947078705,
+      "learning_rate": 7.562060971502383e-07,
+      "loss": 0.1516,
+      "step": 1311
+    },
+    {
+      "epoch": 2.52096,
+      "grad_norm": 0.48662975430488586,
+      "learning_rate": 7.50300728660407e-07,
+      "loss": 0.146,
+      "step": 1312
+    },
+    {
+      "epoch": 2.52288,
+      "grad_norm": 0.5119237303733826,
+      "learning_rate": 7.444166378150014e-07,
+      "loss": 0.1752,
+      "step": 1313
+    },
+    {
+      "epoch": 2.5248,
+      "grad_norm": 0.5494744777679443,
+      "learning_rate": 7.385538540748327e-07,
+      "loss": 0.191,
+      "step": 1314
+    },
+    {
+      "epoch": 2.52672,
+      "grad_norm": 0.5031672120094299,
+      "learning_rate": 7.327124067940311e-07,
+      "loss": 0.1658,
+      "step": 1315
+    },
+    {
+      "epoch": 2.52864,
+      "grad_norm": 0.4784882068634033,
+      "learning_rate": 7.26892325219899e-07,
+      "loss": 0.1676,
+      "step": 1316
+    },
+    {
+      "epoch": 2.53056,
+      "grad_norm": 0.48420923948287964,
+      "learning_rate": 7.210936384927631e-07,
+      "loss": 0.1639,
+      "step": 1317
+    },
+    {
+      "epoch": 2.53248,
+      "grad_norm": 0.5364359617233276,
+      "learning_rate": 7.153163756458287e-07,
+      "loss": 0.2259,
+      "step": 1318
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.4589465856552124,
+      "learning_rate": 7.09560565605037e-07,
+      "loss": 0.1465,
+      "step": 1319
+    },
+    {
+      "epoch": 2.53632,
+      "grad_norm": 0.4575812816619873,
+      "learning_rate": 7.03826237188916e-07,
+      "loss": 0.1506,
+      "step": 1320
+    },
+    {
+      "epoch": 2.53824,
+      "grad_norm": 0.518252432346344,
+      "learning_rate": 6.981134191084388e-07,
+      "loss": 0.1898,
+      "step": 1321
+    },
+    {
+      "epoch": 2.54016,
+      "grad_norm": 0.4594062268733978,
+      "learning_rate": 6.924221399668785e-07,
+      "loss": 0.1701,
+      "step": 1322
+    },
+    {
+      "epoch": 2.54208,
+      "grad_norm": 0.4566185176372528,
+      "learning_rate": 6.867524282596655e-07,
+      "loss": 0.1621,
+      "step": 1323
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 0.5680291652679443,
+      "learning_rate": 6.811043123742494e-07,
+      "loss": 0.1969,
+      "step": 1324
+    },
+    {
+      "epoch": 2.5459199999999997,
+      "grad_norm": 0.5100989937782288,
+      "learning_rate": 6.754778205899465e-07,
+      "loss": 0.1629,
+      "step": 1325
+    },
+    {
+      "epoch": 2.54784,
+      "grad_norm": 0.4661482274532318,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.1657,
+      "step": 1326
+    },
+    {
+      "epoch": 2.54976,
+      "grad_norm": 0.4511263072490692,
+      "learning_rate": 6.642898219004723e-07,
+      "loss": 0.1427,
+      "step": 1327
+    },
+    {
+      "epoch": 2.55168,
+      "grad_norm": 0.4931066036224365,
+      "learning_rate": 6.587283710120324e-07,
+      "loss": 0.176,
+      "step": 1328
+    },
+    {
+      "epoch": 2.5536,
+      "grad_norm": 0.4824303090572357,
+      "learning_rate": 6.531886562578859e-07,
+      "loss": 0.1738,
+      "step": 1329
+    },
+    {
+      "epoch": 2.55552,
+      "grad_norm": 0.4820888042449951,
+      "learning_rate": 6.47670705374604e-07,
+      "loss": 0.1615,
+      "step": 1330
+    },
+    {
+      "epoch": 2.55744,
+      "grad_norm": 0.4837888181209564,
+      "learning_rate": 6.421745459897871e-07,
+      "loss": 0.191,
+      "step": 1331
+    },
+    {
+      "epoch": 2.55936,
+      "grad_norm": 0.4818747341632843,
+      "learning_rate": 6.367002056219285e-07,
+      "loss": 0.1524,
+      "step": 1332
+    },
+    {
+      "epoch": 2.56128,
+      "grad_norm": 0.4666469097137451,
+      "learning_rate": 6.312477116802807e-07,
+      "loss": 0.1524,
+      "step": 1333
+    },
+    {
+      "epoch": 2.5632,
+      "grad_norm": 0.4864393472671509,
+      "learning_rate": 6.258170914647077e-07,
+      "loss": 0.1785,
+      "step": 1334
+    },
+    {
+      "epoch": 2.56512,
+      "grad_norm": 0.45313307642936707,
+      "learning_rate": 6.204083721655607e-07,
+      "loss": 0.1471,
+      "step": 1335
+    },
+    {
+      "epoch": 2.56704,
+      "grad_norm": 0.614441454410553,
+      "learning_rate": 6.150215808635334e-07,
+      "loss": 0.1926,
+      "step": 1336
+    },
+    {
+      "epoch": 2.56896,
+      "grad_norm": 0.4992642402648926,
+      "learning_rate": 6.096567445295298e-07,
+      "loss": 0.1954,
+      "step": 1337
+    },
+    {
+      "epoch": 2.57088,
+      "grad_norm": 0.48243018984794617,
+      "learning_rate": 6.043138900245277e-07,
+      "loss": 0.1583,
+      "step": 1338
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.5750699639320374,
+      "learning_rate": 5.989930440994451e-07,
+      "loss": 0.2119,
+      "step": 1339
+    },
+    {
+      "epoch": 2.57472,
+      "grad_norm": 0.49085164070129395,
+      "learning_rate": 5.936942333950063e-07,
+      "loss": 0.1751,
+      "step": 1340
+    },
+    {
+      "epoch": 2.5766400000000003,
+      "grad_norm": 0.44255247712135315,
+      "learning_rate": 5.884174844416102e-07,
+      "loss": 0.1477,
+      "step": 1341
+    },
+    {
+      "epoch": 2.57856,
+      "grad_norm": 0.4399384558200836,
+      "learning_rate": 5.831628236591929e-07,
+      "loss": 0.1379,
+      "step": 1342
+    },
+    {
+      "epoch": 2.58048,
+      "grad_norm": 0.4502022862434387,
+      "learning_rate": 5.779302773570994e-07,
+      "loss": 0.1613,
+      "step": 1343
+    },
+    {
+      "epoch": 2.5824,
+      "grad_norm": 0.45515891909599304,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.1595,
+      "step": 1344
+    },
+    {
+      "epoch": 2.58432,
+      "grad_norm": 0.4777766168117523,
+      "learning_rate": 5.675316328775126e-07,
+      "loss": 0.1748,
+      "step": 1345
+    },
+    {
+      "epoch": 2.58624,
+      "grad_norm": 0.5000216960906982,
+      "learning_rate": 5.623655867645628e-07,
+      "loss": 0.1841,
+      "step": 1346
+    },
+    {
+      "epoch": 2.5881600000000002,
+      "grad_norm": 0.3877320885658264,
+      "learning_rate": 5.572217592607687e-07,
+      "loss": 0.1304,
+      "step": 1347
+    },
+    {
+      "epoch": 2.59008,
+      "grad_norm": 0.502407968044281,
+      "learning_rate": 5.521001761205441e-07,
+      "loss": 0.1983,
+      "step": 1348
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 0.499590665102005,
+      "learning_rate": 5.470008629869367e-07,
+      "loss": 0.1854,
+      "step": 1349
+    },
+    {
+      "epoch": 2.59392,
+      "grad_norm": 0.5306006073951721,
+      "learning_rate": 5.41923845391486e-07,
+      "loss": 0.1781,
+      "step": 1350
+    },
+    {
+      "epoch": 2.59584,
+      "grad_norm": 0.5286628603935242,
+      "learning_rate": 5.368691487541027e-07,
+      "loss": 0.1782,
+      "step": 1351
+    },
+    {
+      "epoch": 2.59776,
+      "grad_norm": 0.5221961140632629,
+      "learning_rate": 5.318367983829393e-07,
+      "loss": 0.1594,
+      "step": 1352
+    },
+    {
+      "epoch": 2.59968,
+      "grad_norm": 0.4815419614315033,
+      "learning_rate": 5.268268194742638e-07,
+      "loss": 0.1738,
+      "step": 1353
+    },
+    {
+      "epoch": 2.6016,
+      "grad_norm": 0.4583428204059601,
+      "learning_rate": 5.218392371123326e-07,
+      "loss": 0.1454,
+      "step": 1354
+    },
+    {
+      "epoch": 2.60352,
+      "grad_norm": 0.5484034419059753,
+      "learning_rate": 5.168740762692681e-07,
+      "loss": 0.1943,
+      "step": 1355
+    },
+    {
+      "epoch": 2.6054399999999998,
+      "grad_norm": 0.609625518321991,
+      "learning_rate": 5.119313618049309e-07,
+      "loss": 0.2222,
+      "step": 1356
+    },
+    {
+      "epoch": 2.60736,
+      "grad_norm": 0.45864659547805786,
+      "learning_rate": 5.070111184667908e-07,
+      "loss": 0.1435,
+      "step": 1357
+    },
+    {
+      "epoch": 2.60928,
+      "grad_norm": 0.536439836025238,
+      "learning_rate": 5.021133708898146e-07,
+      "loss": 0.2089,
+      "step": 1358
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.4905323386192322,
+      "learning_rate": 4.972381435963331e-07,
+      "loss": 0.1719,
+      "step": 1359
+    },
+    {
+      "epoch": 2.61312,
+      "grad_norm": 0.525477945804596,
+      "learning_rate": 4.9238546099592e-07,
+      "loss": 0.1948,
+      "step": 1360
+    },
+    {
+      "epoch": 2.61504,
+      "grad_norm": 0.488993376493454,
+      "learning_rate": 4.875553473852735e-07,
+      "loss": 0.1505,
+      "step": 1361
+    },
+    {
+      "epoch": 2.6169599999999997,
+      "grad_norm": 0.5714892148971558,
+      "learning_rate": 4.827478269480895e-07,
+      "loss": 0.2252,
+      "step": 1362
+    },
+    {
+      "epoch": 2.61888,
+      "grad_norm": 0.4417460262775421,
+      "learning_rate": 4.779629237549438e-07,
+      "loss": 0.1511,
+      "step": 1363
+    },
+    {
+      "epoch": 2.6208,
+      "grad_norm": 0.48858878016471863,
+      "learning_rate": 4.732006617631729e-07,
+      "loss": 0.1866,
+      "step": 1364
+    },
+    {
+      "epoch": 2.62272,
+      "grad_norm": 0.549006998538971,
+      "learning_rate": 4.6846106481675035e-07,
+      "loss": 0.1854,
+      "step": 1365
+    },
+    {
+      "epoch": 2.62464,
+      "grad_norm": 0.4734170138835907,
+      "learning_rate": 4.637441566461681e-07,
+      "loss": 0.1552,
+      "step": 1366
+    },
+    {
+      "epoch": 2.62656,
+      "grad_norm": 0.4986458420753479,
+      "learning_rate": 4.590499608683202e-07,
+      "loss": 0.1757,
+      "step": 1367
+    },
+    {
+      "epoch": 2.62848,
+      "grad_norm": 0.491531103849411,
+      "learning_rate": 4.54378500986381e-07,
+      "loss": 0.1628,
+      "step": 1368
+    },
+    {
+      "epoch": 2.6304,
+      "grad_norm": 0.4851435124874115,
+      "learning_rate": 4.497298003896944e-07,
+      "loss": 0.1782,
+      "step": 1369
+    },
+    {
+      "epoch": 2.63232,
+      "grad_norm": 0.456939160823822,
+      "learning_rate": 4.451038823536441e-07,
+      "loss": 0.1365,
+      "step": 1370
+    },
+    {
+      "epoch": 2.63424,
+      "grad_norm": 0.601611316204071,
+      "learning_rate": 4.405007700395497e-07,
+      "loss": 0.2089,
+      "step": 1371
+    },
+    {
+      "epoch": 2.63616,
+      "grad_norm": 0.5148964524269104,
+      "learning_rate": 4.35920486494546e-07,
+      "loss": 0.2094,
+      "step": 1372
+    },
+    {
+      "epoch": 2.63808,
+      "grad_norm": 0.4972579777240753,
+      "learning_rate": 4.313630546514663e-07,
+      "loss": 0.1891,
+      "step": 1373
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.5453333854675293,
+      "learning_rate": 4.268284973287273e-07,
+      "loss": 0.1878,
+      "step": 1374
+    },
+    {
+      "epoch": 2.64192,
+      "grad_norm": 0.5713828802108765,
+      "learning_rate": 4.223168372302189e-07,
+      "loss": 0.2096,
+      "step": 1375
+    },
+    {
+      "epoch": 2.64384,
+      "grad_norm": 0.4665101170539856,
+      "learning_rate": 4.1782809694518533e-07,
+      "loss": 0.168,
+      "step": 1376
+    },
+    {
+      "epoch": 2.64576,
+      "grad_norm": 0.4873267412185669,
+      "learning_rate": 4.1336229894811454e-07,
+      "loss": 0.1607,
+      "step": 1377
+    },
+    {
+      "epoch": 2.6476800000000003,
+      "grad_norm": 0.5394080877304077,
+      "learning_rate": 4.089194655986306e-07,
+      "loss": 0.1888,
+      "step": 1378
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.5384389162063599,
+      "learning_rate": 4.044996191413686e-07,
+      "loss": 0.1846,
+      "step": 1379
+    },
+    {
+      "epoch": 2.65152,
+      "grad_norm": 0.4953325390815735,
+      "learning_rate": 4.001027817058789e-07,
+      "loss": 0.192,
+      "step": 1380
+    },
+    {
+      "epoch": 2.65344,
+      "grad_norm": 0.4761173725128174,
+      "learning_rate": 3.957289753065052e-07,
+      "loss": 0.1584,
+      "step": 1381
+    },
+    {
+      "epoch": 2.65536,
+      "grad_norm": 0.5375334024429321,
+      "learning_rate": 3.9137822184227845e-07,
+      "loss": 0.1904,
+      "step": 1382
+    },
+    {
+      "epoch": 2.65728,
+      "grad_norm": 0.4729098677635193,
+      "learning_rate": 3.870505430968069e-07,
+      "loss": 0.186,
+      "step": 1383
+    },
+    {
+      "epoch": 2.6592000000000002,
+      "grad_norm": 0.4598015248775482,
+      "learning_rate": 3.8274596073816784e-07,
+      "loss": 0.1571,
+      "step": 1384
+    },
+    {
+      "epoch": 2.66112,
+      "grad_norm": 0.4553985595703125,
+      "learning_rate": 3.7846449631879664e-07,
+      "loss": 0.1616,
+      "step": 1385
+    },
+    {
+      "epoch": 2.66304,
+      "grad_norm": 0.4976440370082855,
+      "learning_rate": 3.742061712753825e-07,
+      "loss": 0.1808,
+      "step": 1386
+    },
+    {
+      "epoch": 2.6649599999999998,
+      "grad_norm": 0.4888286888599396,
+      "learning_rate": 3.699710069287571e-07,
+      "loss": 0.1783,
+      "step": 1387
+    },
+    {
+      "epoch": 2.66688,
+      "grad_norm": 0.48596081137657166,
+      "learning_rate": 3.657590244837911e-07,
+      "loss": 0.1692,
+      "step": 1388
+    },
+    {
+      "epoch": 2.6688,
+      "grad_norm": 0.4889366924762726,
+      "learning_rate": 3.615702450292857e-07,
+      "loss": 0.1682,
+      "step": 1389
+    },
+    {
+      "epoch": 2.67072,
+      "grad_norm": 0.4651656746864319,
+      "learning_rate": 3.5740468953786854e-07,
+      "loss": 0.1532,
+      "step": 1390
+    },
+    {
+      "epoch": 2.67264,
+      "grad_norm": 0.47233784198760986,
+      "learning_rate": 3.5326237886588734e-07,
+      "loss": 0.1546,
+      "step": 1391
+    },
+    {
+      "epoch": 2.67456,
+      "grad_norm": 0.4697699546813965,
+      "learning_rate": 3.49143333753309e-07,
+      "loss": 0.1582,
+      "step": 1392
+    },
+    {
+      "epoch": 2.6764799999999997,
+      "grad_norm": 0.5064758062362671,
+      "learning_rate": 3.4504757482360817e-07,
+      "loss": 0.1769,
+      "step": 1393
+    },
+    {
+      "epoch": 2.6784,
+      "grad_norm": 0.5022813677787781,
+      "learning_rate": 3.4097512258367385e-07,
+      "loss": 0.1734,
+      "step": 1394
+    },
+    {
+      "epoch": 2.68032,
+      "grad_norm": 0.4810009300708771,
+      "learning_rate": 3.369259974236988e-07,
+      "loss": 0.189,
+      "step": 1395
+    },
+    {
+      "epoch": 2.68224,
+      "grad_norm": 0.4351924657821655,
+      "learning_rate": 3.3290021961708163e-07,
+      "loss": 0.1375,
+      "step": 1396
+    },
+    {
+      "epoch": 2.68416,
+      "grad_norm": 0.5674909353256226,
+      "learning_rate": 3.288978093203227e-07,
+      "loss": 0.1955,
+      "step": 1397
+    },
+    {
+      "epoch": 2.68608,
+      "grad_norm": 0.4637385606765747,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.1507,
+      "step": 1398
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.5261948108673096,
+      "learning_rate": 3.209631712972966e-07,
+      "loss": 0.1933,
+      "step": 1399
+    },
+    {
+      "epoch": 2.68992,
+      "grad_norm": 0.4813753664493561,
+      "learning_rate": 3.1703098329864237e-07,
+      "loss": 0.1592,
+      "step": 1400
+    },
+    {
+      "epoch": 2.69184,
+      "grad_norm": 0.4825710952281952,
+      "learning_rate": 3.131222422648744e-07,
+      "loss": 0.1873,
+      "step": 1401
+    },
+    {
+      "epoch": 2.69376,
+      "grad_norm": 0.5220139622688293,
+      "learning_rate": 3.0923696776650414e-07,
+      "loss": 0.1831,
+      "step": 1402
+    },
+    {
+      "epoch": 2.69568,
+      "grad_norm": 0.497307151556015,
+      "learning_rate": 3.0537517925655567e-07,
+      "loss": 0.1712,
+      "step": 1403
+    },
+    {
+      "epoch": 2.6976,
+      "grad_norm": 0.39419013261795044,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.1303,
+      "step": 1404
+    },
+    {
+      "epoch": 2.69952,
+      "grad_norm": 0.5207510590553284,
+      "learning_rate": 2.9772213742595367e-07,
+      "loss": 0.2105,
+      "step": 1405
+    },
+    {
+      "epoch": 2.70144,
+      "grad_norm": 0.48335328698158264,
+      "learning_rate": 2.9393092242300026e-07,
+      "loss": 0.1611,
+      "step": 1406
+    },
+    {
+      "epoch": 2.70336,
+      "grad_norm": 0.5887119174003601,
+      "learning_rate": 2.901632700436757e-07,
+      "loss": 0.2246,
+      "step": 1407
+    },
+    {
+      "epoch": 2.70528,
+      "grad_norm": 0.46260228753089905,
+      "learning_rate": 2.864191991520848e-07,
+      "loss": 0.1568,
+      "step": 1408
+    },
+    {
+      "epoch": 2.7072000000000003,
+      "grad_norm": 0.5043843388557434,
+      "learning_rate": 2.8269872849426114e-07,
+      "loss": 0.1802,
+      "step": 1409
+    },
+    {
+      "epoch": 2.70912,
+      "grad_norm": 0.46146297454833984,
+      "learning_rate": 2.790018766980773e-07,
+      "loss": 0.1584,
+      "step": 1410
+    },
+    {
+      "epoch": 2.71104,
+      "grad_norm": 0.4933132827281952,
+      "learning_rate": 2.7532866227314714e-07,
+      "loss": 0.1802,
+      "step": 1411
+    },
+    {
+      "epoch": 2.71296,
+      "grad_norm": 0.5017509460449219,
+      "learning_rate": 2.71679103610738e-07,
+      "loss": 0.1908,
+      "step": 1412
+    },
+    {
+      "epoch": 2.71488,
+      "grad_norm": 0.44843459129333496,
+      "learning_rate": 2.6805321898367323e-07,
+      "loss": 0.1747,
+      "step": 1413
+    },
+    {
+      "epoch": 2.7168,
+      "grad_norm": 0.5194143056869507,
+      "learning_rate": 2.64451026546248e-07,
+      "loss": 0.1825,
+      "step": 1414
+    },
+    {
+      "epoch": 2.7187200000000002,
+      "grad_norm": 0.5290195345878601,
+      "learning_rate": 2.6087254433412924e-07,
+      "loss": 0.2024,
+      "step": 1415
+    },
+    {
+      "epoch": 2.72064,
+      "grad_norm": 0.5048720836639404,
+      "learning_rate": 2.573177902642726e-07,
+      "loss": 0.1691,
+      "step": 1416
+    },
+    {
+      "epoch": 2.72256,
+      "grad_norm": 0.45962241291999817,
+      "learning_rate": 2.5378678213483057e-07,
+      "loss": 0.1493,
+      "step": 1417
+    },
+    {
+      "epoch": 2.72448,
+      "grad_norm": 0.5348236560821533,
+      "learning_rate": 2.502795376250622e-07,
+      "loss": 0.2316,
+      "step": 1418
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.5114087462425232,
+      "learning_rate": 2.467960742952463e-07,
+      "loss": 0.2066,
+      "step": 1419
+    },
+    {
+      "epoch": 2.72832,
+      "grad_norm": 0.5293125510215759,
+      "learning_rate": 2.4333640958659144e-07,
+      "loss": 0.2062,
+      "step": 1420
+    },
+    {
+      "epoch": 2.7302400000000002,
+      "grad_norm": 0.49702712893486023,
+      "learning_rate": 2.399005608211502e-07,
+      "loss": 0.1606,
+      "step": 1421
+    },
+    {
+      "epoch": 2.73216,
+      "grad_norm": 0.5160643458366394,
+      "learning_rate": 2.3648854520173237e-07,
+      "loss": 0.1954,
+      "step": 1422
+    },
+    {
+      "epoch": 2.73408,
+      "grad_norm": 0.44420406222343445,
+      "learning_rate": 2.3310037981182088e-07,
+      "loss": 0.156,
+      "step": 1423
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 0.49680715799331665,
+      "learning_rate": 2.2973608161547755e-07,
+      "loss": 0.1781,
+      "step": 1424
+    },
+    {
+      "epoch": 2.73792,
+      "grad_norm": 0.5032156109809875,
+      "learning_rate": 2.2639566745727203e-07,
+      "loss": 0.2062,
+      "step": 1425
+    },
+    {
+      "epoch": 2.73984,
+      "grad_norm": 0.5426006317138672,
+      "learning_rate": 2.2307915406218517e-07,
+      "loss": 0.2092,
+      "step": 1426
+    },
+    {
+      "epoch": 2.74176,
+      "grad_norm": 0.4473648965358734,
+      "learning_rate": 2.1978655803553128e-07,
+      "loss": 0.1415,
+      "step": 1427
+    },
+    {
+      "epoch": 2.74368,
+      "grad_norm": 0.529974639415741,
+      "learning_rate": 2.1651789586287442e-07,
+      "loss": 0.1949,
+      "step": 1428
+    },
+    {
+      "epoch": 2.7456,
+      "grad_norm": 0.4456387460231781,
+      "learning_rate": 2.1327318390994445e-07,
+      "loss": 0.1419,
+      "step": 1429
+    },
+    {
+      "epoch": 2.7475199999999997,
+      "grad_norm": 0.509550154209137,
+      "learning_rate": 2.1005243842255552e-07,
+      "loss": 0.1668,
+      "step": 1430
+    },
+    {
+      "epoch": 2.74944,
+      "grad_norm": 0.48361483216285706,
+      "learning_rate": 2.068556755265272e-07,
+      "loss": 0.2052,
+      "step": 1431
+    },
+    {
+      "epoch": 2.75136,
+      "grad_norm": 0.46098339557647705,
+      "learning_rate": 2.0368291122759898e-07,
+      "loss": 0.1583,
+      "step": 1432
+    },
+    {
+      "epoch": 2.75328,
+      "grad_norm": 0.544738233089447,
+      "learning_rate": 2.005341614113543e-07,
+      "loss": 0.1847,
+      "step": 1433
+    },
+    {
+      "epoch": 2.7552,
+      "grad_norm": 0.5398479700088501,
+      "learning_rate": 1.9740944184313882e-07,
+      "loss": 0.2009,
+      "step": 1434
+    },
+    {
+      "epoch": 2.75712,
+      "grad_norm": 0.4925403296947479,
+      "learning_rate": 1.9430876816798228e-07,
+      "loss": 0.1496,
+      "step": 1435
+    },
+    {
+      "epoch": 2.75904,
+      "grad_norm": 0.5223547220230103,
+      "learning_rate": 1.9123215591052014e-07,
+      "loss": 0.2015,
+      "step": 1436
+    },
+    {
+      "epoch": 2.76096,
+      "grad_norm": 0.5292021632194519,
+      "learning_rate": 1.88179620474917e-07,
+      "loss": 0.1937,
+      "step": 1437
+    },
+    {
+      "epoch": 2.76288,
+      "grad_norm": 0.47044602036476135,
+      "learning_rate": 1.8515117714478447e-07,
+      "loss": 0.179,
+      "step": 1438
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.4776745140552521,
+      "learning_rate": 1.8214684108311286e-07,
+      "loss": 0.1581,
+      "step": 1439
+    },
+    {
+      "epoch": 2.76672,
+      "grad_norm": 0.5121039152145386,
+      "learning_rate": 1.7916662733218848e-07,
+      "loss": 0.1594,
+      "step": 1440
+    },
+    {
+      "epoch": 2.76864,
+      "grad_norm": 0.47272300720214844,
+      "learning_rate": 1.762105508135198e-07,
+      "loss": 0.1455,
+      "step": 1441
+    },
+    {
+      "epoch": 2.77056,
+      "grad_norm": 0.47502782940864563,
+      "learning_rate": 1.7327862632776638e-07,
+      "loss": 0.1777,
+      "step": 1442
+    },
+    {
+      "epoch": 2.77248,
+      "grad_norm": 0.4969771206378937,
+      "learning_rate": 1.7037086855465902e-07,
+      "loss": 0.1908,
+      "step": 1443
+    },
+    {
+      "epoch": 2.7744,
+      "grad_norm": 0.47299256920814514,
+      "learning_rate": 1.6748729205293024e-07,
+      "loss": 0.1736,
+      "step": 1444
+    },
+    {
+      "epoch": 2.77632,
+      "grad_norm": 0.5601415038108826,
+      "learning_rate": 1.646279112602417e-07,
+      "loss": 0.2112,
+      "step": 1445
+    },
+    {
+      "epoch": 2.7782400000000003,
+      "grad_norm": 0.5285223722457886,
+      "learning_rate": 1.6179274049310966e-07,
+      "loss": 0.2041,
+      "step": 1446
+    },
+    {
+      "epoch": 2.78016,
+      "grad_norm": 0.508060097694397,
+      "learning_rate": 1.5898179394683244e-07,
+      "loss": 0.1544,
+      "step": 1447
+    },
+    {
+      "epoch": 2.78208,
+      "grad_norm": 0.5289275050163269,
+      "learning_rate": 1.5619508569542363e-07,
+      "loss": 0.1862,
+      "step": 1448
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 0.4936520457267761,
+      "learning_rate": 1.5343262969153781e-07,
+      "loss": 0.1592,
+      "step": 1449
+    },
+    {
+      "epoch": 2.78592,
+      "grad_norm": 0.45374196767807007,
+      "learning_rate": 1.5069443976640287e-07,
+      "loss": 0.1566,
+      "step": 1450
+    },
+    {
+      "epoch": 2.78784,
+      "grad_norm": 0.48903894424438477,
+      "learning_rate": 1.4798052962974874e-07,
+      "loss": 0.1682,
+      "step": 1451
+    },
+    {
+      "epoch": 2.7897600000000002,
+      "grad_norm": 0.48836973309516907,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.1686,
+      "step": 1452
+    },
+    {
+      "epoch": 2.79168,
+      "grad_norm": 0.47878265380859375,
+      "learning_rate": 1.4262560295290884e-07,
+      "loss": 0.1615,
+      "step": 1453
+    },
+    {
+      "epoch": 2.7936,
+      "grad_norm": 0.4313754737377167,
+      "learning_rate": 1.3998461322408563e-07,
+      "loss": 0.1451,
+      "step": 1454
+    },
+    {
+      "epoch": 2.79552,
+      "grad_norm": 0.4980270266532898,
+      "learning_rate": 1.3736795690633353e-07,
+      "loss": 0.1868,
+      "step": 1455
+    },
+    {
+      "epoch": 2.79744,
+      "grad_norm": 0.5261210203170776,
+      "learning_rate": 1.3477564710088097e-07,
+      "loss": 0.186,
+      "step": 1456
+    },
+    {
+      "epoch": 2.79936,
+      "grad_norm": 0.46927425265312195,
+      "learning_rate": 1.3220769678705724e-07,
+      "loss": 0.1611,
+      "step": 1457
+    },
+    {
+      "epoch": 2.80128,
+      "grad_norm": 0.5142197012901306,
+      "learning_rate": 1.2966411882222695e-07,
+      "loss": 0.1783,
+      "step": 1458
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.5471307039260864,
+      "learning_rate": 1.271449259417268e-07,
+      "loss": 0.1817,
+      "step": 1459
+    },
+    {
+      "epoch": 2.80512,
+      "grad_norm": 0.4760148823261261,
+      "learning_rate": 1.2465013075879884e-07,
+      "loss": 0.1606,
+      "step": 1460
+    },
+    {
+      "epoch": 2.8070399999999998,
+      "grad_norm": 0.5122244954109192,
+      "learning_rate": 1.2217974576453072e-07,
+      "loss": 0.2111,
+      "step": 1461
+    },
+    {
+      "epoch": 2.80896,
+      "grad_norm": 0.5303905010223389,
+      "learning_rate": 1.1973378332779229e-07,
+      "loss": 0.1972,
+      "step": 1462
+    },
+    {
+      "epoch": 2.81088,
+      "grad_norm": 0.461692214012146,
+      "learning_rate": 1.1731225569517113e-07,
+      "loss": 0.1656,
+      "step": 1463
+    },
+    {
+      "epoch": 2.8128,
+      "grad_norm": 0.4992302358150482,
+      "learning_rate": 1.1491517499091498e-07,
+      "loss": 0.1829,
+      "step": 1464
+    },
+    {
+      "epoch": 2.81472,
+      "grad_norm": 0.4717392921447754,
+      "learning_rate": 1.1254255321686836e-07,
+      "loss": 0.1816,
+      "step": 1465
+    },
+    {
+      "epoch": 2.81664,
+      "grad_norm": 0.5641149282455444,
+      "learning_rate": 1.1019440225241317e-07,
+      "loss": 0.1943,
+      "step": 1466
+    },
+    {
+      "epoch": 2.8185599999999997,
+      "grad_norm": 0.5237898230552673,
+      "learning_rate": 1.078707338544105e-07,
+      "loss": 0.1796,
+      "step": 1467
+    },
+    {
+      "epoch": 2.82048,
+      "grad_norm": 0.4788375496864319,
+      "learning_rate": 1.055715596571405e-07,
+      "loss": 0.1613,
+      "step": 1468
+    },
+    {
+      "epoch": 2.8224,
+      "grad_norm": 0.4782055616378784,
+      "learning_rate": 1.0329689117224262e-07,
+      "loss": 0.161,
+      "step": 1469
+    },
+    {
+      "epoch": 2.82432,
+      "grad_norm": 0.43120184540748596,
+      "learning_rate": 1.0104673978866164e-07,
+      "loss": 0.1619,
+      "step": 1470
+    },
+    {
+      "epoch": 2.82624,
+      "grad_norm": 0.48213693499565125,
+      "learning_rate": 9.882111677258777e-08,
+      "loss": 0.1578,
+      "step": 1471
+    },
+    {
+      "epoch": 2.82816,
+      "grad_norm": 0.49023351073265076,
+      "learning_rate": 9.662003326740166e-08,
+      "loss": 0.1954,
+      "step": 1472
+    },
+    {
+      "epoch": 2.83008,
+      "grad_norm": 0.5225573182106018,
+      "learning_rate": 9.444350029361671e-08,
+      "loss": 0.1966,
+      "step": 1473
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 0.4643978774547577,
+      "learning_rate": 9.22915287488274e-08,
+      "loss": 0.1552,
+      "step": 1474
+    },
+    {
+      "epoch": 2.83392,
+      "grad_norm": 0.47553491592407227,
+      "learning_rate": 9.016412940765107e-08,
+      "loss": 0.1694,
+      "step": 1475
+    },
+    {
+      "epoch": 2.83584,
+      "grad_norm": 0.5227478742599487,
+      "learning_rate": 8.80613129216762e-08,
+      "loss": 0.1703,
+      "step": 1476
+    },
+    {
+      "epoch": 2.83776,
+      "grad_norm": 0.5164349675178528,
+      "learning_rate": 8.598308981940751e-08,
+      "loss": 0.194,
+      "step": 1477
+    },
+    {
+      "epoch": 2.83968,
+      "grad_norm": 0.506781280040741,
+      "learning_rate": 8.392947050621603e-08,
+      "loss": 0.1582,
+      "step": 1478
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.4964149296283722,
+      "learning_rate": 8.190046526428241e-08,
+      "loss": 0.171,
+      "step": 1479
+    },
+    {
+      "epoch": 2.84352,
+      "grad_norm": 0.478400856256485,
+      "learning_rate": 7.989608425254924e-08,
+      "loss": 0.1786,
+      "step": 1480
+    },
+    {
+      "epoch": 2.84544,
+      "grad_norm": 0.5768463015556335,
+      "learning_rate": 7.791633750667105e-08,
+      "loss": 0.2625,
+      "step": 1481
+    },
+    {
+      "epoch": 2.84736,
+      "grad_norm": 0.4935254156589508,
+      "learning_rate": 7.59612349389599e-08,
+      "loss": 0.1805,
+      "step": 1482
+    },
+    {
+      "epoch": 2.8492800000000003,
+      "grad_norm": 0.46618959307670593,
+      "learning_rate": 7.403078633833716e-08,
+      "loss": 0.1691,
+      "step": 1483
+    },
+    {
+      "epoch": 2.8512,
+      "grad_norm": 0.49815329909324646,
+      "learning_rate": 7.212500137028789e-08,
+      "loss": 0.1578,
+      "step": 1484
+    },
+    {
+      "epoch": 2.85312,
+      "grad_norm": 0.5078113079071045,
+      "learning_rate": 7.024388957680705e-08,
+      "loss": 0.173,
+      "step": 1485
+    },
+    {
+      "epoch": 2.85504,
+      "grad_norm": 0.4840622842311859,
+      "learning_rate": 6.838746037635735e-08,
+      "loss": 0.1657,
+      "step": 1486
+    },
+    {
+      "epoch": 2.85696,
+      "grad_norm": 0.5306880474090576,
+      "learning_rate": 6.655572306381696e-08,
+      "loss": 0.2141,
+      "step": 1487
+    },
+    {
+      "epoch": 2.85888,
+      "grad_norm": 0.4794956147670746,
+      "learning_rate": 6.474868681043578e-08,
+      "loss": 0.165,
+      "step": 1488
+    },
+    {
+      "epoch": 2.8608000000000002,
+      "grad_norm": 0.4843374490737915,
+      "learning_rate": 6.29663606637898e-08,
+      "loss": 0.1544,
+      "step": 1489
+    },
+    {
+      "epoch": 2.86272,
+      "grad_norm": 0.4867093563079834,
+      "learning_rate": 6.120875354773459e-08,
+      "loss": 0.1625,
+      "step": 1490
+    },
+    {
+      "epoch": 2.86464,
+      "grad_norm": 0.4692665636539459,
+      "learning_rate": 5.947587426236079e-08,
+      "loss": 0.148,
+      "step": 1491
+    },
+    {
+      "epoch": 2.8665599999999998,
+      "grad_norm": 0.4676060974597931,
+      "learning_rate": 5.776773148394976e-08,
+      "loss": 0.1696,
+      "step": 1492
+    },
+    {
+      "epoch": 2.86848,
+      "grad_norm": 0.5762848258018494,
+      "learning_rate": 5.608433376493194e-08,
+      "loss": 0.2113,
+      "step": 1493
+    },
+    {
+      "epoch": 2.8704,
+      "grad_norm": 0.4478662312030792,
+      "learning_rate": 5.4425689533841864e-08,
+      "loss": 0.1344,
+      "step": 1494
+    },
+    {
+      "epoch": 2.87232,
+      "grad_norm": 0.5131320357322693,
+      "learning_rate": 5.279180709527765e-08,
+      "loss": 0.1629,
+      "step": 1495
+    },
+    {
+      "epoch": 2.87424,
+      "grad_norm": 0.5137205123901367,
+      "learning_rate": 5.1182694629857145e-08,
+      "loss": 0.1808,
+      "step": 1496
+    },
+    {
+      "epoch": 2.87616,
+      "grad_norm": 0.5199742913246155,
+      "learning_rate": 4.959836019417963e-08,
+      "loss": 0.1621,
+      "step": 1497
+    },
+    {
+      "epoch": 2.8780799999999997,
+      "grad_norm": 0.529670774936676,
+      "learning_rate": 4.803881172078473e-08,
+      "loss": 0.202,
+      "step": 1498
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.4574511647224426,
+      "learning_rate": 4.650405701811078e-08,
+      "loss": 0.1601,
+      "step": 1499
+    },
+    {
+      "epoch": 2.88192,
+      "grad_norm": 0.4847154915332794,
+      "learning_rate": 4.499410377045765e-08,
+      "loss": 0.1836,
+      "step": 1500
+    },
+    {
+      "epoch": 2.88384,
+      "grad_norm": 0.49493154883384705,
+      "learning_rate": 4.350895953794898e-08,
+      "loss": 0.1639,
+      "step": 1501
+    },
+    {
+      "epoch": 2.88576,
+      "grad_norm": 0.48846614360809326,
+      "learning_rate": 4.2048631756492206e-08,
+      "loss": 0.1742,
+      "step": 1502
+    },
+    {
+      "epoch": 2.88768,
+      "grad_norm": 0.4900778532028198,
+      "learning_rate": 4.0613127737741396e-08,
+      "loss": 0.1528,
+      "step": 1503
+    },
+    {
+      "epoch": 2.8895999999999997,
+      "grad_norm": 0.4664475917816162,
+      "learning_rate": 3.9202454669063915e-08,
+      "loss": 0.1715,
+      "step": 1504
+    },
+    {
+      "epoch": 2.89152,
+      "grad_norm": 0.48522326350212097,
+      "learning_rate": 3.781661961349992e-08,
+      "loss": 0.1609,
+      "step": 1505
+    },
+    {
+      "epoch": 2.89344,
+      "grad_norm": 0.4957093894481659,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 0.1722,
+      "step": 1506
+    },
+    {
+      "epoch": 2.89536,
+      "grad_norm": 0.4857979118824005,
+      "learning_rate": 3.5119491172039836e-08,
+      "loss": 0.1864,
+      "step": 1507
+    },
+    {
+      "epoch": 2.89728,
+      "grad_norm": 0.4926449656486511,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 0.1743,
+      "step": 1508
+    },
+    {
+      "epoch": 2.8992,
+      "grad_norm": 0.4623541533946991,
+      "learning_rate": 3.252179642985909e-08,
+      "loss": 0.1713,
+      "step": 1509
+    },
+    {
+      "epoch": 2.90112,
+      "grad_norm": 0.49395453929901123,
+      "learning_rate": 3.1260253031660247e-08,
+      "loss": 0.1795,
+      "step": 1510
+    },
+    {
+      "epoch": 2.90304,
+      "grad_norm": 0.4932856261730194,
+      "learning_rate": 3.002358741205691e-08,
+      "loss": 0.1964,
+      "step": 1511
+    },
+    {
+      "epoch": 2.90496,
+      "grad_norm": 0.5978785157203674,
+      "learning_rate": 2.8811805762860578e-08,
+      "loss": 0.2168,
+      "step": 1512
+    },
+    {
+      "epoch": 2.90688,
+      "grad_norm": 0.4652256965637207,
+      "learning_rate": 2.762491415128965e-08,
+      "loss": 0.1595,
+      "step": 1513
+    },
+    {
+      "epoch": 2.9088000000000003,
+      "grad_norm": 0.49602875113487244,
+      "learning_rate": 2.6462918519944425e-08,
+      "loss": 0.1905,
+      "step": 1514
+    },
+    {
+      "epoch": 2.91072,
+      "grad_norm": 0.4354414939880371,
+      "learning_rate": 2.5325824686772138e-08,
+      "loss": 0.1589,
+      "step": 1515
+    },
+    {
+      "epoch": 2.91264,
+      "grad_norm": 0.479852557182312,
+      "learning_rate": 2.4213638345040868e-08,
+      "loss": 0.1667,
+      "step": 1516
+    },
+    {
+      "epoch": 2.91456,
+      "grad_norm": 0.4678557813167572,
+      "learning_rate": 2.3126365063311228e-08,
+      "loss": 0.1533,
+      "step": 1517
+    },
+    {
+      "epoch": 2.91648,
+      "grad_norm": 0.4605678617954254,
+      "learning_rate": 2.206401028540639e-08,
+      "loss": 0.1731,
+      "step": 1518
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.49699097871780396,
+      "learning_rate": 2.1026579330387655e-08,
+      "loss": 0.172,
+      "step": 1519
+    },
+    {
+      "epoch": 2.9203200000000002,
+      "grad_norm": 0.5154688358306885,
+      "learning_rate": 2.0014077392525035e-08,
+      "loss": 0.1827,
+      "step": 1520
+    },
+    {
+      "epoch": 2.92224,
+      "grad_norm": 0.4762783944606781,
+      "learning_rate": 1.9026509541272276e-08,
+      "loss": 0.1848,
+      "step": 1521
+    },
+    {
+      "epoch": 2.92416,
+      "grad_norm": 0.5206320285797119,
+      "learning_rate": 1.8063880721242986e-08,
+      "loss": 0.1876,
+      "step": 1522
+    },
+    {
+      "epoch": 2.92608,
+      "grad_norm": 0.4943743646144867,
+      "learning_rate": 1.712619575218344e-08,
+      "loss": 0.1785,
+      "step": 1523
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 0.4939040541648865,
+      "learning_rate": 1.6213459328950355e-08,
+      "loss": 0.1479,
+      "step": 1524
+    },
+    {
+      "epoch": 2.92992,
+      "grad_norm": 0.47102558612823486,
+      "learning_rate": 1.5325676021484826e-08,
+      "loss": 0.1645,
+      "step": 1525
+    },
+    {
+      "epoch": 2.9318400000000002,
+      "grad_norm": 0.49551957845687866,
+      "learning_rate": 1.4462850274794548e-08,
+      "loss": 0.1678,
+      "step": 1526
+    },
+    {
+      "epoch": 2.93376,
+      "grad_norm": 0.5450107455253601,
+      "learning_rate": 1.3624986408924956e-08,
+      "loss": 0.2245,
+      "step": 1527
+    },
+    {
+      "epoch": 2.93568,
+      "grad_norm": 0.45947033166885376,
+      "learning_rate": 1.2812088618942009e-08,
+      "loss": 0.1749,
+      "step": 1528
+    },
+    {
+      "epoch": 2.9375999999999998,
+      "grad_norm": 0.4959351718425751,
+      "learning_rate": 1.2024160974911103e-08,
+      "loss": 0.1705,
+      "step": 1529
+    },
+    {
+      "epoch": 2.93952,
+      "grad_norm": 0.48875778913497925,
+      "learning_rate": 1.1261207421874309e-08,
+      "loss": 0.1539,
+      "step": 1530
+    },
+    {
+      "epoch": 2.94144,
+      "grad_norm": 0.5686318874359131,
+      "learning_rate": 1.0523231779832055e-08,
+      "loss": 0.2185,
+      "step": 1531
+    },
+    {
+      "epoch": 2.94336,
+      "grad_norm": 0.5016399025917053,
+      "learning_rate": 9.810237743724805e-09,
+      "loss": 0.1676,
+      "step": 1532
+    },
+    {
+      "epoch": 2.94528,
+      "grad_norm": 0.5080051422119141,
+      "learning_rate": 9.12222888341252e-09,
+      "loss": 0.1926,
+      "step": 1533
+    },
+    {
+      "epoch": 2.9472,
+      "grad_norm": 0.5564820170402527,
+      "learning_rate": 8.459208643659122e-09,
+      "loss": 0.1922,
+      "step": 1534
+    },
+    {
+      "epoch": 2.9491199999999997,
+      "grad_norm": 0.4664892554283142,
+      "learning_rate": 7.8211803441125e-09,
+      "loss": 0.1648,
+      "step": 1535
+    },
+    {
+      "epoch": 2.95104,
+      "grad_norm": 0.43937167525291443,
+      "learning_rate": 7.2081471792911914e-09,
+      "loss": 0.1484,
+      "step": 1536
+    },
+    {
+      "epoch": 2.95296,
+      "grad_norm": 0.4819045662879944,
+      "learning_rate": 6.6201122185649555e-09,
+      "loss": 0.1758,
+      "step": 1537
+    },
+    {
+      "epoch": 2.95488,
+      "grad_norm": 0.48437193036079407,
+      "learning_rate": 6.057078406142003e-09,
+      "loss": 0.168,
+      "step": 1538
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.4970892369747162,
+      "learning_rate": 5.519048561053453e-09,
+      "loss": 0.1786,
+      "step": 1539
+    },
+    {
+      "epoch": 2.95872,
+      "grad_norm": 0.48164060711860657,
+      "learning_rate": 5.006025377138901e-09,
+      "loss": 0.1719,
+      "step": 1540
+    },
+    {
+      "epoch": 2.96064,
+      "grad_norm": 0.5043368339538574,
+      "learning_rate": 4.518011423032542e-09,
+      "loss": 0.17,
+      "step": 1541
+    },
+    {
+      "epoch": 2.96256,
+      "grad_norm": 0.5187457799911499,
+      "learning_rate": 4.055009142152066e-09,
+      "loss": 0.164,
+      "step": 1542
+    },
+    {
+      "epoch": 2.96448,
+      "grad_norm": 0.48123857378959656,
+      "learning_rate": 3.6170208526836724e-09,
+      "loss": 0.1698,
+      "step": 1543
+    },
+    {
+      "epoch": 2.9664,
+      "grad_norm": 0.46527099609375,
+      "learning_rate": 3.204048747573185e-09,
+      "loss": 0.1625,
+      "step": 1544
+    },
+    {
+      "epoch": 2.96832,
+      "grad_norm": 0.48876380920410156,
+      "learning_rate": 2.816094894513843e-09,
+      "loss": 0.1712,
+      "step": 1545
+    },
+    {
+      "epoch": 2.97024,
+      "grad_norm": 0.5515295267105103,
+      "learning_rate": 2.4531612359363077e-09,
+      "loss": 0.1783,
+      "step": 1546
+    },
+    {
+      "epoch": 2.97216,
+      "grad_norm": 0.48005878925323486,
+      "learning_rate": 2.1152495889970035e-09,
+      "loss": 0.1703,
+      "step": 1547
+    },
+    {
+      "epoch": 2.97408,
+      "grad_norm": 0.4595643877983093,
+      "learning_rate": 1.8023616455731253e-09,
+      "loss": 0.1651,
+      "step": 1548
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 0.4874194264411926,
+      "learning_rate": 1.514498972249312e-09,
+      "loss": 0.1697,
+      "step": 1549
+    },
+    {
+      "epoch": 2.97792,
+      "grad_norm": 0.49300575256347656,
+      "learning_rate": 1.2516630103137638e-09,
+      "loss": 0.1932,
+      "step": 1550
+    },
+    {
+      "epoch": 2.9798400000000003,
+      "grad_norm": 0.4463099539279938,
+      "learning_rate": 1.0138550757493592e-09,
+      "loss": 0.1723,
+      "step": 1551
+    },
+    {
+      "epoch": 2.98176,
+      "grad_norm": 0.49644947052001953,
+      "learning_rate": 8.010763592264381e-10,
+      "loss": 0.1621,
+      "step": 1552
+    },
+    {
+      "epoch": 2.98368,
+      "grad_norm": 0.5428080558776855,
+      "learning_rate": 6.133279260983616e-10,
+      "loss": 0.2144,
+      "step": 1553
+    },
+    {
+      "epoch": 2.9856,
+      "grad_norm": 0.4774559736251831,
+      "learning_rate": 4.506107163948503e-10,
+      "loss": 0.1725,
+      "step": 1554
+    },
+    {
+      "epoch": 2.98752,
+      "grad_norm": 0.4530414044857025,
+      "learning_rate": 3.12925544818099e-10,
+      "loss": 0.1489,
+      "step": 1555
+    },
+    {
+      "epoch": 2.98944,
+      "grad_norm": 0.5247970819473267,
+      "learning_rate": 2.0027310073833516e-10,
+      "loss": 0.1884,
+      "step": 1556
+    },
+    {
+      "epoch": 2.9913600000000002,
+      "grad_norm": 0.459453821182251,
+      "learning_rate": 1.1265394818993358e-10,
+      "loss": 0.1519,
+      "step": 1557
+    },
+    {
+      "epoch": 2.99328,
+      "grad_norm": 0.46953755617141724,
+      "learning_rate": 5.0068525870305974e-11,
+      "loss": 0.1611,
+      "step": 1558
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.48540616035461426,
+      "learning_rate": 1.251714713546015e-11,
+      "loss": 0.1731,
+      "step": 1559
+    },
+    {
+      "epoch": 2.99712,
+      "grad_norm": 0.502263605594635,
+      "learning_rate": 0.0,
+      "loss": 0.1588,
+      "step": 1560
+    },
+    {
+      "epoch": 2.99712,
+      "step": 1560,
+      "total_flos": 153031453360128.0,
+      "train_loss": 0.21554167224810675,
+      "train_runtime": 14482.7308,
+      "train_samples_per_second": 10.357,
+      "train_steps_per_second": 0.108
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1560,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 153031453360128.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}