diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.2742857142857143,
+  "eval_steps": 500,
+  "global_step": 600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.563270330429077,
+      "learning_rate": 4e-05,
+      "loss": 2.4127,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.333299160003662,
+      "learning_rate": 8e-05,
+      "loss": 2.3102,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8613317012786865,
+      "learning_rate": 0.00012,
+      "loss": 2.1983,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.620126485824585,
+      "learning_rate": 0.00016,
+      "loss": 2.0917,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0352119207382202,
+      "learning_rate": 0.0002,
+      "loss": 1.9613,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8601917624473572,
+      "learning_rate": 0.00019979899497487438,
+      "loss": 1.6927,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.273868203163147,
+      "learning_rate": 0.00019959798994974876,
+      "loss": 1.6828,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0873847007751465,
+      "learning_rate": 0.00019939698492462313,
+      "loss": 1.5088,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5800795555114746,
+      "learning_rate": 0.0001991959798994975,
+      "loss": 1.3702,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.693160355091095,
+      "learning_rate": 0.00019899497487437187,
+      "loss": 1.3718,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.49409618973731995,
+      "learning_rate": 0.00019879396984924622,
+      "loss": 1.3583,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.47029319405555725,
+      "learning_rate": 0.00019859296482412062,
+      "loss": 1.2791,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.43618088960647583,
+      "learning_rate": 0.000198391959798995,
+      "loss": 1.3161,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3907912075519562,
+      "learning_rate": 0.00019819095477386937,
+      "loss": 1.2954,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6292415857315063,
+      "learning_rate": 0.0001979899497487437,
+      "loss": 1.3397,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.37423521280288696,
+      "learning_rate": 0.0001977889447236181,
+      "loss": 1.3983,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3845643699169159,
+      "learning_rate": 0.00019758793969849249,
+      "loss": 1.3349,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3657298982143402,
+      "learning_rate": 0.00019738693467336683,
+      "loss": 1.2767,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3727971315383911,
+      "learning_rate": 0.0001971859296482412,
+      "loss": 1.3672,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.35123032331466675,
+      "learning_rate": 0.0001969849246231156,
+      "loss": 1.3692,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4003850221633911,
+      "learning_rate": 0.00019678391959798995,
+      "loss": 1.3412,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3638221323490143,
+      "learning_rate": 0.00019658291457286432,
+      "loss": 1.2813,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.391216903924942,
+      "learning_rate": 0.0001963819095477387,
+      "loss": 1.2853,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4370620846748352,
+      "learning_rate": 0.0001961809045226131,
+      "loss": 1.2524,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3566085696220398,
+      "learning_rate": 0.00019597989949748744,
+      "loss": 1.3192,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.35438084602355957,
+      "learning_rate": 0.00019577889447236181,
+      "loss": 1.2858,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3968108296394348,
+      "learning_rate": 0.0001955778894472362,
+      "loss": 1.3112,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.36512017250061035,
+      "learning_rate": 0.00019537688442211056,
+      "loss": 1.278,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3982504606246948,
+      "learning_rate": 0.00019517587939698493,
+      "loss": 1.2392,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.38377949595451355,
+      "learning_rate": 0.0001949748743718593,
+      "loss": 1.2843,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3582867980003357,
+      "learning_rate": 0.00019477386934673368,
+      "loss": 1.3008,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3572194576263428,
+      "learning_rate": 0.00019457286432160805,
+      "loss": 1.294,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.35502907633781433,
+      "learning_rate": 0.00019437185929648243,
+      "loss": 1.3877,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3649040460586548,
+      "learning_rate": 0.0001941708542713568,
+      "loss": 1.2966,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3649256229400635,
+      "learning_rate": 0.00019396984924623117,
+      "loss": 1.2354,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.36085084080696106,
+      "learning_rate": 0.00019376884422110552,
+      "loss": 1.2409,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.35929059982299805,
+      "learning_rate": 0.00019356783919597992,
+      "loss": 1.243,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3897881805896759,
+      "learning_rate": 0.0001933668341708543,
+      "loss": 1.3945,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.35484543442726135,
+      "learning_rate": 0.00019316582914572864,
+      "loss": 1.3433,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.35691192746162415,
+      "learning_rate": 0.000192964824120603,
+      "loss": 1.3243,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3804129958152771,
+      "learning_rate": 0.0001927638190954774,
+      "loss": 1.2509,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3623339831829071,
+      "learning_rate": 0.00019256281407035178,
+      "loss": 1.1799,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3411855697631836,
+      "learning_rate": 0.00019236180904522613,
+      "loss": 1.2372,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.36590930819511414,
+      "learning_rate": 0.0001921608040201005,
+      "loss": 1.2585,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30974116921424866,
+      "learning_rate": 0.0001919597989949749,
+      "loss": 1.2974,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.32794803380966187,
+      "learning_rate": 0.00019175879396984925,
+      "loss": 1.2696,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33263906836509705,
+      "learning_rate": 0.00019155778894472362,
+      "loss": 1.3209,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31748828291893005,
+      "learning_rate": 0.000191356783919598,
+      "loss": 1.278,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.34738561511039734,
+      "learning_rate": 0.0001911557788944724,
+      "loss": 1.2105,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3313944339752197,
+      "learning_rate": 0.00019095477386934674,
+      "loss": 1.2527,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33137476444244385,
+      "learning_rate": 0.0001907537688442211,
+      "loss": 1.2984,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31752490997314453,
+      "learning_rate": 0.00019055276381909548,
+      "loss": 1.307,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3111082911491394,
+      "learning_rate": 0.00019035175879396986,
+      "loss": 1.2769,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31065696477890015,
+      "learning_rate": 0.00019015075376884423,
+      "loss": 1.3082,
+      "step": 54
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3382773697376251,
+      "learning_rate": 0.0001899497487437186,
+      "loss": 1.2744,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.34320947527885437,
+      "learning_rate": 0.00018974874371859298,
+      "loss": 1.3013,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.33131280541419983,
+      "learning_rate": 0.00018954773869346732,
+      "loss": 1.4066,
+      "step": 57
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3357389569282532,
+      "learning_rate": 0.00018934673366834172,
+      "loss": 1.2841,
+      "step": 58
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3200838267803192,
+      "learning_rate": 0.0001891457286432161,
+      "loss": 1.2654,
+      "step": 59
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3336584270000458,
+      "learning_rate": 0.00018894472361809047,
+      "loss": 1.1716,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3128441274166107,
+      "learning_rate": 0.00018874371859296481,
+      "loss": 1.3009,
+      "step": 61
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30249112844467163,
+      "learning_rate": 0.00018854271356783921,
+      "loss": 1.2311,
+      "step": 62
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3263241946697235,
+      "learning_rate": 0.0001883417085427136,
+      "loss": 1.2344,
+      "step": 63
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.32348358631134033,
+      "learning_rate": 0.00018814070351758793,
+      "loss": 1.3023,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6508419513702393,
+      "learning_rate": 0.0001879396984924623,
+      "loss": 1.2028,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.34560996294021606,
+      "learning_rate": 0.0001877386934673367,
+      "loss": 1.389,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36555996537208557,
+      "learning_rate": 0.00018753768844221108,
+      "loss": 1.3653,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3195466697216034,
+      "learning_rate": 0.00018733668341708543,
+      "loss": 1.2412,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30555933713912964,
+      "learning_rate": 0.0001871356783919598,
+      "loss": 1.2357,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30776411294937134,
+      "learning_rate": 0.0001869346733668342,
+      "loss": 1.3112,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.31933915615081787,
+      "learning_rate": 0.00018673366834170854,
+      "loss": 1.1951,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3241545259952545,
+      "learning_rate": 0.00018653266331658292,
+      "loss": 1.2717,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3117482364177704,
+      "learning_rate": 0.0001863316582914573,
+      "loss": 1.3031,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.33056551218032837,
+      "learning_rate": 0.0001861306532663317,
+      "loss": 1.2098,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.32441195845603943,
+      "learning_rate": 0.00018592964824120604,
+      "loss": 1.2135,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.34216779470443726,
+      "learning_rate": 0.0001857286432160804,
+      "loss": 1.2531,
+      "step": 76
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.32885247468948364,
+      "learning_rate": 0.00018552763819095478,
+      "loss": 1.3054,
+      "step": 77
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34541794657707214,
+      "learning_rate": 0.00018532663316582915,
+      "loss": 1.3207,
+      "step": 78
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30329057574272156,
+      "learning_rate": 0.00018512562814070353,
+      "loss": 1.2652,
+      "step": 79
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31469491124153137,
+      "learning_rate": 0.0001849246231155779,
+      "loss": 1.1961,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3181230127811432,
+      "learning_rate": 0.00018472361809045227,
+      "loss": 1.3111,
+      "step": 81
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3181725740432739,
+      "learning_rate": 0.00018452261306532662,
+      "loss": 1.3353,
+      "step": 82
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3154084384441376,
+      "learning_rate": 0.00018432160804020102,
+      "loss": 1.2418,
+      "step": 83
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.35061103105545044,
+      "learning_rate": 0.0001841206030150754,
+      "loss": 1.2332,
+      "step": 84
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3259966969490051,
+      "learning_rate": 0.00018391959798994977,
+      "loss": 1.3633,
+      "step": 85
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31192857027053833,
+      "learning_rate": 0.0001837185929648241,
+      "loss": 1.1886,
+      "step": 86
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.32024237513542175,
+      "learning_rate": 0.0001835175879396985,
+      "loss": 1.2141,
+      "step": 87
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.302498459815979,
+      "learning_rate": 0.00018331658291457288,
+      "loss": 1.237,
+      "step": 88
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3569789230823517,
+      "learning_rate": 0.00018311557788944723,
+      "loss": 1.3015,
+      "step": 89
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3121156692504883,
+      "learning_rate": 0.0001829145728643216,
+      "loss": 1.281,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31279826164245605,
+      "learning_rate": 0.000182713567839196,
+      "loss": 1.2924,
+      "step": 91
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3210877478122711,
+      "learning_rate": 0.00018251256281407038,
+      "loss": 1.3082,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.331406831741333,
+      "learning_rate": 0.00018231155778894472,
+      "loss": 1.2434,
+      "step": 93
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3135213255882263,
+      "learning_rate": 0.0001821105527638191,
+      "loss": 1.2188,
+      "step": 94
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31146401166915894,
+      "learning_rate": 0.0001819095477386935,
+      "loss": 1.2484,
+      "step": 95
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.32071712613105774,
+      "learning_rate": 0.00018170854271356784,
+      "loss": 1.1927,
+      "step": 96
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3343571722507477,
+      "learning_rate": 0.00018150753768844221,
+      "loss": 1.3443,
+      "step": 97
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3510550558567047,
+      "learning_rate": 0.0001813065326633166,
+      "loss": 1.2832,
+      "step": 98
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33436939120292664,
+      "learning_rate": 0.00018110552763819096,
+      "loss": 1.252,
+      "step": 99
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3175451159477234,
+      "learning_rate": 0.00018090452261306533,
+      "loss": 1.251,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32603979110717773,
+      "learning_rate": 0.0001807035175879397,
+      "loss": 1.228,
+      "step": 101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3073003590106964,
+      "learning_rate": 0.00018050251256281408,
+      "loss": 1.2659,
+      "step": 102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3285619616508484,
+      "learning_rate": 0.00018030150753768845,
+      "loss": 1.2826,
+      "step": 103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3038572072982788,
+      "learning_rate": 0.00018010050251256282,
+      "loss": 1.217,
+      "step": 104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35778746008872986,
+      "learning_rate": 0.0001798994974874372,
+      "loss": 1.2901,
+      "step": 105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2900612950325012,
+      "learning_rate": 0.00017969849246231157,
+      "loss": 1.2651,
+      "step": 106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32928743958473206,
+      "learning_rate": 0.00017949748743718592,
+      "loss": 1.3143,
+      "step": 107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32471874356269836,
+      "learning_rate": 0.00017929648241206032,
+      "loss": 1.1834,
+      "step": 108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30989256501197815,
+      "learning_rate": 0.0001790954773869347,
+      "loss": 1.2216,
+      "step": 109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3371771275997162,
+      "learning_rate": 0.00017889447236180906,
+      "loss": 1.197,
+      "step": 110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31041428446769714,
+      "learning_rate": 0.0001786934673366834,
+      "loss": 1.27,
+      "step": 111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3152185082435608,
+      "learning_rate": 0.0001784924623115578,
+      "loss": 1.2436,
+      "step": 112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3227459490299225,
+      "learning_rate": 0.00017829145728643218,
+      "loss": 1.2401,
+      "step": 113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3246959149837494,
+      "learning_rate": 0.00017809045226130653,
+      "loss": 1.2703,
+      "step": 114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38032859563827515,
+      "learning_rate": 0.0001778894472361809,
+      "loss": 1.3266,
+      "step": 115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33325478434562683,
+      "learning_rate": 0.0001776884422110553,
+      "loss": 1.2954,
+      "step": 116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3178690969944,
+      "learning_rate": 0.00017748743718592967,
+      "loss": 1.1793,
+      "step": 117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31393784284591675,
+      "learning_rate": 0.00017728643216080402,
+      "loss": 1.277,
+      "step": 118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3150279223918915,
+      "learning_rate": 0.0001770854271356784,
+      "loss": 1.293,
+      "step": 119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32476913928985596,
+      "learning_rate": 0.0001768844221105528,
+      "loss": 1.2569,
+      "step": 120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36075925827026367,
+      "learning_rate": 0.00017668341708542714,
+      "loss": 1.205,
+      "step": 121
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33134496212005615,
+      "learning_rate": 0.0001764824120603015,
+      "loss": 1.2299,
+      "step": 122
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30507662892341614,
+      "learning_rate": 0.00017628140703517588,
+      "loss": 1.2883,
+      "step": 123
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34049952030181885,
+      "learning_rate": 0.00017608040201005026,
+      "loss": 1.214,
+      "step": 124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3405919373035431,
+      "learning_rate": 0.00017587939698492463,
+      "loss": 1.2738,
+      "step": 125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3306083679199219,
+      "learning_rate": 0.000175678391959799,
+      "loss": 1.2415,
+      "step": 126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33770737051963806,
+      "learning_rate": 0.00017547738693467338,
+      "loss": 1.3233,
+      "step": 127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3261878788471222,
+      "learning_rate": 0.00017527638190954775,
+      "loss": 1.2695,
+      "step": 128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3433193266391754,
+      "learning_rate": 0.00017507537688442212,
+      "loss": 1.2052,
+      "step": 129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3111405670642853,
+      "learning_rate": 0.0001748743718592965,
+      "loss": 1.2798,
+      "step": 130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3630310297012329,
+      "learning_rate": 0.00017467336683417087,
+      "loss": 1.2567,
+      "step": 131
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31963038444519043,
+      "learning_rate": 0.00017447236180904521,
+      "loss": 1.2455,
+      "step": 132
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.299695760011673,
+      "learning_rate": 0.00017427135678391961,
+      "loss": 1.207,
+      "step": 133
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3167514503002167,
+      "learning_rate": 0.000174070351758794,
+      "loss": 1.2378,
+      "step": 134
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31375688314437866,
+      "learning_rate": 0.00017386934673366836,
+      "loss": 1.2658,
+      "step": 135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34311383962631226,
+      "learning_rate": 0.0001736683417085427,
+      "loss": 1.2004,
+      "step": 136
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31706517934799194,
+      "learning_rate": 0.0001734673366834171,
+      "loss": 1.1879,
+      "step": 137
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31296172738075256,
+      "learning_rate": 0.00017326633165829148,
+      "loss": 1.1866,
+      "step": 138
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3254072368144989,
+      "learning_rate": 0.00017306532663316582,
+      "loss": 1.1952,
+      "step": 139
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3165453374385834,
+      "learning_rate": 0.0001728643216080402,
+      "loss": 1.3459,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.35455992817878723,
+      "learning_rate": 0.0001726633165829146,
+      "loss": 1.2494,
+      "step": 141
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3116908073425293,
+      "learning_rate": 0.00017246231155778897,
+      "loss": 1.2225,
+      "step": 142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3141638934612274,
+      "learning_rate": 0.00017226130653266332,
+      "loss": 1.3385,
+      "step": 143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3096507787704468,
+      "learning_rate": 0.0001720603015075377,
+      "loss": 1.3257,
+      "step": 144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3160630464553833,
+      "learning_rate": 0.00017185929648241206,
+      "loss": 1.2683,
+      "step": 145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3342824876308441,
+      "learning_rate": 0.00017165829145728644,
+      "loss": 1.2551,
+      "step": 146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3086145222187042,
+      "learning_rate": 0.0001714572864321608,
+      "loss": 1.2456,
+      "step": 147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3001709282398224,
+      "learning_rate": 0.00017125628140703518,
+      "loss": 1.2287,
+      "step": 148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3277103304862976,
+      "learning_rate": 0.00017105527638190955,
+      "loss": 1.3302,
+      "step": 149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33616161346435547,
+      "learning_rate": 0.00017085427135678393,
+      "loss": 1.2604,
+      "step": 150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3231915533542633,
+      "learning_rate": 0.0001706532663316583,
+      "loss": 1.2367,
+      "step": 151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3305569291114807,
+      "learning_rate": 0.00017045226130653267,
+      "loss": 1.2387,
+      "step": 152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.35031118988990784,
+      "learning_rate": 0.00017025125628140705,
+      "loss": 1.2464,
+      "step": 153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3142334222793579,
+      "learning_rate": 0.00017005025125628142,
+      "loss": 1.3614,
+      "step": 154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31159430742263794,
+      "learning_rate": 0.0001698492462311558,
+      "loss": 1.2556,
+      "step": 155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3273050785064697,
+      "learning_rate": 0.00016964824120603016,
+      "loss": 1.3519,
+      "step": 156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3299296796321869,
+      "learning_rate": 0.0001694472361809045,
+      "loss": 1.1763,
+      "step": 157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33138513565063477,
+      "learning_rate": 0.0001692462311557789,
+      "loss": 1.17,
+      "step": 158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30424776673316956,
+      "learning_rate": 0.00016904522613065328,
+      "loss": 1.123,
+      "step": 159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3452983498573303,
+      "learning_rate": 0.00016884422110552766,
+      "loss": 1.2999,
+      "step": 160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33614206314086914,
+      "learning_rate": 0.000168643216080402,
+      "loss": 1.262,
+      "step": 161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32416558265686035,
+      "learning_rate": 0.0001684422110552764,
+      "loss": 1.2514,
+      "step": 162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29827457666397095,
+      "learning_rate": 0.00016824120603015078,
+      "loss": 1.2461,
+      "step": 163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32572871446609497,
+      "learning_rate": 0.00016804020100502512,
+      "loss": 1.2393,
+      "step": 164
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32171282172203064,
+      "learning_rate": 0.0001678391959798995,
+      "loss": 1.3045,
+      "step": 165
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34592801332473755,
+      "learning_rate": 0.0001676381909547739,
+      "loss": 1.2669,
+      "step": 166
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33795440196990967,
+      "learning_rate": 0.00016743718592964827,
+      "loss": 1.1404,
+      "step": 167
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32598641514778137,
+      "learning_rate": 0.0001672361809045226,
+      "loss": 1.2495,
+      "step": 168
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31816181540489197,
+      "learning_rate": 0.00016703517587939699,
+      "loss": 1.3003,
+      "step": 169
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3340943157672882,
+      "learning_rate": 0.00016683417085427136,
+      "loss": 1.2615,
+      "step": 170
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3242477476596832,
+      "learning_rate": 0.00016663316582914573,
+      "loss": 1.2527,
+      "step": 171
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.308652400970459,
+      "learning_rate": 0.0001664321608040201,
+      "loss": 1.3241,
+      "step": 172
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31818273663520813,
+      "learning_rate": 0.00016623115577889448,
+      "loss": 1.3712,
+      "step": 173
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32885751128196716,
+      "learning_rate": 0.00016603015075376885,
+      "loss": 1.2583,
+      "step": 174
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32561740279197693,
+      "learning_rate": 0.00016582914572864322,
+      "loss": 1.2458,
+      "step": 175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3278496563434601,
+      "learning_rate": 0.0001656281407035176,
+      "loss": 1.2205,
+      "step": 176
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32530438899993896,
+      "learning_rate": 0.00016542713567839197,
+      "loss": 1.2235,
+      "step": 177
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31232836842536926,
+      "learning_rate": 0.00016522613065326634,
+      "loss": 1.199,
+      "step": 178
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3209743797779083,
+      "learning_rate": 0.00016502512562814072,
+      "loss": 1.2717,
+      "step": 179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.329940527677536,
+      "learning_rate": 0.0001648241206030151,
+      "loss": 1.2425,
+      "step": 180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3144824802875519,
+      "learning_rate": 0.00016462311557788946,
+      "loss": 1.2444,
+      "step": 181
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3218553066253662,
+      "learning_rate": 0.0001644221105527638,
+      "loss": 1.2815,
+      "step": 182
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33460506796836853,
+      "learning_rate": 0.0001642211055276382,
+      "loss": 1.3774,
+      "step": 183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3300727605819702,
+      "learning_rate": 0.00016402010050251258,
+      "loss": 1.3436,
+      "step": 184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3530360460281372,
+      "learning_rate": 0.00016381909547738695,
+      "loss": 1.2605,
+      "step": 185
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3326485753059387,
+      "learning_rate": 0.0001636180904522613,
+      "loss": 1.2202,
+      "step": 186
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31355732679367065,
+      "learning_rate": 0.0001634170854271357,
+      "loss": 1.2798,
+      "step": 187
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3162304759025574,
+      "learning_rate": 0.00016321608040201007,
+      "loss": 1.2118,
+      "step": 188
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32264095544815063,
+      "learning_rate": 0.00016301507537688442,
+      "loss": 1.2775,
+      "step": 189
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.30425918102264404,
+      "learning_rate": 0.0001628140703517588,
+      "loss": 1.1438,
+      "step": 190
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33907556533813477,
+      "learning_rate": 0.00016261306532663316,
+      "loss": 1.4077,
+      "step": 191
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32334232330322266,
+      "learning_rate": 0.00016241206030150756,
+      "loss": 1.2673,
+      "step": 192
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32999834418296814,
+      "learning_rate": 0.0001622110552763819,
+      "loss": 1.257,
+      "step": 193
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3223746120929718,
+      "learning_rate": 0.00016201005025125628,
+      "loss": 1.2125,
+      "step": 194
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3236989378929138,
+      "learning_rate": 0.00016180904522613066,
+      "loss": 1.3645,
+      "step": 195
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3303336203098297,
+      "learning_rate": 0.00016160804020100503,
+      "loss": 1.2786,
+      "step": 196
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3135005831718445,
+      "learning_rate": 0.0001614070351758794,
+      "loss": 1.2775,
+      "step": 197
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3185466527938843,
+      "learning_rate": 0.00016120603015075378,
+      "loss": 1.2128,
+      "step": 198
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3355714976787567,
+      "learning_rate": 0.00016100502512562815,
+      "loss": 1.307,
+      "step": 199
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.339216023683548,
+      "learning_rate": 0.00016080402010050252,
+      "loss": 1.2845,
+      "step": 200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.335781455039978,
+      "learning_rate": 0.0001606030150753769,
+      "loss": 1.3783,
+      "step": 201
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3324490189552307,
+      "learning_rate": 0.00016040201005025127,
+      "loss": 1.3027,
+      "step": 202
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3381625711917877,
+      "learning_rate": 0.00016020100502512564,
+      "loss": 1.1912,
+      "step": 203
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3119088113307953,
+      "learning_rate": 0.00016,
+      "loss": 1.2789,
+      "step": 204
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32320040464401245,
+      "learning_rate": 0.00015979899497487439,
+      "loss": 1.2454,
+      "step": 205
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32925647497177124,
+      "learning_rate": 0.00015959798994974876,
+      "loss": 1.2602,
+      "step": 206
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32569676637649536,
+      "learning_rate": 0.0001593969849246231,
+      "loss": 1.2034,
+      "step": 207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3359280824661255,
+      "learning_rate": 0.0001591959798994975,
+      "loss": 1.2624,
+      "step": 208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3207138180732727,
+      "learning_rate": 0.00015899497487437188,
+      "loss": 1.2814,
+      "step": 209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32691851258277893,
+      "learning_rate": 0.00015879396984924625,
+      "loss": 1.2213,
+      "step": 210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33548569679260254,
+      "learning_rate": 0.0001585929648241206,
+      "loss": 1.133,
+      "step": 211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3024287819862366,
+      "learning_rate": 0.000158391959798995,
+      "loss": 1.1925,
+      "step": 212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3605235517024994,
+      "learning_rate": 0.00015819095477386937,
+      "loss": 1.3439,
+      "step": 213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33820798993110657,
+      "learning_rate": 0.00015798994974874372,
+      "loss": 1.1816,
+      "step": 214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32633164525032043,
+      "learning_rate": 0.0001577889447236181,
+      "loss": 1.2521,
+      "step": 215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34430956840515137,
+      "learning_rate": 0.00015758793969849246,
+      "loss": 1.3119,
+      "step": 216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32302767038345337,
+      "learning_rate": 0.00015738693467336686,
+      "loss": 1.2437,
+      "step": 217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3343082070350647,
+      "learning_rate": 0.0001571859296482412,
+      "loss": 1.298,
+      "step": 218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34785017371177673,
+      "learning_rate": 0.00015698492462311558,
+      "loss": 1.2371,
+      "step": 219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3445768356323242,
+      "learning_rate": 0.00015678391959798995,
+      "loss": 1.2617,
+      "step": 220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3664666414260864,
+      "learning_rate": 0.00015658291457286433,
+      "loss": 1.1819,
+      "step": 221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3051821291446686,
+      "learning_rate": 0.0001563819095477387,
+      "loss": 1.3238,
+      "step": 222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.31191888451576233,
+      "learning_rate": 0.00015618090452261307,
+      "loss": 1.2955,
+      "step": 223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33101990818977356,
+      "learning_rate": 0.00015597989949748745,
+      "loss": 1.2489,
+      "step": 224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3157234489917755,
+      "learning_rate": 0.00015577889447236182,
+      "loss": 1.3097,
+      "step": 225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29264160990715027,
+      "learning_rate": 0.0001555778894472362,
+      "loss": 1.2254,
+      "step": 226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34887459874153137,
+      "learning_rate": 0.00015537688442211056,
+      "loss": 1.2773,
+      "step": 227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3264656960964203,
+      "learning_rate": 0.00015517587939698494,
+      "loss": 1.2045,
+      "step": 228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3183201849460602,
+      "learning_rate": 0.0001549748743718593,
+      "loss": 1.3232,
+      "step": 229
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3423653542995453,
+      "learning_rate": 0.00015477386934673368,
+      "loss": 1.2111,
+      "step": 230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.32361966371536255,
+      "learning_rate": 0.00015457286432160806,
+      "loss": 1.2766,
+      "step": 231
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.35348379611968994,
+      "learning_rate": 0.0001543718592964824,
+      "loss": 1.2533,
+      "step": 232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.348850280046463,
+      "learning_rate": 0.0001541708542713568,
+      "loss": 1.2907,
+      "step": 233
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31669101119041443,
+      "learning_rate": 0.00015396984924623117,
+      "loss": 1.1785,
+      "step": 234
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3331408202648163,
+      "learning_rate": 0.00015376884422110555,
+      "loss": 1.2429,
+      "step": 235
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3349299728870392,
+      "learning_rate": 0.0001535678391959799,
+      "loss": 1.1955,
+      "step": 236
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3368314802646637,
+      "learning_rate": 0.00015336683417085427,
+      "loss": 1.3476,
+      "step": 237
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3111830949783325,
+      "learning_rate": 0.00015316582914572867,
+      "loss": 1.2382,
+      "step": 238
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.32492902874946594,
+      "learning_rate": 0.000152964824120603,
+      "loss": 1.2453,
+      "step": 239
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3176097869873047,
+      "learning_rate": 0.00015276381909547739,
+      "loss": 1.2639,
+      "step": 240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3306467831134796,
+      "learning_rate": 0.00015256281407035176,
+      "loss": 1.2979,
+      "step": 241
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33482393622398376,
+      "learning_rate": 0.00015236180904522613,
+      "loss": 1.2855,
+      "step": 242
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.317231684923172,
+      "learning_rate": 0.0001521608040201005,
+      "loss": 1.336,
+      "step": 243
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3562380075454712,
+      "learning_rate": 0.00015195979899497488,
+      "loss": 1.2538,
+      "step": 244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3557191491127014,
+      "learning_rate": 0.00015175879396984925,
+      "loss": 1.1941,
+      "step": 245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33679717779159546,
+      "learning_rate": 0.00015155778894472362,
+      "loss": 1.2929,
+      "step": 246
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.32738903164863586,
+      "learning_rate": 0.000151356783919598,
+      "loss": 1.1497,
+      "step": 247
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3145580589771271,
+      "learning_rate": 0.00015115577889447237,
+      "loss": 1.2982,
+      "step": 248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3444727659225464,
+      "learning_rate": 0.00015095477386934674,
+      "loss": 1.3321,
+      "step": 249
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3220258355140686,
+      "learning_rate": 0.00015075376884422112,
+      "loss": 1.1777,
+      "step": 250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.32968461513519287,
+      "learning_rate": 0.0001505527638190955,
+      "loss": 1.2707,
+      "step": 251
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3543086647987366,
+      "learning_rate": 0.00015035175879396986,
+      "loss": 1.2997,
+      "step": 252
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32566267251968384,
+      "learning_rate": 0.00015015075376884423,
+      "loss": 1.2313,
+      "step": 253
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31076309084892273,
+      "learning_rate": 0.0001499497487437186,
+      "loss": 1.3003,
+      "step": 254
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33521464467048645,
+      "learning_rate": 0.00014974874371859298,
+      "loss": 1.2593,
+      "step": 255
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3666183054447174,
+      "learning_rate": 0.00014954773869346735,
+      "loss": 1.3473,
+      "step": 256
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36615288257598877,
+      "learning_rate": 0.0001493467336683417,
+      "loss": 1.2521,
+      "step": 257
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3385326564311981,
+      "learning_rate": 0.0001491457286432161,
+      "loss": 1.3117,
+      "step": 258
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32243263721466064,
+      "learning_rate": 0.00014894472361809047,
+      "loss": 1.1798,
+      "step": 259
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3227294087409973,
+      "learning_rate": 0.00014874371859296482,
+      "loss": 1.2362,
+      "step": 260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3345654606819153,
+      "learning_rate": 0.0001485427135678392,
+      "loss": 1.2136,
+      "step": 261
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31397944688796997,
+      "learning_rate": 0.00014834170854271356,
+      "loss": 1.1987,
+      "step": 262
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3394251763820648,
+      "learning_rate": 0.00014814070351758796,
+      "loss": 1.2022,
+      "step": 263
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33496084809303284,
+      "learning_rate": 0.0001479396984924623,
+      "loss": 1.302,
+      "step": 264
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.35757845640182495,
+      "learning_rate": 0.00014773869346733668,
+      "loss": 1.2574,
+      "step": 265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.332405686378479,
+      "learning_rate": 0.00014753768844221106,
+      "loss": 1.2773,
+      "step": 266
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32756730914115906,
+      "learning_rate": 0.00014733668341708543,
+      "loss": 1.2189,
+      "step": 267
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3382038176059723,
+      "learning_rate": 0.0001471356783919598,
+      "loss": 1.2143,
+      "step": 268
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.35607925057411194,
+      "learning_rate": 0.00014693467336683417,
+      "loss": 1.3101,
+      "step": 269
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3490254282951355,
+      "learning_rate": 0.00014673366834170855,
+      "loss": 1.2813,
+      "step": 270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34010350704193115,
+      "learning_rate": 0.00014653266331658292,
+      "loss": 1.1883,
+      "step": 271
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33997610211372375,
+      "learning_rate": 0.0001463316582914573,
+      "loss": 1.2936,
+      "step": 272
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3269011378288269,
+      "learning_rate": 0.00014613065326633167,
+      "loss": 1.2682,
+      "step": 273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34441789984703064,
+      "learning_rate": 0.00014592964824120604,
+      "loss": 1.2556,
+      "step": 274
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3339982330799103,
+      "learning_rate": 0.0001457286432160804,
+      "loss": 1.3609,
+      "step": 275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3199785649776459,
+      "learning_rate": 0.00014552763819095479,
+      "loss": 1.276,
+      "step": 276
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33970314264297485,
+      "learning_rate": 0.00014532663316582916,
+      "loss": 1.2971,
+      "step": 277
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3045497536659241,
+      "learning_rate": 0.00014512562814070353,
+      "loss": 1.2067,
+      "step": 278
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.334547221660614,
+      "learning_rate": 0.0001449246231155779,
+      "loss": 1.2418,
+      "step": 279
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32451215386390686,
+      "learning_rate": 0.00014472361809045228,
+      "loss": 1.3319,
+      "step": 280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33704888820648193,
+      "learning_rate": 0.00014452261306532665,
+      "loss": 1.3065,
+      "step": 281
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3138297200202942,
+      "learning_rate": 0.000144321608040201,
+      "loss": 1.1864,
+      "step": 282
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31389278173446655,
+      "learning_rate": 0.00014412060301507537,
+      "loss": 1.2409,
+      "step": 283
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33956632018089294,
+      "learning_rate": 0.00014391959798994977,
+      "loss": 1.2591,
+      "step": 284
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3188384473323822,
+      "learning_rate": 0.00014371859296482411,
+      "loss": 1.232,
+      "step": 285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36883220076560974,
+      "learning_rate": 0.0001435175879396985,
+      "loss": 1.2569,
+      "step": 286
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3178212344646454,
+      "learning_rate": 0.00014331658291457286,
+      "loss": 1.2337,
+      "step": 287
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33480581641197205,
+      "learning_rate": 0.00014311557788944726,
+      "loss": 1.2784,
+      "step": 288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3228307366371155,
+      "learning_rate": 0.0001429145728643216,
+      "loss": 1.2186,
+      "step": 289
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3270285427570343,
+      "learning_rate": 0.00014271356783919598,
+      "loss": 1.252,
+      "step": 290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32656142115592957,
+      "learning_rate": 0.00014251256281407035,
+      "loss": 1.2598,
+      "step": 291
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3287805914878845,
+      "learning_rate": 0.00014231155778894473,
+      "loss": 1.2528,
+      "step": 292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.351793110370636,
+      "learning_rate": 0.0001421105527638191,
+      "loss": 1.1774,
+      "step": 293
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.334957093000412,
+      "learning_rate": 0.00014190954773869347,
+      "loss": 1.2204,
+      "step": 294
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3303321897983551,
+      "learning_rate": 0.00014170854271356784,
+      "loss": 1.2792,
+      "step": 295
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.325514018535614,
+      "learning_rate": 0.00014150753768844222,
+      "loss": 1.2095,
+      "step": 296
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32064923644065857,
+      "learning_rate": 0.0001413065326633166,
+      "loss": 1.2369,
+      "step": 297
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3173045516014099,
+      "learning_rate": 0.00014110552763819096,
+      "loss": 1.2092,
+      "step": 298
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3447834551334381,
+      "learning_rate": 0.00014090452261306534,
+      "loss": 1.3058,
+      "step": 299
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.355277419090271,
+      "learning_rate": 0.0001407035175879397,
+      "loss": 1.2147,
+      "step": 300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.321415513753891,
+      "learning_rate": 0.00014050251256281408,
+      "loss": 1.3128,
+      "step": 301
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.316572904586792,
+      "learning_rate": 0.00014030150753768846,
+      "loss": 1.2563,
+      "step": 302
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.35804814100265503,
+      "learning_rate": 0.0001401005025125628,
+      "loss": 1.3382,
+      "step": 303
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32747843861579895,
+      "learning_rate": 0.0001398994974874372,
+      "loss": 1.198,
+      "step": 304
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.35342931747436523,
+      "learning_rate": 0.00013969849246231157,
+      "loss": 1.2672,
+      "step": 305
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32692718505859375,
+      "learning_rate": 0.00013949748743718595,
+      "loss": 1.2641,
+      "step": 306
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31664589047431946,
+      "learning_rate": 0.0001392964824120603,
+      "loss": 1.2382,
+      "step": 307
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.33936336636543274,
+      "learning_rate": 0.00013909547738693467,
+      "loss": 1.2823,
+      "step": 308
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.342006117105484,
+      "learning_rate": 0.00013889447236180907,
+      "loss": 1.2712,
+      "step": 309
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31698647141456604,
+      "learning_rate": 0.0001386934673366834,
+      "loss": 1.2551,
+      "step": 310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32440513372421265,
+      "learning_rate": 0.00013849246231155778,
+      "loss": 1.2512,
+      "step": 311
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3394576907157898,
+      "learning_rate": 0.00013829145728643216,
+      "loss": 1.3612,
+      "step": 312
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3272732198238373,
+      "learning_rate": 0.00013809045226130656,
+      "loss": 1.1537,
+      "step": 313
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.33435794711112976,
+      "learning_rate": 0.0001378894472361809,
+      "loss": 1.2832,
+      "step": 314
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3546105921268463,
+      "learning_rate": 0.00013768844221105528,
+      "loss": 1.2398,
+      "step": 315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3501565754413605,
+      "learning_rate": 0.00013748743718592965,
+      "loss": 1.2111,
+      "step": 316
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.35097482800483704,
+      "learning_rate": 0.00013728643216080402,
+      "loss": 1.3033,
+      "step": 317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3313996195793152,
+      "learning_rate": 0.0001370854271356784,
+      "loss": 1.2819,
+      "step": 318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.32861942052841187,
+      "learning_rate": 0.00013688442211055277,
+      "loss": 1.2566,
+      "step": 319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33757033944129944,
+      "learning_rate": 0.00013668341708542714,
+      "loss": 1.2951,
+      "step": 320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33206436038017273,
+      "learning_rate": 0.00013648241206030151,
+      "loss": 1.1828,
+      "step": 321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.34780171513557434,
+      "learning_rate": 0.0001362814070351759,
+      "loss": 1.2766,
+      "step": 322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3237352669239044,
+      "learning_rate": 0.00013608040201005026,
+      "loss": 1.2885,
+      "step": 323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.32298171520233154,
+      "learning_rate": 0.00013587939698492463,
+      "loss": 1.1687,
+      "step": 324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3167840540409088,
+      "learning_rate": 0.000135678391959799,
+      "loss": 1.2617,
+      "step": 325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3790431618690491,
+      "learning_rate": 0.00013547738693467338,
+      "loss": 1.238,
+      "step": 326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.34648290276527405,
+      "learning_rate": 0.00013527638190954775,
+      "loss": 1.2802,
+      "step": 327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.32841557264328003,
+      "learning_rate": 0.0001350753768844221,
+      "loss": 1.2493,
+      "step": 328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3556242287158966,
+      "learning_rate": 0.00013487437185929647,
+      "loss": 1.2269,
+      "step": 329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38643625378608704,
+      "learning_rate": 0.00013467336683417087,
+      "loss": 1.3036,
+      "step": 330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33449631929397583,
+      "learning_rate": 0.00013447236180904524,
+      "loss": 1.2595,
+      "step": 331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33076462149620056,
+      "learning_rate": 0.0001342713567839196,
+      "loss": 1.2376,
+      "step": 332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3205597698688507,
+      "learning_rate": 0.00013407035175879396,
+      "loss": 1.2323,
+      "step": 333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.31788671016693115,
+      "learning_rate": 0.00013386934673366836,
+      "loss": 1.2318,
+      "step": 334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3543769419193268,
+      "learning_rate": 0.0001336683417085427,
+      "loss": 1.2916,
+      "step": 335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.31310132145881653,
+      "learning_rate": 0.00013346733668341708,
+      "loss": 1.1925,
+      "step": 336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3153649866580963,
+      "learning_rate": 0.00013326633165829146,
+      "loss": 1.282,
+      "step": 337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33127865195274353,
+      "learning_rate": 0.00013306532663316586,
+      "loss": 1.2753,
+      "step": 338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36392638087272644,
+      "learning_rate": 0.0001328643216080402,
+      "loss": 1.2895,
+      "step": 339
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33100610971450806,
+      "learning_rate": 0.00013266331658291457,
+      "loss": 1.2334,
+      "step": 340
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3381032347679138,
+      "learning_rate": 0.00013246231155778895,
+      "loss": 1.2819,
+      "step": 341
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31431353092193604,
+      "learning_rate": 0.00013226130653266332,
+      "loss": 1.2468,
+      "step": 342
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31695157289505005,
+      "learning_rate": 0.0001320603015075377,
+      "loss": 1.2349,
+      "step": 343
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3341975808143616,
+      "learning_rate": 0.00013185929648241207,
+      "loss": 1.266,
+      "step": 344
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.32825469970703125,
+      "learning_rate": 0.00013165829145728644,
+      "loss": 1.28,
+      "step": 345
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3523416817188263,
+      "learning_rate": 0.0001314572864321608,
+      "loss": 1.1085,
+      "step": 346
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3559586703777313,
+      "learning_rate": 0.00013125628140703518,
+      "loss": 1.3506,
+      "step": 347
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33866238594055176,
+      "learning_rate": 0.00013105527638190956,
+      "loss": 1.1631,
+      "step": 348
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3517313301563263,
+      "learning_rate": 0.00013085427135678393,
+      "loss": 1.2348,
+      "step": 349
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3219257891178131,
+      "learning_rate": 0.0001306532663316583,
+      "loss": 1.1565,
+      "step": 350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3182775676250458,
+      "learning_rate": 0.00013045226130653268,
+      "loss": 1.3332,
+      "step": 351
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33609387278556824,
+      "learning_rate": 0.00013025125628140705,
+      "loss": 1.2197,
+      "step": 352
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3392961621284485,
+      "learning_rate": 0.0001300502512562814,
+      "loss": 1.3303,
+      "step": 353
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3233344256877899,
+      "learning_rate": 0.00012984924623115577,
+      "loss": 1.1869,
+      "step": 354
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3254396617412567,
+      "learning_rate": 0.00012964824120603017,
+      "loss": 1.2693,
+      "step": 355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3262885510921478,
+      "learning_rate": 0.00012944723618090454,
+      "loss": 1.2837,
+      "step": 356
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3186768889427185,
+      "learning_rate": 0.0001292462311557789,
+      "loss": 1.2705,
+      "step": 357
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3470524251461029,
+      "learning_rate": 0.00012904522613065326,
+      "loss": 1.2626,
+      "step": 358
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31644207239151,
+      "learning_rate": 0.00012884422110552766,
+      "loss": 1.2217,
+      "step": 359
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3402186930179596,
+      "learning_rate": 0.000128643216080402,
+      "loss": 1.2456,
+      "step": 360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32568660378456116,
+      "learning_rate": 0.00012844221105527638,
+      "loss": 1.3024,
+      "step": 361
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31550562381744385,
+      "learning_rate": 0.00012824120603015075,
+      "loss": 1.2512,
+      "step": 362
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3515610098838806,
+      "learning_rate": 0.00012804020100502515,
+      "loss": 1.2323,
+      "step": 363
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3370158076286316,
+      "learning_rate": 0.0001278391959798995,
+      "loss": 1.3072,
+      "step": 364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3440285623073578,
+      "learning_rate": 0.00012763819095477387,
+      "loss": 1.2268,
+      "step": 365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37662672996520996,
+      "learning_rate": 0.00012743718592964824,
+      "loss": 1.2495,
+      "step": 366
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32195794582366943,
+      "learning_rate": 0.00012723618090452262,
+      "loss": 1.2977,
+      "step": 367
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3311251103878021,
+      "learning_rate": 0.000127035175879397,
+      "loss": 1.3372,
+      "step": 368
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3319653272628784,
+      "learning_rate": 0.00012683417085427136,
+      "loss": 1.2813,
+      "step": 369
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3444850742816925,
+      "learning_rate": 0.00012663316582914574,
+      "loss": 1.2877,
+      "step": 370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3433425724506378,
+      "learning_rate": 0.0001264321608040201,
+      "loss": 1.2696,
+      "step": 371
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35098111629486084,
+      "learning_rate": 0.00012623115577889448,
+      "loss": 1.28,
+      "step": 372
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3203146457672119,
+      "learning_rate": 0.00012603015075376885,
+      "loss": 1.2533,
+      "step": 373
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30470582842826843,
+      "learning_rate": 0.00012582914572864323,
+      "loss": 1.2678,
+      "step": 374
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.324220210313797,
+      "learning_rate": 0.0001256281407035176,
+      "loss": 1.1909,
+      "step": 375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3313479423522949,
+      "learning_rate": 0.00012542713567839197,
+      "loss": 1.2528,
+      "step": 376
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31879833340644836,
+      "learning_rate": 0.00012522613065326635,
+      "loss": 1.1431,
+      "step": 377
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3483116626739502,
+      "learning_rate": 0.0001250251256281407,
+      "loss": 1.1517,
+      "step": 378
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3220193684101105,
+      "learning_rate": 0.00012482412060301507,
+      "loss": 1.2532,
+      "step": 379
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3391655683517456,
+      "learning_rate": 0.00012462311557788947,
+      "loss": 1.2565,
+      "step": 380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3446550667285919,
+      "learning_rate": 0.00012442211055276384,
+      "loss": 1.2253,
+      "step": 381
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3528743386268616,
+      "learning_rate": 0.00012422110552763818,
+      "loss": 1.2919,
+      "step": 382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.32574883103370667,
+      "learning_rate": 0.00012402010050251256,
+      "loss": 1.2885,
+      "step": 383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3145955502986908,
+      "learning_rate": 0.00012381909547738696,
+      "loss": 1.2307,
+      "step": 384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3239680230617523,
+      "learning_rate": 0.0001236180904522613,
+      "loss": 1.2612,
+      "step": 385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3375207185745239,
+      "learning_rate": 0.00012341708542713568,
+      "loss": 1.274,
+      "step": 386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3346465528011322,
+      "learning_rate": 0.00012321608040201005,
+      "loss": 1.2074,
+      "step": 387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3280505836009979,
+      "learning_rate": 0.00012301507537688445,
+      "loss": 1.2776,
+      "step": 388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3411586880683899,
+      "learning_rate": 0.0001228140703517588,
+      "loss": 1.3012,
+      "step": 389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.32394883036613464,
+      "learning_rate": 0.00012261306532663317,
+      "loss": 1.3018,
+      "step": 390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34626421332359314,
+      "learning_rate": 0.00012241206030150754,
+      "loss": 1.2882,
+      "step": 391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3305688500404358,
+      "learning_rate": 0.00012221105527638191,
+      "loss": 1.2484,
+      "step": 392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33277568221092224,
+      "learning_rate": 0.00012201005025125629,
+      "loss": 1.2088,
+      "step": 393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3431893289089203,
+      "learning_rate": 0.00012180904522613066,
+      "loss": 1.1882,
+      "step": 394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3225569725036621,
+      "learning_rate": 0.00012160804020100502,
+      "loss": 1.2162,
+      "step": 395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33999207615852356,
+      "learning_rate": 0.00012140703517587942,
+      "loss": 1.2532,
+      "step": 396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3457259237766266,
+      "learning_rate": 0.00012120603015075378,
+      "loss": 1.2045,
+      "step": 397
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.35479119420051575,
+      "learning_rate": 0.00012100502512562815,
+      "loss": 1.2475,
+      "step": 398
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34909576177597046,
+      "learning_rate": 0.00012080402010050251,
+      "loss": 1.2581,
+      "step": 399
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33657559752464294,
+      "learning_rate": 0.00012060301507537688,
+      "loss": 1.1876,
+      "step": 400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33252567052841187,
+      "learning_rate": 0.00012040201005025127,
+      "loss": 1.2108,
+      "step": 401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.35932156443595886,
+      "learning_rate": 0.00012020100502512563,
+      "loss": 1.25,
+      "step": 402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3509422242641449,
+      "learning_rate": 0.00012,
+      "loss": 1.2748,
+      "step": 403
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3509500324726105,
+      "learning_rate": 0.00011979899497487436,
+      "loss": 1.2704,
+      "step": 404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32239145040512085,
+      "learning_rate": 0.00011959798994974876,
+      "loss": 1.2476,
+      "step": 405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33603423833847046,
+      "learning_rate": 0.00011939698492462312,
+      "loss": 1.278,
+      "step": 406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3381786346435547,
+      "learning_rate": 0.0001191959798994975,
+      "loss": 1.2382,
+      "step": 407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31310775876045227,
+      "learning_rate": 0.00011899497487437185,
+      "loss": 1.2827,
+      "step": 408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3387271463871002,
+      "learning_rate": 0.00011879396984924624,
+      "loss": 1.2891,
+      "step": 409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3353903293609619,
+      "learning_rate": 0.00011859296482412061,
+      "loss": 1.326,
+      "step": 410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.322992742061615,
+      "learning_rate": 0.00011839195979899497,
+      "loss": 1.2513,
+      "step": 411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3425077199935913,
+      "learning_rate": 0.00011819095477386935,
+      "loss": 1.1482,
+      "step": 412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3305937647819519,
+      "learning_rate": 0.00011798994974874373,
+      "loss": 1.1919,
+      "step": 413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3408913016319275,
+      "learning_rate": 0.0001177889447236181,
+      "loss": 1.2535,
+      "step": 414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43716689944267273,
+      "learning_rate": 0.00011758793969849247,
+      "loss": 1.33,
+      "step": 415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34090203046798706,
+      "learning_rate": 0.00011738693467336684,
+      "loss": 1.1666,
+      "step": 416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.35914671421051025,
+      "learning_rate": 0.00011718592964824122,
+      "loss": 1.2683,
+      "step": 417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3459693193435669,
+      "learning_rate": 0.00011698492462311558,
+      "loss": 1.2502,
+      "step": 418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3254222273826599,
+      "learning_rate": 0.00011678391959798996,
+      "loss": 1.2369,
+      "step": 419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33233174681663513,
+      "learning_rate": 0.00011658291457286432,
+      "loss": 1.1769,
+      "step": 420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3394586145877838,
+      "learning_rate": 0.00011638190954773872,
+      "loss": 1.1441,
+      "step": 421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3482055962085724,
+      "learning_rate": 0.00011618090452261308,
+      "loss": 1.1411,
+      "step": 422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33943256735801697,
+      "learning_rate": 0.00011597989949748745,
+      "loss": 1.274,
+      "step": 423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34545761346817017,
+      "learning_rate": 0.00011577889447236181,
+      "loss": 1.1839,
+      "step": 424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3279217481613159,
+      "learning_rate": 0.00011557788944723618,
+      "loss": 1.2555,
+      "step": 425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3297037184238434,
+      "learning_rate": 0.00011537688442211057,
+      "loss": 1.2645,
+      "step": 426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3765062391757965,
+      "learning_rate": 0.00011517587939698493,
+      "loss": 1.2166,
+      "step": 427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34099629521369934,
+      "learning_rate": 0.0001149748743718593,
+      "loss": 1.1985,
+      "step": 428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3941348195075989,
+      "learning_rate": 0.00011477386934673366,
+      "loss": 1.2124,
+      "step": 429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3421550989151001,
+      "learning_rate": 0.00011457286432160806,
+      "loss": 1.2202,
+      "step": 430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3587627708911896,
+      "learning_rate": 0.00011437185929648242,
+      "loss": 1.2119,
+      "step": 431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.318024605512619,
+      "learning_rate": 0.00011417085427135679,
+      "loss": 1.1717,
+      "step": 432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3441738486289978,
+      "learning_rate": 0.00011396984924623115,
+      "loss": 1.2637,
+      "step": 433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32831835746765137,
+      "learning_rate": 0.00011376884422110554,
+      "loss": 1.1351,
+      "step": 434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40580299496650696,
+      "learning_rate": 0.00011356783919597991,
+      "loss": 1.2777,
+      "step": 435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3455897867679596,
+      "learning_rate": 0.00011336683417085427,
+      "loss": 1.2711,
+      "step": 436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34554949402809143,
+      "learning_rate": 0.00011316582914572864,
+      "loss": 1.2605,
+      "step": 437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.333046019077301,
+      "learning_rate": 0.00011296482412060303,
+      "loss": 1.1934,
+      "step": 438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37090611457824707,
+      "learning_rate": 0.0001127638190954774,
+      "loss": 1.2289,
+      "step": 439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3464914560317993,
+      "learning_rate": 0.00011256281407035176,
+      "loss": 1.2573,
+      "step": 440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3522966504096985,
+      "learning_rate": 0.00011236180904522614,
+      "loss": 1.233,
+      "step": 441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34124764800071716,
+      "learning_rate": 0.00011216080402010052,
+      "loss": 1.2167,
+      "step": 442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32103344798088074,
+      "learning_rate": 0.00011195979899497488,
+      "loss": 1.2823,
+      "step": 443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.35400694608688354,
+      "learning_rate": 0.00011175879396984925,
+      "loss": 1.275,
+      "step": 444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32526448369026184,
+      "learning_rate": 0.00011155778894472361,
+      "loss": 1.2298,
+      "step": 445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32495397329330444,
+      "learning_rate": 0.00011135678391959799,
+      "loss": 1.2328,
+      "step": 446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33901894092559814,
+      "learning_rate": 0.00011115577889447237,
+      "loss": 1.2864,
+      "step": 447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3219538927078247,
+      "learning_rate": 0.00011095477386934675,
+      "loss": 1.1061,
+      "step": 448
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3175276219844818,
+      "learning_rate": 0.0001107537688442211,
+      "loss": 1.2714,
+      "step": 449
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.32311904430389404,
+      "learning_rate": 0.00011055276381909548,
+      "loss": 1.2635,
+      "step": 450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3252653181552887,
+      "learning_rate": 0.00011035175879396986,
+      "loss": 1.2694,
+      "step": 451
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.337410032749176,
+      "learning_rate": 0.00011015075376884422,
+      "loss": 1.2197,
+      "step": 452
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34929850697517395,
+      "learning_rate": 0.0001099497487437186,
+      "loss": 1.2773,
+      "step": 453
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3558543622493744,
+      "learning_rate": 0.00010974874371859296,
+      "loss": 1.2673,
+      "step": 454
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.309593141078949,
+      "learning_rate": 0.00010954773869346736,
+      "loss": 1.2945,
+      "step": 455
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30904704332351685,
+      "learning_rate": 0.00010934673366834172,
+      "loss": 1.2362,
+      "step": 456
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3560062646865845,
+      "learning_rate": 0.00010914572864321609,
+      "loss": 1.2874,
+      "step": 457
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3410942554473877,
+      "learning_rate": 0.00010894472361809045,
+      "loss": 1.2621,
+      "step": 458
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33610406517982483,
+      "learning_rate": 0.00010874371859296483,
+      "loss": 1.2572,
+      "step": 459
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3685830235481262,
+      "learning_rate": 0.00010854271356783921,
+      "loss": 1.292,
+      "step": 460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3263039290904999,
+      "learning_rate": 0.00010834170854271357,
+      "loss": 1.1467,
+      "step": 461
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33784759044647217,
+      "learning_rate": 0.00010814070351758794,
+      "loss": 1.2536,
+      "step": 462
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3310985565185547,
+      "learning_rate": 0.00010793969849246233,
+      "loss": 1.2865,
+      "step": 463
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3608328104019165,
+      "learning_rate": 0.0001077386934673367,
+      "loss": 1.3144,
+      "step": 464
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3107350468635559,
+      "learning_rate": 0.00010753768844221106,
+      "loss": 1.1831,
+      "step": 465
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3376270532608032,
+      "learning_rate": 0.00010733668341708543,
+      "loss": 1.2699,
+      "step": 466
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34757518768310547,
+      "learning_rate": 0.00010713567839195982,
+      "loss": 1.3423,
+      "step": 467
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3217342495918274,
+      "learning_rate": 0.00010693467336683418,
+      "loss": 1.2629,
+      "step": 468
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3594968020915985,
+      "learning_rate": 0.00010673366834170855,
+      "loss": 1.2613,
+      "step": 469
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34216034412384033,
+      "learning_rate": 0.00010653266331658291,
+      "loss": 1.2807,
+      "step": 470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33661434054374695,
+      "learning_rate": 0.00010633165829145728,
+      "loss": 1.2175,
+      "step": 471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3459634780883789,
+      "learning_rate": 0.00010613065326633167,
+      "loss": 1.2234,
+      "step": 472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.31939029693603516,
+      "learning_rate": 0.00010592964824120604,
+      "loss": 1.2566,
+      "step": 473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3308617174625397,
+      "learning_rate": 0.0001057286432160804,
+      "loss": 1.2249,
+      "step": 474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3457432985305786,
+      "learning_rate": 0.00010552763819095478,
+      "loss": 1.3377,
+      "step": 475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3623298704624176,
+      "learning_rate": 0.00010532663316582916,
+      "loss": 1.295,
+      "step": 476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3287794888019562,
+      "learning_rate": 0.00010512562814070352,
+      "loss": 1.3167,
+      "step": 477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32969963550567627,
+      "learning_rate": 0.0001049246231155779,
+      "loss": 1.1917,
+      "step": 478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3520050346851349,
+      "learning_rate": 0.00010472361809045225,
+      "loss": 1.2786,
+      "step": 479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33835569024086,
+      "learning_rate": 0.00010452261306532664,
+      "loss": 1.2734,
+      "step": 480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32975468039512634,
+      "learning_rate": 0.00010432160804020101,
+      "loss": 1.2263,
+      "step": 481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3600429594516754,
+      "learning_rate": 0.00010412060301507539,
+      "loss": 1.308,
+      "step": 482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35504522919654846,
+      "learning_rate": 0.00010391959798994975,
+      "loss": 1.3255,
+      "step": 483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.334204763174057,
+      "learning_rate": 0.00010371859296482413,
+      "loss": 1.2029,
+      "step": 484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32885733246803284,
+      "learning_rate": 0.0001035175879396985,
+      "loss": 1.258,
+      "step": 485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3293534815311432,
+      "learning_rate": 0.00010331658291457286,
+      "loss": 1.2438,
+      "step": 486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3399008810520172,
+      "learning_rate": 0.00010311557788944724,
+      "loss": 1.2508,
+      "step": 487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3626408576965332,
+      "learning_rate": 0.00010291457286432162,
+      "loss": 1.2555,
+      "step": 488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3435054123401642,
+      "learning_rate": 0.00010271356783919598,
+      "loss": 1.1557,
+      "step": 489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35252466797828674,
+      "learning_rate": 0.00010251256281407036,
+      "loss": 1.2756,
+      "step": 490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3346278667449951,
+      "learning_rate": 0.00010231155778894473,
+      "loss": 1.2722,
+      "step": 491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33955395221710205,
+      "learning_rate": 0.00010211055276381909,
+      "loss": 1.2445,
+      "step": 492
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33230292797088623,
+      "learning_rate": 0.00010190954773869348,
+      "loss": 1.2836,
+      "step": 493
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.32656341791152954,
+      "learning_rate": 0.00010170854271356785,
+      "loss": 1.2414,
+      "step": 494
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3479287922382355,
+      "learning_rate": 0.00010150753768844221,
+      "loss": 1.2442,
+      "step": 495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3435857892036438,
+      "learning_rate": 0.00010130653266331658,
+      "loss": 1.2344,
+      "step": 496
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3324833810329437,
+      "learning_rate": 0.00010110552763819097,
+      "loss": 1.2261,
+      "step": 497
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3612833023071289,
+      "learning_rate": 0.00010090452261306533,
+      "loss": 1.3509,
+      "step": 498
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3381580710411072,
+      "learning_rate": 0.0001007035175879397,
+      "loss": 1.2508,
+      "step": 499
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3283715546131134,
+      "learning_rate": 0.00010050251256281407,
+      "loss": 1.223,
+      "step": 500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3245822787284851,
+      "learning_rate": 0.00010030150753768846,
+      "loss": 1.214,
+      "step": 501
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33582326769828796,
+      "learning_rate": 0.00010010050251256282,
+      "loss": 1.1891,
+      "step": 502
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.34178048372268677,
+      "learning_rate": 9.989949748743719e-05,
+      "loss": 1.2328,
+      "step": 503
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3429703414440155,
+      "learning_rate": 9.969849246231156e-05,
+      "loss": 1.243,
+      "step": 504
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3459545969963074,
+      "learning_rate": 9.949748743718594e-05,
+      "loss": 1.2272,
+      "step": 505
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.32898616790771484,
+      "learning_rate": 9.929648241206031e-05,
+      "loss": 1.2651,
+      "step": 506
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33411911129951477,
+      "learning_rate": 9.909547738693468e-05,
+      "loss": 1.2215,
+      "step": 507
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3196270763874054,
+      "learning_rate": 9.889447236180906e-05,
+      "loss": 1.2338,
+      "step": 508
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3360273241996765,
+      "learning_rate": 9.869346733668342e-05,
+      "loss": 1.1259,
+      "step": 509
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3572694957256317,
+      "learning_rate": 9.84924623115578e-05,
+      "loss": 1.2741,
+      "step": 510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3491540849208832,
+      "learning_rate": 9.829145728643216e-05,
+      "loss": 1.2844,
+      "step": 511
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3704037666320801,
+      "learning_rate": 9.809045226130655e-05,
+      "loss": 1.2974,
+      "step": 512
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3397068381309509,
+      "learning_rate": 9.788944723618091e-05,
+      "loss": 1.2155,
+      "step": 513
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3439743220806122,
+      "learning_rate": 9.768844221105528e-05,
+      "loss": 1.2449,
+      "step": 514
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3374169170856476,
+      "learning_rate": 9.748743718592965e-05,
+      "loss": 1.1984,
+      "step": 515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3484777510166168,
+      "learning_rate": 9.728643216080403e-05,
+      "loss": 1.2707,
+      "step": 516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34569051861763,
+      "learning_rate": 9.70854271356784e-05,
+      "loss": 1.2669,
+      "step": 517
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3295353949069977,
+      "learning_rate": 9.688442211055276e-05,
+      "loss": 1.3073,
+      "step": 518
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3298560380935669,
+      "learning_rate": 9.668341708542715e-05,
+      "loss": 1.1989,
+      "step": 519
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.342427521944046,
+      "learning_rate": 9.64824120603015e-05,
+      "loss": 1.2467,
+      "step": 520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34153419733047485,
+      "learning_rate": 9.628140703517589e-05,
+      "loss": 1.1852,
+      "step": 521
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33842480182647705,
+      "learning_rate": 9.608040201005025e-05,
+      "loss": 1.277,
+      "step": 522
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3495989739894867,
+      "learning_rate": 9.587939698492462e-05,
+      "loss": 1.2363,
+      "step": 523
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34229379892349243,
+      "learning_rate": 9.5678391959799e-05,
+      "loss": 1.2327,
+      "step": 524
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3576783835887909,
+      "learning_rate": 9.547738693467337e-05,
+      "loss": 1.2011,
+      "step": 525
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3553127646446228,
+      "learning_rate": 9.527638190954774e-05,
+      "loss": 1.2718,
+      "step": 526
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34771883487701416,
+      "learning_rate": 9.507537688442212e-05,
+      "loss": 1.2399,
+      "step": 527
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34981489181518555,
+      "learning_rate": 9.487437185929649e-05,
+      "loss": 1.2305,
+      "step": 528
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35138848423957825,
+      "learning_rate": 9.467336683417086e-05,
+      "loss": 1.3227,
+      "step": 529
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32845598459243774,
+      "learning_rate": 9.447236180904523e-05,
+      "loss": 1.3256,
+      "step": 530
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35754063725471497,
+      "learning_rate": 9.427135678391961e-05,
+      "loss": 1.28,
+      "step": 531
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3293386697769165,
+      "learning_rate": 9.407035175879397e-05,
+      "loss": 1.2435,
+      "step": 532
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32942119240760803,
+      "learning_rate": 9.386934673366835e-05,
+      "loss": 1.2703,
+      "step": 533
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36065587401390076,
+      "learning_rate": 9.366834170854271e-05,
+      "loss": 1.1724,
+      "step": 534
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32289040088653564,
+      "learning_rate": 9.34673366834171e-05,
+      "loss": 1.2722,
+      "step": 535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3467089533805847,
+      "learning_rate": 9.326633165829146e-05,
+      "loss": 1.2233,
+      "step": 536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3310029208660126,
+      "learning_rate": 9.306532663316585e-05,
+      "loss": 1.2507,
+      "step": 537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3369174301624298,
+      "learning_rate": 9.28643216080402e-05,
+      "loss": 1.2127,
+      "step": 538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34335601329803467,
+      "learning_rate": 9.266331658291458e-05,
+      "loss": 1.2038,
+      "step": 539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3327116072177887,
+      "learning_rate": 9.246231155778895e-05,
+      "loss": 1.3228,
+      "step": 540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3617047071456909,
+      "learning_rate": 9.226130653266331e-05,
+      "loss": 1.3661,
+      "step": 541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3532153069972992,
+      "learning_rate": 9.20603015075377e-05,
+      "loss": 1.2159,
+      "step": 542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3590898811817169,
+      "learning_rate": 9.185929648241206e-05,
+      "loss": 1.2507,
+      "step": 543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3496554493904114,
+      "learning_rate": 9.165829145728644e-05,
+      "loss": 1.2163,
+      "step": 544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34077826142311096,
+      "learning_rate": 9.14572864321608e-05,
+      "loss": 1.2127,
+      "step": 545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33904626965522766,
+      "learning_rate": 9.125628140703519e-05,
+      "loss": 1.1938,
+      "step": 546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34727028012275696,
+      "learning_rate": 9.105527638190955e-05,
+      "loss": 1.2685,
+      "step": 547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3311704397201538,
+      "learning_rate": 9.085427135678392e-05,
+      "loss": 1.2014,
+      "step": 548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3330560028553009,
+      "learning_rate": 9.06532663316583e-05,
+      "loss": 1.2679,
+      "step": 549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3092857003211975,
+      "learning_rate": 9.045226130653267e-05,
+      "loss": 1.1688,
+      "step": 550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3591139018535614,
+      "learning_rate": 9.025125628140704e-05,
+      "loss": 1.1914,
+      "step": 551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3442589044570923,
+      "learning_rate": 9.005025125628141e-05,
+      "loss": 1.2051,
+      "step": 552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.35013848543167114,
+      "learning_rate": 8.984924623115579e-05,
+      "loss": 1.2408,
+      "step": 553
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3634118139743805,
+      "learning_rate": 8.964824120603016e-05,
+      "loss": 1.2016,
+      "step": 554
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.32844385504722595,
+      "learning_rate": 8.944723618090453e-05,
+      "loss": 1.186,
+      "step": 555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3430251181125641,
+      "learning_rate": 8.92462311557789e-05,
+      "loss": 1.3732,
+      "step": 556
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.32605141401290894,
+      "learning_rate": 8.904522613065326e-05,
+      "loss": 1.2225,
+      "step": 557
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3491004407405853,
+      "learning_rate": 8.884422110552765e-05,
+      "loss": 1.277,
+      "step": 558
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34751296043395996,
+      "learning_rate": 8.864321608040201e-05,
+      "loss": 1.2069,
+      "step": 559
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31487154960632324,
+      "learning_rate": 8.84422110552764e-05,
+      "loss": 1.1973,
+      "step": 560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.33653488755226135,
+      "learning_rate": 8.824120603015076e-05,
+      "loss": 1.3103,
+      "step": 561
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3140471279621124,
+      "learning_rate": 8.804020100502513e-05,
+      "loss": 1.1919,
+      "step": 562
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37627485394477844,
+      "learning_rate": 8.78391959798995e-05,
+      "loss": 1.2185,
+      "step": 563
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34721139073371887,
+      "learning_rate": 8.763819095477387e-05,
+      "loss": 1.2937,
+      "step": 564
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31878143548965454,
+      "learning_rate": 8.743718592964825e-05,
+      "loss": 1.1966,
+      "step": 565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3476640284061432,
+      "learning_rate": 8.723618090452261e-05,
+      "loss": 1.1035,
+      "step": 566
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32448452711105347,
+      "learning_rate": 8.7035175879397e-05,
+      "loss": 1.2245,
+      "step": 567
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3310175836086273,
+      "learning_rate": 8.683417085427135e-05,
+      "loss": 1.2231,
+      "step": 568
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3311636745929718,
+      "learning_rate": 8.663316582914574e-05,
+      "loss": 1.2363,
+      "step": 569
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3398733139038086,
+      "learning_rate": 8.64321608040201e-05,
+      "loss": 1.2684,
+      "step": 570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35292455554008484,
+      "learning_rate": 8.623115577889449e-05,
+      "loss": 1.2251,
+      "step": 571
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3688778579235077,
+      "learning_rate": 8.603015075376884e-05,
+      "loss": 1.2493,
+      "step": 572
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3263533413410187,
+      "learning_rate": 8.582914572864322e-05,
+      "loss": 1.2479,
+      "step": 573
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.362587034702301,
+      "learning_rate": 8.562814070351759e-05,
+      "loss": 1.2296,
+      "step": 574
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3232894241809845,
+      "learning_rate": 8.542713567839196e-05,
+      "loss": 1.1811,
+      "step": 575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3491204082965851,
+      "learning_rate": 8.522613065326634e-05,
+      "loss": 1.306,
+      "step": 576
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3507627546787262,
+      "learning_rate": 8.502512562814071e-05,
+      "loss": 1.2438,
+      "step": 577
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34256500005722046,
+      "learning_rate": 8.482412060301508e-05,
+      "loss": 1.2765,
+      "step": 578
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.333816260099411,
+      "learning_rate": 8.462311557788946e-05,
+      "loss": 1.2372,
+      "step": 579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3247378170490265,
+      "learning_rate": 8.442211055276383e-05,
+      "loss": 1.2541,
+      "step": 580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31364428997039795,
+      "learning_rate": 8.42211055276382e-05,
+      "loss": 1.2676,
+      "step": 581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3368885815143585,
+      "learning_rate": 8.402010050251256e-05,
+      "loss": 1.214,
+      "step": 582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3350681662559509,
+      "learning_rate": 8.381909547738695e-05,
+      "loss": 1.1894,
+      "step": 583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3448706269264221,
+      "learning_rate": 8.36180904522613e-05,
+      "loss": 1.2945,
+      "step": 584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3268083930015564,
+      "learning_rate": 8.341708542713568e-05,
+      "loss": 1.1992,
+      "step": 585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3421708941459656,
+      "learning_rate": 8.321608040201005e-05,
+      "loss": 1.2229,
+      "step": 586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3358671963214874,
+      "learning_rate": 8.301507537688443e-05,
+      "loss": 1.1979,
+      "step": 587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3192279636859894,
+      "learning_rate": 8.28140703517588e-05,
+      "loss": 1.182,
+      "step": 588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32525762915611267,
+      "learning_rate": 8.261306532663317e-05,
+      "loss": 1.2582,
+      "step": 589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3301040232181549,
+      "learning_rate": 8.241206030150754e-05,
+      "loss": 1.1714,
+      "step": 590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33362287282943726,
+      "learning_rate": 8.22110552763819e-05,
+      "loss": 1.2263,
+      "step": 591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33714091777801514,
+      "learning_rate": 8.201005025125629e-05,
+      "loss": 1.2911,
+      "step": 592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33390864729881287,
+      "learning_rate": 8.180904522613065e-05,
+      "loss": 1.3051,
+      "step": 593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3238459527492523,
+      "learning_rate": 8.160804020100504e-05,
+      "loss": 1.2372,
+      "step": 594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.35366594791412354,
+      "learning_rate": 8.14070351758794e-05,
+      "loss": 1.1972,
+      "step": 595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33592501282691956,
+      "learning_rate": 8.120603015075378e-05,
+      "loss": 1.2653,
+      "step": 596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3256325423717499,
+      "learning_rate": 8.100502512562814e-05,
+      "loss": 1.2153,
+      "step": 597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3295387029647827,
+      "learning_rate": 8.080402010050251e-05,
+      "loss": 1.2275,
+      "step": 598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31947025656700134,
+      "learning_rate": 8.060301507537689e-05,
+      "loss": 1.1627,
+      "step": 599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3440350890159607,
+      "learning_rate": 8.040201005025126e-05,
+      "loss": 1.2488,
+      "step": 600
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "total_flos": 2.7787708687132262e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}