diff --git "a/checkpoint-1276/trainer_state.json" "b/checkpoint-1276/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1276/trainer_state.json"
@@ -0,0 +1,8966 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.998956158663883,
+  "eval_steps": 500,
+  "global_step": 1276,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0015657620041753654,
+      "grad_norm": 658.1201782226562,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 1.3039,
+      "step": 1
+    },
+    {
+      "epoch": 0.003131524008350731,
+      "grad_norm": 749.5796508789062,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 1.4164,
+      "step": 2
+    },
+    {
+      "epoch": 0.004697286012526096,
+      "grad_norm": 394.2350769042969,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 1.3367,
+      "step": 3
+    },
+    {
+      "epoch": 0.006263048016701462,
+      "grad_norm": 384.08319091796875,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.3192,
+      "step": 4
+    },
+    {
+      "epoch": 0.007828810020876827,
+      "grad_norm": 376.9790954589844,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 1.2756,
+      "step": 5
+    },
+    {
+      "epoch": 0.009394572025052192,
+      "grad_norm": 175.25262451171875,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.3992,
+      "step": 6
+    },
+    {
+      "epoch": 0.010960334029227558,
+      "grad_norm": 100.1114730834961,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 1.2348,
+      "step": 7
+    },
+    {
+      "epoch": 0.012526096033402923,
+      "grad_norm": 131.44432067871094,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.2685,
+      "step": 8
+    },
+    {
+      "epoch": 0.014091858037578288,
+      "grad_norm": 67.03341674804688,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 1.3272,
+      "step": 9
+    },
+    {
+      "epoch": 0.015657620041753653,
+      "grad_norm": 32.0869140625,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.3458,
+      "step": 10
+    },
+    {
+      "epoch": 0.01722338204592902,
+      "grad_norm": 17.701913833618164,
+      "learning_rate": 5.5e-07,
+      "loss": 1.3608,
+      "step": 11
+    },
+    {
+      "epoch": 0.018789144050104383,
+      "grad_norm": 9.164765357971191,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.2961,
+      "step": 12
+    },
+    {
+      "epoch": 0.02035490605427975,
+      "grad_norm": 4.315349578857422,
+      "learning_rate": 6.5e-07,
+      "loss": 1.2069,
+      "step": 13
+    },
+    {
+      "epoch": 0.021920668058455117,
+      "grad_norm": 6.652385234832764,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 1.2588,
+      "step": 14
+    },
+    {
+      "epoch": 0.02348643006263048,
+      "grad_norm": 3.3322548866271973,
+      "learning_rate": 7.5e-07,
+      "loss": 1.2505,
+      "step": 15
+    },
+    {
+      "epoch": 0.025052192066805846,
+      "grad_norm": 9.599331855773926,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.1343,
+      "step": 16
+    },
+    {
+      "epoch": 0.02661795407098121,
+      "grad_norm": 10.33918285369873,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 1.3334,
+      "step": 17
+    },
+    {
+      "epoch": 0.028183716075156576,
+      "grad_norm": 9.453723907470703,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.3099,
+      "step": 18
+    },
+    {
+      "epoch": 0.029749478079331943,
+      "grad_norm": 7.075769901275635,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 1.1121,
+      "step": 19
+    },
+    {
+      "epoch": 0.031315240083507306,
+      "grad_norm": 7.770864009857178,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.1836,
+      "step": 20
+    },
+    {
+      "epoch": 0.03288100208768267,
+      "grad_norm": 9.582944869995117,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 1.1939,
+      "step": 21
+    },
+    {
+      "epoch": 0.03444676409185804,
+      "grad_norm": 9.496583938598633,
+      "learning_rate": 1.1e-06,
+      "loss": 1.2395,
+      "step": 22
+    },
+    {
+      "epoch": 0.0360125260960334,
+      "grad_norm": 5.912876129150391,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 1.1769,
+      "step": 23
+    },
+    {
+      "epoch": 0.037578288100208766,
+      "grad_norm": 2.6761083602905273,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.204,
+      "step": 24
+    },
+    {
+      "epoch": 0.03914405010438413,
+      "grad_norm": 11.136463165283203,
+      "learning_rate": 1.25e-06,
+      "loss": 1.2271,
+      "step": 25
+    },
+    {
+      "epoch": 0.0407098121085595,
+      "grad_norm": 23.912893295288086,
+      "learning_rate": 1.3e-06,
+      "loss": 1.1771,
+      "step": 26
+    },
+    {
+      "epoch": 0.042275574112734866,
+      "grad_norm": 57.08906555175781,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 1.2253,
+      "step": 27
+    },
+    {
+      "epoch": 0.04384133611691023,
+      "grad_norm": 36.11408615112305,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.1014,
+      "step": 28
+    },
+    {
+      "epoch": 0.04540709812108559,
+      "grad_norm": 51.39445495605469,
+      "learning_rate": 1.45e-06,
+      "loss": 1.1978,
+      "step": 29
+    },
+    {
+      "epoch": 0.04697286012526096,
+      "grad_norm": 27.394500732421875,
+      "learning_rate": 1.5e-06,
+      "loss": 1.1155,
+      "step": 30
+    },
+    {
+      "epoch": 0.048538622129436326,
+      "grad_norm": 6.844716548919678,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 1.1798,
+      "step": 31
+    },
+    {
+      "epoch": 0.05010438413361169,
+      "grad_norm": 2.397536277770996,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.1846,
+      "step": 32
+    },
+    {
+      "epoch": 0.05167014613778706,
+      "grad_norm": 7.821156024932861,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 1.1312,
+      "step": 33
+    },
+    {
+      "epoch": 0.05323590814196242,
+      "grad_norm": 5.455922603607178,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.2153,
+      "step": 34
+    },
+    {
+      "epoch": 0.054801670146137786,
+      "grad_norm": 5.377617835998535,
+      "learning_rate": 1.75e-06,
+      "loss": 1.1292,
+      "step": 35
+    },
+    {
+      "epoch": 0.05636743215031315,
+      "grad_norm": 7.172031879425049,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.1459,
+      "step": 36
+    },
+    {
+      "epoch": 0.05793319415448852,
+      "grad_norm": 2.1464905738830566,
+      "learning_rate": 1.85e-06,
+      "loss": 1.1649,
+      "step": 37
+    },
+    {
+      "epoch": 0.059498956158663886,
+      "grad_norm": 2.7782726287841797,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.159,
+      "step": 38
+    },
+    {
+      "epoch": 0.061064718162839246,
+      "grad_norm": 20.818374633789062,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 1.0174,
+      "step": 39
+    },
+    {
+      "epoch": 0.06263048016701461,
+      "grad_norm": 22.776456832885742,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.1425,
+      "step": 40
+    },
+    {
+      "epoch": 0.06419624217118998,
+      "grad_norm": 33.601287841796875,
+      "learning_rate": 2.05e-06,
+      "loss": 1.0498,
+      "step": 41
+    },
+    {
+      "epoch": 0.06576200417536535,
+      "grad_norm": 12.884684562683105,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 1.1231,
+      "step": 42
+    },
+    {
+      "epoch": 0.06732776617954071,
+      "grad_norm": 1.951470971107483,
+      "learning_rate": 2.15e-06,
+      "loss": 1.1256,
+      "step": 43
+    },
+    {
+      "epoch": 0.06889352818371608,
+      "grad_norm": 2.584127426147461,
+      "learning_rate": 2.2e-06,
+      "loss": 1.1242,
+      "step": 44
+    },
+    {
+      "epoch": 0.07045929018789145,
+      "grad_norm": 6.982508659362793,
+      "learning_rate": 2.25e-06,
+      "loss": 1.0769,
+      "step": 45
+    },
+    {
+      "epoch": 0.0720250521920668,
+      "grad_norm": 7.940657138824463,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 1.0516,
+      "step": 46
+    },
+    {
+      "epoch": 0.07359081419624217,
+      "grad_norm": 5.456567287445068,
+      "learning_rate": 2.35e-06,
+      "loss": 0.9894,
+      "step": 47
+    },
+    {
+      "epoch": 0.07515657620041753,
+      "grad_norm": 4.623636722564697,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.9912,
+      "step": 48
+    },
+    {
+      "epoch": 0.0767223382045929,
+      "grad_norm": 2.182708978652954,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 1.1303,
+      "step": 49
+    },
+    {
+      "epoch": 0.07828810020876827,
+      "grad_norm": 2.388091802597046,
+      "learning_rate": 2.5e-06,
+      "loss": 1.0725,
+      "step": 50
+    },
+    {
+      "epoch": 0.07985386221294363,
+      "grad_norm": 2.2816522121429443,
+      "learning_rate": 2.55e-06,
+      "loss": 1.0895,
+      "step": 51
+    },
+    {
+      "epoch": 0.081419624217119,
+      "grad_norm": 3.8391623497009277,
+      "learning_rate": 2.6e-06,
+      "loss": 1.0587,
+      "step": 52
+    },
+    {
+      "epoch": 0.08298538622129437,
+      "grad_norm": 3.700389862060547,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.9795,
+      "step": 53
+    },
+    {
+      "epoch": 0.08455114822546973,
+      "grad_norm": 3.180649995803833,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 1.0434,
+      "step": 54
+    },
+    {
+      "epoch": 0.0861169102296451,
+      "grad_norm": 3.790774345397949,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.9652,
+      "step": 55
+    },
+    {
+      "epoch": 0.08768267223382047,
+      "grad_norm": 1.8411438465118408,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.947,
+      "step": 56
+    },
+    {
+      "epoch": 0.08924843423799582,
+      "grad_norm": 2.100008010864258,
+      "learning_rate": 2.85e-06,
+      "loss": 1.0129,
+      "step": 57
+    },
+    {
+      "epoch": 0.09081419624217119,
+      "grad_norm": 1.103857159614563,
+      "learning_rate": 2.9e-06,
+      "loss": 0.9993,
+      "step": 58
+    },
+    {
+      "epoch": 0.09237995824634655,
+      "grad_norm": 1.232283353805542,
+      "learning_rate": 2.95e-06,
+      "loss": 0.9988,
+      "step": 59
+    },
+    {
+      "epoch": 0.09394572025052192,
+      "grad_norm": 2.4731297492980957,
+      "learning_rate": 3e-06,
+      "loss": 0.9873,
+      "step": 60
+    },
+    {
+      "epoch": 0.09551148225469729,
+      "grad_norm": 1.1080410480499268,
+      "learning_rate": 3.05e-06,
+      "loss": 1.0657,
+      "step": 61
+    },
+    {
+      "epoch": 0.09707724425887265,
+      "grad_norm": 1.1155121326446533,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.974,
+      "step": 62
+    },
+    {
+      "epoch": 0.09864300626304802,
+      "grad_norm": 1.763231873512268,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 1.0215,
+      "step": 63
+    },
+    {
+      "epoch": 0.10020876826722339,
+      "grad_norm": 1.1961321830749512,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.9955,
+      "step": 64
+    },
+    {
+      "epoch": 0.10177453027139875,
+      "grad_norm": 1.377040982246399,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.9834,
+      "step": 65
+    },
+    {
+      "epoch": 0.10334029227557412,
+      "grad_norm": 1.373260498046875,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.984,
+      "step": 66
+    },
+    {
+      "epoch": 0.10490605427974947,
+      "grad_norm": 2.034064292907715,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.973,
+      "step": 67
+    },
+    {
+      "epoch": 0.10647181628392484,
+      "grad_norm": 1.41798734664917,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.925,
+      "step": 68
+    },
+    {
+      "epoch": 0.1080375782881002,
+      "grad_norm": 1.7425439357757568,
+      "learning_rate": 3.45e-06,
+      "loss": 0.9884,
+      "step": 69
+    },
+    {
+      "epoch": 0.10960334029227557,
+      "grad_norm": 2.369657278060913,
+      "learning_rate": 3.5e-06,
+      "loss": 1.0048,
+      "step": 70
+    },
+    {
+      "epoch": 0.11116910229645094,
+      "grad_norm": 1.4699137210845947,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.9155,
+      "step": 71
+    },
+    {
+      "epoch": 0.1127348643006263,
+      "grad_norm": 1.1473263502120972,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.8445,
+      "step": 72
+    },
+    {
+      "epoch": 0.11430062630480167,
+      "grad_norm": 2.315568685531616,
+      "learning_rate": 3.65e-06,
+      "loss": 0.977,
+      "step": 73
+    },
+    {
+      "epoch": 0.11586638830897704,
+      "grad_norm": 1.8554916381835938,
+      "learning_rate": 3.7e-06,
+      "loss": 0.841,
+      "step": 74
+    },
+    {
+      "epoch": 0.1174321503131524,
+      "grad_norm": 1.1517397165298462,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.9478,
+      "step": 75
+    },
+    {
+      "epoch": 0.11899791231732777,
+      "grad_norm": 2.799316883087158,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.9873,
+      "step": 76
+    },
+    {
+      "epoch": 0.12056367432150313,
+      "grad_norm": 1.11930251121521,
+      "learning_rate": 3.85e-06,
+      "loss": 0.9214,
+      "step": 77
+    },
+    {
+      "epoch": 0.12212943632567849,
+      "grad_norm": 1.0082950592041016,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.9115,
+      "step": 78
+    },
+    {
+      "epoch": 0.12369519832985386,
+      "grad_norm": 1.5534541606903076,
+      "learning_rate": 3.95e-06,
+      "loss": 1.04,
+      "step": 79
+    },
+    {
+      "epoch": 0.12526096033402923,
+      "grad_norm": 1.3469066619873047,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.9091,
+      "step": 80
+    },
+    {
+      "epoch": 0.1268267223382046,
+      "grad_norm": 1.1149636507034302,
+      "learning_rate": 4.05e-06,
+      "loss": 0.8829,
+      "step": 81
+    },
+    {
+      "epoch": 0.12839248434237996,
+      "grad_norm": 2.5933806896209717,
+      "learning_rate": 4.1e-06,
+      "loss": 0.9343,
+      "step": 82
+    },
+    {
+      "epoch": 0.1299582463465553,
+      "grad_norm": 1.980901837348938,
+      "learning_rate": 4.15e-06,
+      "loss": 0.9028,
+      "step": 83
+    },
+    {
+      "epoch": 0.1315240083507307,
+      "grad_norm": 0.8476768732070923,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.9038,
+      "step": 84
+    },
+    {
+      "epoch": 0.13308977035490605,
+      "grad_norm": 1.1239815950393677,
+      "learning_rate": 4.25e-06,
+      "loss": 1.0359,
+      "step": 85
+    },
+    {
+      "epoch": 0.13465553235908143,
+      "grad_norm": 1.0159156322479248,
+      "learning_rate": 4.3e-06,
+      "loss": 0.8705,
+      "step": 86
+    },
+    {
+      "epoch": 0.13622129436325678,
+      "grad_norm": 2.180856466293335,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.9834,
+      "step": 87
+    },
+    {
+      "epoch": 0.13778705636743216,
+      "grad_norm": 1.5725276470184326,
+      "learning_rate": 4.4e-06,
+      "loss": 0.9449,
+      "step": 88
+    },
+    {
+      "epoch": 0.1393528183716075,
+      "grad_norm": 1.5560917854309082,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 1.0109,
+      "step": 89
+    },
+    {
+      "epoch": 0.1409185803757829,
+      "grad_norm": 1.1096960306167603,
+      "learning_rate": 4.5e-06,
+      "loss": 0.9081,
+      "step": 90
+    },
+    {
+      "epoch": 0.14248434237995825,
+      "grad_norm": 1.2525160312652588,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.9907,
+      "step": 91
+    },
+    {
+      "epoch": 0.1440501043841336,
+      "grad_norm": 0.9198770523071289,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.9266,
+      "step": 92
+    },
+    {
+      "epoch": 0.14561586638830898,
+      "grad_norm": 0.992057740688324,
+      "learning_rate": 4.65e-06,
+      "loss": 0.9556,
+      "step": 93
+    },
+    {
+      "epoch": 0.14718162839248433,
+      "grad_norm": 1.00990891456604,
+      "learning_rate": 4.7e-06,
+      "loss": 0.9712,
+      "step": 94
+    },
+    {
+      "epoch": 0.1487473903966597,
+      "grad_norm": 1.1176848411560059,
+      "learning_rate": 4.75e-06,
+      "loss": 0.9604,
+      "step": 95
+    },
+    {
+      "epoch": 0.15031315240083507,
+      "grad_norm": 0.9323613047599792,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.9708,
+      "step": 96
+    },
+    {
+      "epoch": 0.15187891440501045,
+      "grad_norm": 0.9800642728805542,
+      "learning_rate": 4.85e-06,
+      "loss": 0.9868,
+      "step": 97
+    },
+    {
+      "epoch": 0.1534446764091858,
+      "grad_norm": 1.487841010093689,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.8679,
+      "step": 98
+    },
+    {
+      "epoch": 0.15501043841336118,
+      "grad_norm": 0.95517897605896,
+      "learning_rate": 4.95e-06,
+      "loss": 0.7949,
+      "step": 99
+    },
+    {
+      "epoch": 0.15657620041753653,
+      "grad_norm": 0.9686336517333984,
+      "learning_rate": 5e-06,
+      "loss": 0.8655,
+      "step": 100
+    },
+    {
+      "epoch": 0.1581419624217119,
+      "grad_norm": 1.9481035470962524,
+      "learning_rate": 4.999999112316954e-06,
+      "loss": 0.9328,
+      "step": 101
+    },
+    {
+      "epoch": 0.15970772442588727,
+      "grad_norm": 1.272631287574768,
+      "learning_rate": 4.9999964492684465e-06,
+      "loss": 0.7869,
+      "step": 102
+    },
+    {
+      "epoch": 0.16127348643006262,
+      "grad_norm": 1.1117815971374512,
+      "learning_rate": 4.999992010856368e-06,
+      "loss": 0.949,
+      "step": 103
+    },
+    {
+      "epoch": 0.162839248434238,
+      "grad_norm": 0.9681039452552795,
+      "learning_rate": 4.999985797083871e-06,
+      "loss": 1.0396,
+      "step": 104
+    },
+    {
+      "epoch": 0.16440501043841335,
+      "grad_norm": 0.9797878265380859,
+      "learning_rate": 4.999977807955368e-06,
+      "loss": 0.9393,
+      "step": 105
+    },
+    {
+      "epoch": 0.16597077244258873,
+      "grad_norm": 1.0239158868789673,
+      "learning_rate": 4.999968043476532e-06,
+      "loss": 0.8766,
+      "step": 106
+    },
+    {
+      "epoch": 0.16753653444676408,
+      "grad_norm": 0.9314751029014587,
+      "learning_rate": 4.999956503654299e-06,
+      "loss": 0.8545,
+      "step": 107
+    },
+    {
+      "epoch": 0.16910229645093947,
+      "grad_norm": 1.1627179384231567,
+      "learning_rate": 4.999943188496862e-06,
+      "loss": 0.9241,
+      "step": 108
+    },
+    {
+      "epoch": 0.17066805845511482,
+      "grad_norm": 1.022752285003662,
+      "learning_rate": 4.9999280980136765e-06,
+      "loss": 0.9662,
+      "step": 109
+    },
+    {
+      "epoch": 0.1722338204592902,
+      "grad_norm": 1.0513653755187988,
+      "learning_rate": 4.99991123221546e-06,
+      "loss": 0.9054,
+      "step": 110
+    },
+    {
+      "epoch": 0.17379958246346555,
+      "grad_norm": 0.9488444328308105,
+      "learning_rate": 4.99989259111419e-06,
+      "loss": 0.8921,
+      "step": 111
+    },
+    {
+      "epoch": 0.17536534446764093,
+      "grad_norm": 0.9934154748916626,
+      "learning_rate": 4.999872174723104e-06,
+      "loss": 0.9848,
+      "step": 112
+    },
+    {
+      "epoch": 0.17693110647181629,
+      "grad_norm": 1.2035515308380127,
+      "learning_rate": 4.9998499830567e-06,
+      "loss": 0.9833,
+      "step": 113
+    },
+    {
+      "epoch": 0.17849686847599164,
+      "grad_norm": 1.027538537979126,
+      "learning_rate": 4.999826016130739e-06,
+      "loss": 1.0089,
+      "step": 114
+    },
+    {
+      "epoch": 0.18006263048016702,
+      "grad_norm": 1.0635300874710083,
+      "learning_rate": 4.999800273962238e-06,
+      "loss": 0.8863,
+      "step": 115
+    },
+    {
+      "epoch": 0.18162839248434237,
+      "grad_norm": 0.9978796243667603,
+      "learning_rate": 4.999772756569482e-06,
+      "loss": 0.899,
+      "step": 116
+    },
+    {
+      "epoch": 0.18319415448851775,
+      "grad_norm": 0.9971838593482971,
+      "learning_rate": 4.999743463972008e-06,
+      "loss": 0.9515,
+      "step": 117
+    },
+    {
+      "epoch": 0.1847599164926931,
+      "grad_norm": 1.2436699867248535,
+      "learning_rate": 4.999712396190622e-06,
+      "loss": 0.9101,
+      "step": 118
+    },
+    {
+      "epoch": 0.18632567849686849,
+      "grad_norm": 0.9573557376861572,
+      "learning_rate": 4.999679553247383e-06,
+      "loss": 0.8817,
+      "step": 119
+    },
+    {
+      "epoch": 0.18789144050104384,
+      "grad_norm": 0.9105668067932129,
+      "learning_rate": 4.9996449351656165e-06,
+      "loss": 1.0077,
+      "step": 120
+    },
+    {
+      "epoch": 0.18945720250521922,
+      "grad_norm": 0.8741400241851807,
+      "learning_rate": 4.999608541969905e-06,
+      "loss": 0.912,
+      "step": 121
+    },
+    {
+      "epoch": 0.19102296450939457,
+      "grad_norm": 1.1172688007354736,
+      "learning_rate": 4.999570373686095e-06,
+      "loss": 0.8062,
+      "step": 122
+    },
+    {
+      "epoch": 0.19258872651356992,
+      "grad_norm": 0.940539538860321,
+      "learning_rate": 4.9995304303412905e-06,
+      "loss": 0.8944,
+      "step": 123
+    },
+    {
+      "epoch": 0.1941544885177453,
+      "grad_norm": 2.1082208156585693,
+      "learning_rate": 4.999488711963857e-06,
+      "loss": 0.9668,
+      "step": 124
+    },
+    {
+      "epoch": 0.19572025052192066,
+      "grad_norm": 1.0358761548995972,
+      "learning_rate": 4.99944521858342e-06,
+      "loss": 0.9032,
+      "step": 125
+    },
+    {
+      "epoch": 0.19728601252609604,
+      "grad_norm": 2.0804238319396973,
+      "learning_rate": 4.999399950230867e-06,
+      "loss": 0.9819,
+      "step": 126
+    },
+    {
+      "epoch": 0.1988517745302714,
+      "grad_norm": 1.0986278057098389,
+      "learning_rate": 4.9993529069383465e-06,
+      "loss": 0.9034,
+      "step": 127
+    },
+    {
+      "epoch": 0.20041753653444677,
+      "grad_norm": 1.2374722957611084,
+      "learning_rate": 4.999304088739263e-06,
+      "loss": 0.9255,
+      "step": 128
+    },
+    {
+      "epoch": 0.20198329853862212,
+      "grad_norm": 1.0443146228790283,
+      "learning_rate": 4.999253495668287e-06,
+      "loss": 0.9352,
+      "step": 129
+    },
+    {
+      "epoch": 0.2035490605427975,
+      "grad_norm": 1.009800910949707,
+      "learning_rate": 4.999201127761346e-06,
+      "loss": 0.8324,
+      "step": 130
+    },
+    {
+      "epoch": 0.20511482254697286,
+      "grad_norm": 0.892861008644104,
+      "learning_rate": 4.9991469850556295e-06,
+      "loss": 0.886,
+      "step": 131
+    },
+    {
+      "epoch": 0.20668058455114824,
+      "grad_norm": 1.1675922870635986,
+      "learning_rate": 4.999091067589587e-06,
+      "loss": 0.967,
+      "step": 132
+    },
+    {
+      "epoch": 0.2082463465553236,
+      "grad_norm": 1.0169954299926758,
+      "learning_rate": 4.999033375402926e-06,
+      "loss": 0.8498,
+      "step": 133
+    },
+    {
+      "epoch": 0.20981210855949894,
+      "grad_norm": 1.0006227493286133,
+      "learning_rate": 4.99897390853662e-06,
+      "loss": 0.8481,
+      "step": 134
+    },
+    {
+      "epoch": 0.21137787056367432,
+      "grad_norm": 0.9049347639083862,
+      "learning_rate": 4.9989126670328945e-06,
+      "loss": 0.9186,
+      "step": 135
+    },
+    {
+      "epoch": 0.21294363256784968,
+      "grad_norm": 1.2151603698730469,
+      "learning_rate": 4.9988496509352444e-06,
+      "loss": 0.9067,
+      "step": 136
+    },
+    {
+      "epoch": 0.21450939457202506,
+      "grad_norm": 1.0112006664276123,
+      "learning_rate": 4.998784860288417e-06,
+      "loss": 0.8843,
+      "step": 137
+    },
+    {
+      "epoch": 0.2160751565762004,
+      "grad_norm": 0.9316662549972534,
+      "learning_rate": 4.9987182951384264e-06,
+      "loss": 0.8635,
+      "step": 138
+    },
+    {
+      "epoch": 0.2176409185803758,
+      "grad_norm": 0.942743718624115,
+      "learning_rate": 4.9986499555325406e-06,
+      "loss": 0.8481,
+      "step": 139
+    },
+    {
+      "epoch": 0.21920668058455114,
+      "grad_norm": 1.0298817157745361,
+      "learning_rate": 4.998579841519292e-06,
+      "loss": 0.865,
+      "step": 140
+    },
+    {
+      "epoch": 0.22077244258872653,
+      "grad_norm": 1.1556729078292847,
+      "learning_rate": 4.998507953148472e-06,
+      "loss": 0.8731,
+      "step": 141
+    },
+    {
+      "epoch": 0.22233820459290188,
+      "grad_norm": 0.997958779335022,
+      "learning_rate": 4.9984342904711315e-06,
+      "loss": 0.7949,
+      "step": 142
+    },
+    {
+      "epoch": 0.22390396659707723,
+      "grad_norm": 0.9565998315811157,
+      "learning_rate": 4.998358853539582e-06,
+      "loss": 0.8875,
+      "step": 143
+    },
+    {
+      "epoch": 0.2254697286012526,
+      "grad_norm": 0.9667196273803711,
+      "learning_rate": 4.998281642407394e-06,
+      "loss": 0.8537,
+      "step": 144
+    },
+    {
+      "epoch": 0.22703549060542796,
+      "grad_norm": 0.9701303839683533,
+      "learning_rate": 4.998202657129401e-06,
+      "loss": 0.873,
+      "step": 145
+    },
+    {
+      "epoch": 0.22860125260960334,
+      "grad_norm": 0.9936159253120422,
+      "learning_rate": 4.998121897761692e-06,
+      "loss": 0.8915,
+      "step": 146
+    },
+    {
+      "epoch": 0.2301670146137787,
+      "grad_norm": 1.5054311752319336,
+      "learning_rate": 4.998039364361617e-06,
+      "loss": 0.9944,
+      "step": 147
+    },
+    {
+      "epoch": 0.23173277661795408,
+      "grad_norm": 1.5895891189575195,
+      "learning_rate": 4.99795505698779e-06,
+      "loss": 0.9282,
+      "step": 148
+    },
+    {
+      "epoch": 0.23329853862212943,
+      "grad_norm": 1.0141469240188599,
+      "learning_rate": 4.9978689757000785e-06,
+      "loss": 0.9166,
+      "step": 149
+    },
+    {
+      "epoch": 0.2348643006263048,
+      "grad_norm": 1.1720311641693115,
+      "learning_rate": 4.997781120559615e-06,
+      "loss": 0.9404,
+      "step": 150
+    },
+    {
+      "epoch": 0.23643006263048016,
+      "grad_norm": 0.9536433815956116,
+      "learning_rate": 4.997691491628789e-06,
+      "loss": 0.9339,
+      "step": 151
+    },
+    {
+      "epoch": 0.23799582463465555,
+      "grad_norm": 0.9407872557640076,
+      "learning_rate": 4.997600088971249e-06,
+      "loss": 0.9019,
+      "step": 152
+    },
+    {
+      "epoch": 0.2395615866388309,
+      "grad_norm": 0.9202889204025269,
+      "learning_rate": 4.997506912651906e-06,
+      "loss": 0.9598,
+      "step": 153
+    },
+    {
+      "epoch": 0.24112734864300625,
+      "grad_norm": 1.0630513429641724,
+      "learning_rate": 4.997411962736927e-06,
+      "loss": 0.8554,
+      "step": 154
+    },
+    {
+      "epoch": 0.24269311064718163,
+      "grad_norm": 0.8378801941871643,
+      "learning_rate": 4.997315239293742e-06,
+      "loss": 0.8583,
+      "step": 155
+    },
+    {
+      "epoch": 0.24425887265135698,
+      "grad_norm": 0.9442815184593201,
+      "learning_rate": 4.997216742391038e-06,
+      "loss": 0.8518,
+      "step": 156
+    },
+    {
+      "epoch": 0.24582463465553236,
+      "grad_norm": 0.9273308515548706,
+      "learning_rate": 4.997116472098763e-06,
+      "loss": 0.8942,
+      "step": 157
+    },
+    {
+      "epoch": 0.24739039665970772,
+      "grad_norm": 0.9028997421264648,
+      "learning_rate": 4.997014428488121e-06,
+      "loss": 0.8592,
+      "step": 158
+    },
+    {
+      "epoch": 0.2489561586638831,
+      "grad_norm": 0.9877567887306213,
+      "learning_rate": 4.996910611631582e-06,
+      "loss": 0.9316,
+      "step": 159
+    },
+    {
+      "epoch": 0.25052192066805845,
+      "grad_norm": 0.961592435836792,
+      "learning_rate": 4.996805021602868e-06,
+      "loss": 0.8875,
+      "step": 160
+    },
+    {
+      "epoch": 0.2520876826722338,
+      "grad_norm": 0.9923007488250732,
+      "learning_rate": 4.996697658476965e-06,
+      "loss": 0.8554,
+      "step": 161
+    },
+    {
+      "epoch": 0.2536534446764092,
+      "grad_norm": 0.956763744354248,
+      "learning_rate": 4.996588522330114e-06,
+      "loss": 0.9381,
+      "step": 162
+    },
+    {
+      "epoch": 0.25521920668058456,
+      "grad_norm": 0.919992208480835,
+      "learning_rate": 4.996477613239822e-06,
+      "loss": 0.9016,
+      "step": 163
+    },
+    {
+      "epoch": 0.2567849686847599,
+      "grad_norm": 0.9655581712722778,
+      "learning_rate": 4.996364931284847e-06,
+      "loss": 0.896,
+      "step": 164
+    },
+    {
+      "epoch": 0.25835073068893527,
+      "grad_norm": 0.9159606099128723,
+      "learning_rate": 4.996250476545212e-06,
+      "loss": 0.9475,
+      "step": 165
+    },
+    {
+      "epoch": 0.2599164926931106,
+      "grad_norm": 1.1147626638412476,
+      "learning_rate": 4.996134249102195e-06,
+      "loss": 0.8793,
+      "step": 166
+    },
+    {
+      "epoch": 0.26148225469728603,
+      "grad_norm": 1.112069010734558,
+      "learning_rate": 4.996016249038334e-06,
+      "loss": 0.8765,
+      "step": 167
+    },
+    {
+      "epoch": 0.2630480167014614,
+      "grad_norm": 1.0366944074630737,
+      "learning_rate": 4.995896476437428e-06,
+      "loss": 0.9356,
+      "step": 168
+    },
+    {
+      "epoch": 0.26461377870563674,
+      "grad_norm": 0.9494239091873169,
+      "learning_rate": 4.995774931384534e-06,
+      "loss": 0.8246,
+      "step": 169
+    },
+    {
+      "epoch": 0.2661795407098121,
+      "grad_norm": 0.9819200038909912,
+      "learning_rate": 4.995651613965964e-06,
+      "loss": 0.9403,
+      "step": 170
+    },
+    {
+      "epoch": 0.2677453027139875,
+      "grad_norm": 1.7329354286193848,
+      "learning_rate": 4.995526524269293e-06,
+      "loss": 0.9091,
+      "step": 171
+    },
+    {
+      "epoch": 0.26931106471816285,
+      "grad_norm": 1.1415139436721802,
+      "learning_rate": 4.995399662383352e-06,
+      "loss": 0.9106,
+      "step": 172
+    },
+    {
+      "epoch": 0.2708768267223382,
+      "grad_norm": 0.9471906423568726,
+      "learning_rate": 4.995271028398233e-06,
+      "loss": 0.8858,
+      "step": 173
+    },
+    {
+      "epoch": 0.27244258872651356,
+      "grad_norm": 1.096053123474121,
+      "learning_rate": 4.995140622405284e-06,
+      "loss": 0.7502,
+      "step": 174
+    },
+    {
+      "epoch": 0.2740083507306889,
+      "grad_norm": 1.285410761833191,
+      "learning_rate": 4.995008444497113e-06,
+      "loss": 0.9288,
+      "step": 175
+    },
+    {
+      "epoch": 0.2755741127348643,
+      "grad_norm": 0.9714595079421997,
+      "learning_rate": 4.994874494767585e-06,
+      "loss": 0.8824,
+      "step": 176
+    },
+    {
+      "epoch": 0.27713987473903967,
+      "grad_norm": 0.8399011492729187,
+      "learning_rate": 4.994738773311824e-06,
+      "loss": 0.8703,
+      "step": 177
+    },
+    {
+      "epoch": 0.278705636743215,
+      "grad_norm": 0.9454224109649658,
+      "learning_rate": 4.994601280226212e-06,
+      "loss": 0.9627,
+      "step": 178
+    },
+    {
+      "epoch": 0.2802713987473904,
+      "grad_norm": 0.8983661532402039,
+      "learning_rate": 4.99446201560839e-06,
+      "loss": 0.9086,
+      "step": 179
+    },
+    {
+      "epoch": 0.2818371607515658,
+      "grad_norm": 0.9814153909683228,
+      "learning_rate": 4.994320979557256e-06,
+      "loss": 0.908,
+      "step": 180
+    },
+    {
+      "epoch": 0.28340292275574114,
+      "grad_norm": 0.8405355215072632,
+      "learning_rate": 4.9941781721729645e-06,
+      "loss": 0.8414,
+      "step": 181
+    },
+    {
+      "epoch": 0.2849686847599165,
+      "grad_norm": 0.950950026512146,
+      "learning_rate": 4.994033593556933e-06,
+      "loss": 0.9091,
+      "step": 182
+    },
+    {
+      "epoch": 0.28653444676409184,
+      "grad_norm": 1.1019747257232666,
+      "learning_rate": 4.99388724381183e-06,
+      "loss": 0.8305,
+      "step": 183
+    },
+    {
+      "epoch": 0.2881002087682672,
+      "grad_norm": 0.9339331984519958,
+      "learning_rate": 4.9937391230415875e-06,
+      "loss": 0.8542,
+      "step": 184
+    },
+    {
+      "epoch": 0.2896659707724426,
+      "grad_norm": 0.8813616633415222,
+      "learning_rate": 4.993589231351394e-06,
+      "loss": 0.848,
+      "step": 185
+    },
+    {
+      "epoch": 0.29123173277661796,
+      "grad_norm": 0.8916418552398682,
+      "learning_rate": 4.9934375688476905e-06,
+      "loss": 0.8032,
+      "step": 186
+    },
+    {
+      "epoch": 0.2927974947807933,
+      "grad_norm": 0.9891005754470825,
+      "learning_rate": 4.993284135638182e-06,
+      "loss": 0.7937,
+      "step": 187
+    },
+    {
+      "epoch": 0.29436325678496866,
+      "grad_norm": 0.9842082262039185,
+      "learning_rate": 4.99312893183183e-06,
+      "loss": 0.9532,
+      "step": 188
+    },
+    {
+      "epoch": 0.29592901878914407,
+      "grad_norm": 0.993478536605835,
+      "learning_rate": 4.99297195753885e-06,
+      "loss": 0.9406,
+      "step": 189
+    },
+    {
+      "epoch": 0.2974947807933194,
+      "grad_norm": 1.0654981136322021,
+      "learning_rate": 4.992813212870718e-06,
+      "loss": 0.907,
+      "step": 190
+    },
+    {
+      "epoch": 0.2990605427974948,
+      "grad_norm": 0.8937190771102905,
+      "learning_rate": 4.992652697940164e-06,
+      "loss": 0.9307,
+      "step": 191
+    },
+    {
+      "epoch": 0.30062630480167013,
+      "grad_norm": 0.9006052613258362,
+      "learning_rate": 4.992490412861178e-06,
+      "loss": 0.8808,
+      "step": 192
+    },
+    {
+      "epoch": 0.30219206680584554,
+      "grad_norm": 0.9601522088050842,
+      "learning_rate": 4.992326357749007e-06,
+      "loss": 0.8495,
+      "step": 193
+    },
+    {
+      "epoch": 0.3037578288100209,
+      "grad_norm": 0.9814243316650391,
+      "learning_rate": 4.992160532720153e-06,
+      "loss": 0.8204,
+      "step": 194
+    },
+    {
+      "epoch": 0.30532359081419624,
+      "grad_norm": 0.9660456776618958,
+      "learning_rate": 4.991992937892377e-06,
+      "loss": 0.8209,
+      "step": 195
+    },
+    {
+      "epoch": 0.3068893528183716,
+      "grad_norm": 0.8504270315170288,
+      "learning_rate": 4.991823573384695e-06,
+      "loss": 0.8415,
+      "step": 196
+    },
+    {
+      "epoch": 0.30845511482254695,
+      "grad_norm": 0.8733190894126892,
+      "learning_rate": 4.991652439317382e-06,
+      "loss": 0.8753,
+      "step": 197
+    },
+    {
+      "epoch": 0.31002087682672236,
+      "grad_norm": 0.9761794805526733,
+      "learning_rate": 4.991479535811967e-06,
+      "loss": 0.8923,
+      "step": 198
+    },
+    {
+      "epoch": 0.3115866388308977,
+      "grad_norm": 1.0400285720825195,
+      "learning_rate": 4.991304862991238e-06,
+      "loss": 0.899,
+      "step": 199
+    },
+    {
+      "epoch": 0.31315240083507306,
+      "grad_norm": 0.8561192750930786,
+      "learning_rate": 4.991128420979237e-06,
+      "loss": 0.9117,
+      "step": 200
+    },
+    {
+      "epoch": 0.3147181628392484,
+      "grad_norm": 0.8839634656906128,
+      "learning_rate": 4.990950209901263e-06,
+      "loss": 0.9022,
+      "step": 201
+    },
+    {
+      "epoch": 0.3162839248434238,
+      "grad_norm": 0.9569464921951294,
+      "learning_rate": 4.990770229883873e-06,
+      "loss": 0.9013,
+      "step": 202
+    },
+    {
+      "epoch": 0.3178496868475992,
+      "grad_norm": 1.0600861310958862,
+      "learning_rate": 4.990588481054881e-06,
+      "loss": 0.8387,
+      "step": 203
+    },
+    {
+      "epoch": 0.31941544885177453,
+      "grad_norm": 0.9492761492729187,
+      "learning_rate": 4.990404963543352e-06,
+      "loss": 0.8521,
+      "step": 204
+    },
+    {
+      "epoch": 0.3209812108559499,
+      "grad_norm": 0.9534817934036255,
+      "learning_rate": 4.990219677479612e-06,
+      "loss": 0.8009,
+      "step": 205
+    },
+    {
+      "epoch": 0.32254697286012524,
+      "grad_norm": 0.9918960928916931,
+      "learning_rate": 4.990032622995242e-06,
+      "loss": 0.8215,
+      "step": 206
+    },
+    {
+      "epoch": 0.32411273486430064,
+      "grad_norm": 0.8751630783081055,
+      "learning_rate": 4.989843800223077e-06,
+      "loss": 0.8744,
+      "step": 207
+    },
+    {
+      "epoch": 0.325678496868476,
+      "grad_norm": 1.0034785270690918,
+      "learning_rate": 4.9896532092972085e-06,
+      "loss": 0.884,
+      "step": 208
+    },
+    {
+      "epoch": 0.32724425887265135,
+      "grad_norm": 1.2693432569503784,
+      "learning_rate": 4.989460850352984e-06,
+      "loss": 0.9689,
+      "step": 209
+    },
+    {
+      "epoch": 0.3288100208768267,
+      "grad_norm": 0.9923811554908752,
+      "learning_rate": 4.989266723527007e-06,
+      "loss": 0.9622,
+      "step": 210
+    },
+    {
+      "epoch": 0.3303757828810021,
+      "grad_norm": 0.9439074397087097,
+      "learning_rate": 4.989070828957137e-06,
+      "loss": 0.9155,
+      "step": 211
+    },
+    {
+      "epoch": 0.33194154488517746,
+      "grad_norm": 1.2564488649368286,
+      "learning_rate": 4.988873166782485e-06,
+      "loss": 0.8712,
+      "step": 212
+    },
+    {
+      "epoch": 0.3335073068893528,
+      "grad_norm": 1.1441881656646729,
+      "learning_rate": 4.988673737143423e-06,
+      "loss": 0.8895,
+      "step": 213
+    },
+    {
+      "epoch": 0.33507306889352817,
+      "grad_norm": 0.972155749797821,
+      "learning_rate": 4.988472540181574e-06,
+      "loss": 0.9158,
+      "step": 214
+    },
+    {
+      "epoch": 0.3366388308977035,
+      "grad_norm": 0.9536120295524597,
+      "learning_rate": 4.988269576039817e-06,
+      "loss": 0.7963,
+      "step": 215
+    },
+    {
+      "epoch": 0.33820459290187893,
+      "grad_norm": 0.8956594467163086,
+      "learning_rate": 4.988064844862288e-06,
+      "loss": 0.8947,
+      "step": 216
+    },
+    {
+      "epoch": 0.3397703549060543,
+      "grad_norm": 0.9503691792488098,
+      "learning_rate": 4.987858346794373e-06,
+      "loss": 0.7857,
+      "step": 217
+    },
+    {
+      "epoch": 0.34133611691022964,
+      "grad_norm": 0.9471456408500671,
+      "learning_rate": 4.987650081982719e-06,
+      "loss": 0.9118,
+      "step": 218
+    },
+    {
+      "epoch": 0.342901878914405,
+      "grad_norm": 0.9485713839530945,
+      "learning_rate": 4.987440050575222e-06,
+      "loss": 0.8188,
+      "step": 219
+    },
+    {
+      "epoch": 0.3444676409185804,
+      "grad_norm": 1.0775927305221558,
+      "learning_rate": 4.987228252721037e-06,
+      "loss": 0.9469,
+      "step": 220
+    },
+    {
+      "epoch": 0.34603340292275575,
+      "grad_norm": 0.9173988103866577,
+      "learning_rate": 4.9870146885705705e-06,
+      "loss": 0.8566,
+      "step": 221
+    },
+    {
+      "epoch": 0.3475991649269311,
+      "grad_norm": 1.008735179901123,
+      "learning_rate": 4.986799358275485e-06,
+      "loss": 0.8637,
+      "step": 222
+    },
+    {
+      "epoch": 0.34916492693110646,
+      "grad_norm": 0.8183198571205139,
+      "learning_rate": 4.986582261988695e-06,
+      "loss": 0.9321,
+      "step": 223
+    },
+    {
+      "epoch": 0.35073068893528186,
+      "grad_norm": 0.8951883316040039,
+      "learning_rate": 4.986363399864372e-06,
+      "loss": 0.8015,
+      "step": 224
+    },
+    {
+      "epoch": 0.3522964509394572,
+      "grad_norm": 0.9275704622268677,
+      "learning_rate": 4.986142772057941e-06,
+      "loss": 0.8242,
+      "step": 225
+    },
+    {
+      "epoch": 0.35386221294363257,
+      "grad_norm": 0.9548516273498535,
+      "learning_rate": 4.985920378726077e-06,
+      "loss": 0.7889,
+      "step": 226
+    },
+    {
+      "epoch": 0.3554279749478079,
+      "grad_norm": 1.1203585863113403,
+      "learning_rate": 4.985696220026714e-06,
+      "loss": 0.835,
+      "step": 227
+    },
+    {
+      "epoch": 0.3569937369519833,
+      "grad_norm": 0.9525771141052246,
+      "learning_rate": 4.985470296119038e-06,
+      "loss": 0.9162,
+      "step": 228
+    },
+    {
+      "epoch": 0.3585594989561587,
+      "grad_norm": 1.150404930114746,
+      "learning_rate": 4.985242607163488e-06,
+      "loss": 0.9456,
+      "step": 229
+    },
+    {
+      "epoch": 0.36012526096033404,
+      "grad_norm": 1.01652991771698,
+      "learning_rate": 4.985013153321755e-06,
+      "loss": 0.87,
+      "step": 230
+    },
+    {
+      "epoch": 0.3616910229645094,
+      "grad_norm": 0.9455218315124512,
+      "learning_rate": 4.984781934756786e-06,
+      "loss": 0.8466,
+      "step": 231
+    },
+    {
+      "epoch": 0.36325678496868474,
+      "grad_norm": 0.9125218391418457,
+      "learning_rate": 4.984548951632779e-06,
+      "loss": 0.8764,
+      "step": 232
+    },
+    {
+      "epoch": 0.36482254697286015,
+      "grad_norm": 2.0192980766296387,
+      "learning_rate": 4.984314204115187e-06,
+      "loss": 0.7866,
+      "step": 233
+    },
+    {
+      "epoch": 0.3663883089770355,
+      "grad_norm": 1.8118515014648438,
+      "learning_rate": 4.984077692370716e-06,
+      "loss": 0.8703,
+      "step": 234
+    },
+    {
+      "epoch": 0.36795407098121086,
+      "grad_norm": 1.2995024919509888,
+      "learning_rate": 4.983839416567322e-06,
+      "loss": 0.9459,
+      "step": 235
+    },
+    {
+      "epoch": 0.3695198329853862,
+      "grad_norm": 1.0611528158187866,
+      "learning_rate": 4.983599376874216e-06,
+      "loss": 0.8456,
+      "step": 236
+    },
+    {
+      "epoch": 0.37108559498956156,
+      "grad_norm": 1.2018043994903564,
+      "learning_rate": 4.983357573461863e-06,
+      "loss": 0.914,
+      "step": 237
+    },
+    {
+      "epoch": 0.37265135699373697,
+      "grad_norm": 0.9091911315917969,
+      "learning_rate": 4.983114006501978e-06,
+      "loss": 0.791,
+      "step": 238
+    },
+    {
+      "epoch": 0.3742171189979123,
+      "grad_norm": 0.980873167514801,
+      "learning_rate": 4.982868676167528e-06,
+      "loss": 0.77,
+      "step": 239
+    },
+    {
+      "epoch": 0.3757828810020877,
+      "grad_norm": 0.8945695161819458,
+      "learning_rate": 4.982621582632735e-06,
+      "loss": 0.8038,
+      "step": 240
+    },
+    {
+      "epoch": 0.37734864300626303,
+      "grad_norm": 0.9325360655784607,
+      "learning_rate": 4.9823727260730705e-06,
+      "loss": 0.8908,
+      "step": 241
+    },
+    {
+      "epoch": 0.37891440501043844,
+      "grad_norm": 0.987630307674408,
+      "learning_rate": 4.98212210666526e-06,
+      "loss": 0.8642,
+      "step": 242
+    },
+    {
+      "epoch": 0.3804801670146138,
+      "grad_norm": 0.9151357412338257,
+      "learning_rate": 4.98186972458728e-06,
+      "loss": 0.8437,
+      "step": 243
+    },
+    {
+      "epoch": 0.38204592901878914,
+      "grad_norm": 0.9084850549697876,
+      "learning_rate": 4.981615580018358e-06,
+      "loss": 0.8775,
+      "step": 244
+    },
+    {
+      "epoch": 0.3836116910229645,
+      "grad_norm": 0.9786385297775269,
+      "learning_rate": 4.981359673138974e-06,
+      "loss": 0.8251,
+      "step": 245
+    },
+    {
+      "epoch": 0.38517745302713985,
+      "grad_norm": 0.8947877883911133,
+      "learning_rate": 4.98110200413086e-06,
+      "loss": 0.7709,
+      "step": 246
+    },
+    {
+      "epoch": 0.38674321503131526,
+      "grad_norm": 0.967019259929657,
+      "learning_rate": 4.9808425731769985e-06,
+      "loss": 0.8954,
+      "step": 247
+    },
+    {
+      "epoch": 0.3883089770354906,
+      "grad_norm": 1.1066704988479614,
+      "learning_rate": 4.980581380461622e-06,
+      "loss": 0.8544,
+      "step": 248
+    },
+    {
+      "epoch": 0.38987473903966596,
+      "grad_norm": 0.9750981330871582,
+      "learning_rate": 4.980318426170218e-06,
+      "loss": 0.8354,
+      "step": 249
+    },
+    {
+      "epoch": 0.3914405010438413,
+      "grad_norm": 0.8727232217788696,
+      "learning_rate": 4.980053710489521e-06,
+      "loss": 0.8911,
+      "step": 250
+    },
+    {
+      "epoch": 0.3930062630480167,
+      "grad_norm": 0.9337188601493835,
+      "learning_rate": 4.979787233607518e-06,
+      "loss": 0.8567,
+      "step": 251
+    },
+    {
+      "epoch": 0.3945720250521921,
+      "grad_norm": 0.890701174736023,
+      "learning_rate": 4.979518995713448e-06,
+      "loss": 0.8327,
+      "step": 252
+    },
+    {
+      "epoch": 0.39613778705636743,
+      "grad_norm": 0.9973195195198059,
+      "learning_rate": 4.979248996997797e-06,
+      "loss": 0.9065,
+      "step": 253
+    },
+    {
+      "epoch": 0.3977035490605428,
+      "grad_norm": 0.9894055128097534,
+      "learning_rate": 4.978977237652306e-06,
+      "loss": 0.8729,
+      "step": 254
+    },
+    {
+      "epoch": 0.39926931106471814,
+      "grad_norm": 1.0398536920547485,
+      "learning_rate": 4.978703717869961e-06,
+      "loss": 0.861,
+      "step": 255
+    },
+    {
+      "epoch": 0.40083507306889354,
+      "grad_norm": 1.1713942289352417,
+      "learning_rate": 4.978428437845003e-06,
+      "loss": 0.8183,
+      "step": 256
+    },
+    {
+      "epoch": 0.4024008350730689,
+      "grad_norm": 0.9868776202201843,
+      "learning_rate": 4.9781513977729225e-06,
+      "loss": 0.941,
+      "step": 257
+    },
+    {
+      "epoch": 0.40396659707724425,
+      "grad_norm": 0.9260863065719604,
+      "learning_rate": 4.977872597850455e-06,
+      "loss": 0.8963,
+      "step": 258
+    },
+    {
+      "epoch": 0.4055323590814196,
+      "grad_norm": 0.9553239941596985,
+      "learning_rate": 4.977592038275592e-06,
+      "loss": 0.8102,
+      "step": 259
+    },
+    {
+      "epoch": 0.407098121085595,
+      "grad_norm": 0.9426907300949097,
+      "learning_rate": 4.977309719247571e-06,
+      "loss": 0.89,
+      "step": 260
+    },
+    {
+      "epoch": 0.40866388308977036,
+      "grad_norm": 0.9691527485847473,
+      "learning_rate": 4.97702564096688e-06,
+      "loss": 0.7967,
+      "step": 261
+    },
+    {
+      "epoch": 0.4102296450939457,
+      "grad_norm": 0.9917487502098083,
+      "learning_rate": 4.976739803635256e-06,
+      "loss": 0.8327,
+      "step": 262
+    },
+    {
+      "epoch": 0.41179540709812107,
+      "grad_norm": 0.992340087890625,
+      "learning_rate": 4.976452207455686e-06,
+      "loss": 0.9462,
+      "step": 263
+    },
+    {
+      "epoch": 0.4133611691022965,
+      "grad_norm": 0.9554483890533447,
+      "learning_rate": 4.9761628526324035e-06,
+      "loss": 0.8294,
+      "step": 264
+    },
+    {
+      "epoch": 0.41492693110647183,
+      "grad_norm": 0.9318789839744568,
+      "learning_rate": 4.975871739370895e-06,
+      "loss": 0.8079,
+      "step": 265
+    },
+    {
+      "epoch": 0.4164926931106472,
+      "grad_norm": 0.9158215522766113,
+      "learning_rate": 4.975578867877893e-06,
+      "loss": 0.8632,
+      "step": 266
+    },
+    {
+      "epoch": 0.41805845511482254,
+      "grad_norm": 0.9444370865821838,
+      "learning_rate": 4.975284238361378e-06,
+      "loss": 0.8897,
+      "step": 267
+    },
+    {
+      "epoch": 0.4196242171189979,
+      "grad_norm": 0.9949231743812561,
+      "learning_rate": 4.974987851030581e-06,
+      "loss": 0.7905,
+      "step": 268
+    },
+    {
+      "epoch": 0.4211899791231733,
+      "grad_norm": 0.9819440245628357,
+      "learning_rate": 4.974689706095981e-06,
+      "loss": 0.9183,
+      "step": 269
+    },
+    {
+      "epoch": 0.42275574112734865,
+      "grad_norm": 1.0787206888198853,
+      "learning_rate": 4.974389803769304e-06,
+      "loss": 0.7739,
+      "step": 270
+    },
+    {
+      "epoch": 0.424321503131524,
+      "grad_norm": 0.8871361017227173,
+      "learning_rate": 4.9740881442635235e-06,
+      "loss": 0.9454,
+      "step": 271
+    },
+    {
+      "epoch": 0.42588726513569936,
+      "grad_norm": 0.9125893115997314,
+      "learning_rate": 4.973784727792863e-06,
+      "loss": 0.8017,
+      "step": 272
+    },
+    {
+      "epoch": 0.42745302713987476,
+      "grad_norm": 0.9102838635444641,
+      "learning_rate": 4.973479554572793e-06,
+      "loss": 0.8307,
+      "step": 273
+    },
+    {
+      "epoch": 0.4290187891440501,
+      "grad_norm": 0.8780474662780762,
+      "learning_rate": 4.973172624820031e-06,
+      "loss": 0.9402,
+      "step": 274
+    },
+    {
+      "epoch": 0.43058455114822547,
+      "grad_norm": 1.0059378147125244,
+      "learning_rate": 4.972863938752541e-06,
+      "loss": 0.7438,
+      "step": 275
+    },
+    {
+      "epoch": 0.4321503131524008,
+      "grad_norm": 0.9684615135192871,
+      "learning_rate": 4.972553496589537e-06,
+      "loss": 0.8897,
+      "step": 276
+    },
+    {
+      "epoch": 0.4337160751565762,
+      "grad_norm": 0.9615057110786438,
+      "learning_rate": 4.972241298551477e-06,
+      "loss": 0.8157,
+      "step": 277
+    },
+    {
+      "epoch": 0.4352818371607516,
+      "grad_norm": 0.8589702844619751,
+      "learning_rate": 4.971927344860068e-06,
+      "loss": 0.7813,
+      "step": 278
+    },
+    {
+      "epoch": 0.43684759916492694,
+      "grad_norm": 0.900324285030365,
+      "learning_rate": 4.971611635738263e-06,
+      "loss": 0.7534,
+      "step": 279
+    },
+    {
+      "epoch": 0.4384133611691023,
+      "grad_norm": 0.8881833553314209,
+      "learning_rate": 4.971294171410261e-06,
+      "loss": 0.8255,
+      "step": 280
+    },
+    {
+      "epoch": 0.43997912317327764,
+      "grad_norm": 0.9723751544952393,
+      "learning_rate": 4.9709749521015085e-06,
+      "loss": 0.8796,
+      "step": 281
+    },
+    {
+      "epoch": 0.44154488517745305,
+      "grad_norm": 0.9341713190078735,
+      "learning_rate": 4.9706539780386995e-06,
+      "loss": 0.7429,
+      "step": 282
+    },
+    {
+      "epoch": 0.4431106471816284,
+      "grad_norm": 0.9354205131530762,
+      "learning_rate": 4.97033124944977e-06,
+      "loss": 0.8545,
+      "step": 283
+    },
+    {
+      "epoch": 0.44467640918580376,
+      "grad_norm": 0.9294624328613281,
+      "learning_rate": 4.970006766563906e-06,
+      "loss": 0.7783,
+      "step": 284
+    },
+    {
+      "epoch": 0.4462421711899791,
+      "grad_norm": 0.9584887027740479,
+      "learning_rate": 4.969680529611538e-06,
+      "loss": 0.8068,
+      "step": 285
+    },
+    {
+      "epoch": 0.44780793319415446,
+      "grad_norm": 0.9525735378265381,
+      "learning_rate": 4.9693525388243424e-06,
+      "loss": 0.8055,
+      "step": 286
+    },
+    {
+      "epoch": 0.44937369519832987,
+      "grad_norm": 0.8845317959785461,
+      "learning_rate": 4.96902279443524e-06,
+      "loss": 0.8318,
+      "step": 287
+    },
+    {
+      "epoch": 0.4509394572025052,
+      "grad_norm": 0.8932276964187622,
+      "learning_rate": 4.9686912966783965e-06,
+      "loss": 0.8185,
+      "step": 288
+    },
+    {
+      "epoch": 0.4525052192066806,
+      "grad_norm": 0.9881951808929443,
+      "learning_rate": 4.968358045789225e-06,
+      "loss": 0.9001,
+      "step": 289
+    },
+    {
+      "epoch": 0.45407098121085593,
+      "grad_norm": 0.995212197303772,
+      "learning_rate": 4.968023042004384e-06,
+      "loss": 0.75,
+      "step": 290
+    },
+    {
+      "epoch": 0.45563674321503134,
+      "grad_norm": 0.9997846484184265,
+      "learning_rate": 4.967686285561773e-06,
+      "loss": 0.8993,
+      "step": 291
+    },
+    {
+      "epoch": 0.4572025052192067,
+      "grad_norm": 0.9438545107841492,
+      "learning_rate": 4.967347776700538e-06,
+      "loss": 0.8964,
+      "step": 292
+    },
+    {
+      "epoch": 0.45876826722338204,
+      "grad_norm": 0.9164395332336426,
+      "learning_rate": 4.967007515661072e-06,
+      "loss": 0.8671,
+      "step": 293
+    },
+    {
+      "epoch": 0.4603340292275574,
+      "grad_norm": 0.895302951335907,
+      "learning_rate": 4.966665502685009e-06,
+      "loss": 0.8785,
+      "step": 294
+    },
+    {
+      "epoch": 0.4618997912317328,
+      "grad_norm": 0.9122863411903381,
+      "learning_rate": 4.966321738015227e-06,
+      "loss": 0.8499,
+      "step": 295
+    },
+    {
+      "epoch": 0.46346555323590816,
+      "grad_norm": 1.0097336769104004,
+      "learning_rate": 4.965976221895852e-06,
+      "loss": 0.83,
+      "step": 296
+    },
+    {
+      "epoch": 0.4650313152400835,
+      "grad_norm": 1.0664499998092651,
+      "learning_rate": 4.965628954572249e-06,
+      "loss": 0.7799,
+      "step": 297
+    },
+    {
+      "epoch": 0.46659707724425886,
+      "grad_norm": 0.9602341055870056,
+      "learning_rate": 4.96527993629103e-06,
+      "loss": 0.8061,
+      "step": 298
+    },
+    {
+      "epoch": 0.4681628392484342,
+      "grad_norm": 1.0133620500564575,
+      "learning_rate": 4.964929167300048e-06,
+      "loss": 0.9148,
+      "step": 299
+    },
+    {
+      "epoch": 0.4697286012526096,
+      "grad_norm": 0.9351868033409119,
+      "learning_rate": 4.964576647848401e-06,
+      "loss": 0.83,
+      "step": 300
+    },
+    {
+      "epoch": 0.471294363256785,
+      "grad_norm": 1.0078860521316528,
+      "learning_rate": 4.964222378186429e-06,
+      "loss": 0.8391,
+      "step": 301
+    },
+    {
+      "epoch": 0.47286012526096033,
+      "grad_norm": 0.8966493606567383,
+      "learning_rate": 4.963866358565716e-06,
+      "loss": 0.8677,
+      "step": 302
+    },
+    {
+      "epoch": 0.4744258872651357,
+      "grad_norm": 0.9283657670021057,
+      "learning_rate": 4.963508589239087e-06,
+      "loss": 0.8787,
+      "step": 303
+    },
+    {
+      "epoch": 0.4759916492693111,
+      "grad_norm": 0.9540023803710938,
+      "learning_rate": 4.963149070460611e-06,
+      "loss": 0.8667,
+      "step": 304
+    },
+    {
+      "epoch": 0.47755741127348644,
+      "grad_norm": 0.9580283164978027,
+      "learning_rate": 4.9627878024855995e-06,
+      "loss": 0.8442,
+      "step": 305
+    },
+    {
+      "epoch": 0.4791231732776618,
+      "grad_norm": 0.9008278846740723,
+      "learning_rate": 4.962424785570605e-06,
+      "loss": 0.8605,
+      "step": 306
+    },
+    {
+      "epoch": 0.48068893528183715,
+      "grad_norm": 0.9681883454322815,
+      "learning_rate": 4.962060019973423e-06,
+      "loss": 0.7865,
+      "step": 307
+    },
+    {
+      "epoch": 0.4822546972860125,
+      "grad_norm": 0.8492597341537476,
+      "learning_rate": 4.9616935059530915e-06,
+      "loss": 0.8578,
+      "step": 308
+    },
+    {
+      "epoch": 0.4838204592901879,
+      "grad_norm": 0.9323164820671082,
+      "learning_rate": 4.9613252437698865e-06,
+      "loss": 0.8491,
+      "step": 309
+    },
+    {
+      "epoch": 0.48538622129436326,
+      "grad_norm": 0.9417833685874939,
+      "learning_rate": 4.960955233685331e-06,
+      "loss": 0.8108,
+      "step": 310
+    },
+    {
+      "epoch": 0.4869519832985386,
+      "grad_norm": 0.901364803314209,
+      "learning_rate": 4.960583475962185e-06,
+      "loss": 0.8958,
+      "step": 311
+    },
+    {
+      "epoch": 0.48851774530271397,
+      "grad_norm": 0.8929979205131531,
+      "learning_rate": 4.96020997086445e-06,
+      "loss": 0.8237,
+      "step": 312
+    },
+    {
+      "epoch": 0.4900835073068894,
+      "grad_norm": 0.9422337412834167,
+      "learning_rate": 4.9598347186573716e-06,
+      "loss": 0.8024,
+      "step": 313
+    },
+    {
+      "epoch": 0.49164926931106473,
+      "grad_norm": 0.9165504574775696,
+      "learning_rate": 4.959457719607432e-06,
+      "loss": 0.8124,
+      "step": 314
+    },
+    {
+      "epoch": 0.4932150313152401,
+      "grad_norm": 1.0233370065689087,
+      "learning_rate": 4.959078973982356e-06,
+      "loss": 0.788,
+      "step": 315
+    },
+    {
+      "epoch": 0.49478079331941544,
+      "grad_norm": 0.9375972151756287,
+      "learning_rate": 4.958698482051109e-06,
+      "loss": 0.839,
+      "step": 316
+    },
+    {
+      "epoch": 0.4963465553235908,
+      "grad_norm": 0.8805437684059143,
+      "learning_rate": 4.958316244083895e-06,
+      "loss": 0.7445,
+      "step": 317
+    },
+    {
+      "epoch": 0.4979123173277662,
+      "grad_norm": 1.0231893062591553,
+      "learning_rate": 4.95793226035216e-06,
+      "loss": 0.8257,
+      "step": 318
+    },
+    {
+      "epoch": 0.49947807933194155,
+      "grad_norm": 0.9125928282737732,
+      "learning_rate": 4.957546531128589e-06,
+      "loss": 0.7699,
+      "step": 319
+    },
+    {
+      "epoch": 0.5010438413361169,
+      "grad_norm": 0.8844016790390015,
+      "learning_rate": 4.957159056687104e-06,
+      "loss": 0.8784,
+      "step": 320
+    },
+    {
+      "epoch": 0.5026096033402923,
+      "grad_norm": 0.9770572781562805,
+      "learning_rate": 4.956769837302871e-06,
+      "loss": 0.8611,
+      "step": 321
+    },
+    {
+      "epoch": 0.5041753653444676,
+      "grad_norm": 0.9413725733757019,
+      "learning_rate": 4.956378873252293e-06,
+      "loss": 0.8056,
+      "step": 322
+    },
+    {
+      "epoch": 0.505741127348643,
+      "grad_norm": 0.9058850407600403,
+      "learning_rate": 4.955986164813008e-06,
+      "loss": 0.8162,
+      "step": 323
+    },
+    {
+      "epoch": 0.5073068893528184,
+      "grad_norm": 0.893100917339325,
+      "learning_rate": 4.955591712263901e-06,
+      "loss": 0.7554,
+      "step": 324
+    },
+    {
+      "epoch": 0.5088726513569938,
+      "grad_norm": 0.8967075943946838,
+      "learning_rate": 4.955195515885089e-06,
+      "loss": 0.8143,
+      "step": 325
+    },
+    {
+      "epoch": 0.5104384133611691,
+      "grad_norm": 0.9542785286903381,
+      "learning_rate": 4.954797575957929e-06,
+      "loss": 0.8102,
+      "step": 326
+    },
+    {
+      "epoch": 0.5120041753653445,
+      "grad_norm": 1.006451964378357,
+      "learning_rate": 4.954397892765018e-06,
+      "loss": 0.86,
+      "step": 327
+    },
+    {
+      "epoch": 0.5135699373695198,
+      "grad_norm": 0.9964715838432312,
+      "learning_rate": 4.953996466590189e-06,
+      "loss": 0.7166,
+      "step": 328
+    },
+    {
+      "epoch": 0.5151356993736952,
+      "grad_norm": 0.9568808674812317,
+      "learning_rate": 4.953593297718512e-06,
+      "loss": 0.8753,
+      "step": 329
+    },
+    {
+      "epoch": 0.5167014613778705,
+      "grad_norm": 0.9435825347900391,
+      "learning_rate": 4.953188386436298e-06,
+      "loss": 0.8961,
+      "step": 330
+    },
+    {
+      "epoch": 0.5182672233820459,
+      "grad_norm": 0.9386457800865173,
+      "learning_rate": 4.952781733031092e-06,
+      "loss": 0.8468,
+      "step": 331
+    },
+    {
+      "epoch": 0.5198329853862212,
+      "grad_norm": 0.9916554689407349,
+      "learning_rate": 4.952373337791678e-06,
+      "loss": 0.8274,
+      "step": 332
+    },
+    {
+      "epoch": 0.5213987473903967,
+      "grad_norm": 0.9351572394371033,
+      "learning_rate": 4.9519632010080765e-06,
+      "loss": 0.7964,
+      "step": 333
+    },
+    {
+      "epoch": 0.5229645093945721,
+      "grad_norm": 0.9936269521713257,
+      "learning_rate": 4.951551322971545e-06,
+      "loss": 0.8354,
+      "step": 334
+    },
+    {
+      "epoch": 0.5245302713987474,
+      "grad_norm": 0.9678221940994263,
+      "learning_rate": 4.951137703974576e-06,
+      "loss": 0.9115,
+      "step": 335
+    },
+    {
+      "epoch": 0.5260960334029228,
+      "grad_norm": 0.9453133940696716,
+      "learning_rate": 4.9507223443109005e-06,
+      "loss": 0.7725,
+      "step": 336
+    },
+    {
+      "epoch": 0.5276617954070981,
+      "grad_norm": 0.9577226638793945,
+      "learning_rate": 4.950305244275484e-06,
+      "loss": 0.8679,
+      "step": 337
+    },
+    {
+      "epoch": 0.5292275574112735,
+      "grad_norm": 0.998489260673523,
+      "learning_rate": 4.94988640416453e-06,
+      "loss": 0.7505,
+      "step": 338
+    },
+    {
+      "epoch": 0.5307933194154488,
+      "grad_norm": 0.976026713848114,
+      "learning_rate": 4.949465824275475e-06,
+      "loss": 0.9083,
+      "step": 339
+    },
+    {
+      "epoch": 0.5323590814196242,
+      "grad_norm": 1.0041502714157104,
+      "learning_rate": 4.9490435049069925e-06,
+      "loss": 0.8815,
+      "step": 340
+    },
+    {
+      "epoch": 0.5339248434237995,
+      "grad_norm": 0.9839801788330078,
+      "learning_rate": 4.948619446358991e-06,
+      "loss": 0.8517,
+      "step": 341
+    },
+    {
+      "epoch": 0.535490605427975,
+      "grad_norm": 0.9995140433311462,
+      "learning_rate": 4.948193648932616e-06,
+      "loss": 0.8157,
+      "step": 342
+    },
+    {
+      "epoch": 0.5370563674321504,
+      "grad_norm": 1.0782225131988525,
+      "learning_rate": 4.947766112930243e-06,
+      "loss": 0.9206,
+      "step": 343
+    },
+    {
+      "epoch": 0.5386221294363257,
+      "grad_norm": 0.9981802105903625,
+      "learning_rate": 4.947336838655487e-06,
+      "loss": 0.8389,
+      "step": 344
+    },
+    {
+      "epoch": 0.5401878914405011,
+      "grad_norm": 0.9261236190795898,
+      "learning_rate": 4.946905826413195e-06,
+      "loss": 0.8035,
+      "step": 345
+    },
+    {
+      "epoch": 0.5417536534446764,
+      "grad_norm": 0.9195284247398376,
+      "learning_rate": 4.946473076509449e-06,
+      "loss": 0.8265,
+      "step": 346
+    },
+    {
+      "epoch": 0.5433194154488518,
+      "grad_norm": 0.9535123109817505,
+      "learning_rate": 4.946038589251566e-06,
+      "loss": 0.7528,
+      "step": 347
+    },
+    {
+      "epoch": 0.5448851774530271,
+      "grad_norm": 0.9969532489776611,
+      "learning_rate": 4.9456023649480935e-06,
+      "loss": 0.7787,
+      "step": 348
+    },
+    {
+      "epoch": 0.5464509394572025,
+      "grad_norm": 0.9612399935722351,
+      "learning_rate": 4.945164403908816e-06,
+      "loss": 0.8298,
+      "step": 349
+    },
+    {
+      "epoch": 0.5480167014613778,
+      "grad_norm": 0.9541627764701843,
+      "learning_rate": 4.944724706444749e-06,
+      "loss": 0.8039,
+      "step": 350
+    },
+    {
+      "epoch": 0.5495824634655533,
+      "grad_norm": 0.9916557669639587,
+      "learning_rate": 4.944283272868143e-06,
+      "loss": 0.9012,
+      "step": 351
+    },
+    {
+      "epoch": 0.5511482254697286,
+      "grad_norm": 0.8807661533355713,
+      "learning_rate": 4.94384010349248e-06,
+      "loss": 0.9005,
+      "step": 352
+    },
+    {
+      "epoch": 0.552713987473904,
+      "grad_norm": 1.0179708003997803,
+      "learning_rate": 4.943395198632476e-06,
+      "loss": 0.8845,
+      "step": 353
+    },
+    {
+      "epoch": 0.5542797494780793,
+      "grad_norm": 0.923014223575592,
+      "learning_rate": 4.942948558604077e-06,
+      "loss": 0.8173,
+      "step": 354
+    },
+    {
+      "epoch": 0.5558455114822547,
+      "grad_norm": 0.9318749308586121,
+      "learning_rate": 4.942500183724463e-06,
+      "loss": 0.8828,
+      "step": 355
+    },
+    {
+      "epoch": 0.55741127348643,
+      "grad_norm": 1.0907000303268433,
+      "learning_rate": 4.942050074312048e-06,
+      "loss": 0.7951,
+      "step": 356
+    },
+    {
+      "epoch": 0.5589770354906054,
+      "grad_norm": 0.9811620712280273,
+      "learning_rate": 4.9415982306864744e-06,
+      "loss": 0.8826,
+      "step": 357
+    },
+    {
+      "epoch": 0.5605427974947808,
+      "grad_norm": 0.9888150691986084,
+      "learning_rate": 4.9411446531686156e-06,
+      "loss": 0.8338,
+      "step": 358
+    },
+    {
+      "epoch": 0.5621085594989561,
+      "grad_norm": 0.9389257431030273,
+      "learning_rate": 4.9406893420805804e-06,
+      "loss": 0.8316,
+      "step": 359
+    },
+    {
+      "epoch": 0.5636743215031316,
+      "grad_norm": 0.9500603675842285,
+      "learning_rate": 4.940232297745705e-06,
+      "loss": 0.8764,
+      "step": 360
+    },
+    {
+      "epoch": 0.5652400835073069,
+      "grad_norm": 0.9748769998550415,
+      "learning_rate": 4.939773520488559e-06,
+      "loss": 0.8308,
+      "step": 361
+    },
+    {
+      "epoch": 0.5668058455114823,
+      "grad_norm": 0.9179681539535522,
+      "learning_rate": 4.93931301063494e-06,
+      "loss": 0.789,
+      "step": 362
+    },
+    {
+      "epoch": 0.5683716075156576,
+      "grad_norm": 0.8609325289726257,
+      "learning_rate": 4.938850768511878e-06,
+      "loss": 0.7368,
+      "step": 363
+    },
+    {
+      "epoch": 0.569937369519833,
+      "grad_norm": 0.9375531673431396,
+      "learning_rate": 4.9383867944476325e-06,
+      "loss": 0.8164,
+      "step": 364
+    },
+    {
+      "epoch": 0.5715031315240083,
+      "grad_norm": 0.9386453032493591,
+      "learning_rate": 4.937921088771694e-06,
+      "loss": 0.7898,
+      "step": 365
+    },
+    {
+      "epoch": 0.5730688935281837,
+      "grad_norm": 1.025040864944458,
+      "learning_rate": 4.937453651814781e-06,
+      "loss": 0.8465,
+      "step": 366
+    },
+    {
+      "epoch": 0.574634655532359,
+      "grad_norm": 1.1037245988845825,
+      "learning_rate": 4.936984483908842e-06,
+      "loss": 0.7935,
+      "step": 367
+    },
+    {
+      "epoch": 0.5762004175365344,
+      "grad_norm": 1.230821132659912,
+      "learning_rate": 4.936513585387055e-06,
+      "loss": 0.8043,
+      "step": 368
+    },
+    {
+      "epoch": 0.5777661795407099,
+      "grad_norm": 1.090235948562622,
+      "learning_rate": 4.9360409565838265e-06,
+      "loss": 0.8726,
+      "step": 369
+    },
+    {
+      "epoch": 0.5793319415448852,
+      "grad_norm": 0.8977810144424438,
+      "learning_rate": 4.935566597834793e-06,
+      "loss": 0.8667,
+      "step": 370
+    },
+    {
+      "epoch": 0.5808977035490606,
+      "grad_norm": 0.96401447057724,
+      "learning_rate": 4.935090509476818e-06,
+      "loss": 0.8494,
+      "step": 371
+    },
+    {
+      "epoch": 0.5824634655532359,
+      "grad_norm": 0.9290149211883545,
+      "learning_rate": 4.934612691847995e-06,
+      "loss": 0.7678,
+      "step": 372
+    },
+    {
+      "epoch": 0.5840292275574113,
+      "grad_norm": 0.9765584468841553,
+      "learning_rate": 4.934133145287644e-06,
+      "loss": 0.7947,
+      "step": 373
+    },
+    {
+      "epoch": 0.5855949895615866,
+      "grad_norm": 0.8939430117607117,
+      "learning_rate": 4.933651870136313e-06,
+      "loss": 0.885,
+      "step": 374
+    },
+    {
+      "epoch": 0.587160751565762,
+      "grad_norm": 0.9512760043144226,
+      "learning_rate": 4.933168866735777e-06,
+      "loss": 0.8863,
+      "step": 375
+    },
+    {
+      "epoch": 0.5887265135699373,
+      "grad_norm": 1.0334441661834717,
+      "learning_rate": 4.932684135429041e-06,
+      "loss": 0.8362,
+      "step": 376
+    },
+    {
+      "epoch": 0.5902922755741128,
+      "grad_norm": 0.9402525424957275,
+      "learning_rate": 4.932197676560334e-06,
+      "loss": 0.8758,
+      "step": 377
+    },
+    {
+      "epoch": 0.5918580375782881,
+      "grad_norm": 0.9615777730941772,
+      "learning_rate": 4.931709490475113e-06,
+      "loss": 0.8324,
+      "step": 378
+    },
+    {
+      "epoch": 0.5934237995824635,
+      "grad_norm": 0.9407138228416443,
+      "learning_rate": 4.931219577520061e-06,
+      "loss": 0.8454,
+      "step": 379
+    },
+    {
+      "epoch": 0.5949895615866388,
+      "grad_norm": 1.0415815114974976,
+      "learning_rate": 4.930727938043091e-06,
+      "loss": 0.8649,
+      "step": 380
+    },
+    {
+      "epoch": 0.5965553235908142,
+      "grad_norm": 1.0946165323257446,
+      "learning_rate": 4.9302345723933344e-06,
+      "loss": 0.7794,
+      "step": 381
+    },
+    {
+      "epoch": 0.5981210855949896,
+      "grad_norm": 1.0532523393630981,
+      "learning_rate": 4.929739480921157e-06,
+      "loss": 0.8555,
+      "step": 382
+    },
+    {
+      "epoch": 0.5996868475991649,
+      "grad_norm": 0.9441399574279785,
+      "learning_rate": 4.929242663978144e-06,
+      "loss": 0.7613,
+      "step": 383
+    },
+    {
+      "epoch": 0.6012526096033403,
+      "grad_norm": 0.9970793128013611,
+      "learning_rate": 4.92874412191711e-06,
+      "loss": 0.9037,
+      "step": 384
+    },
+    {
+      "epoch": 0.6028183716075156,
+      "grad_norm": 0.9777354001998901,
+      "learning_rate": 4.928243855092091e-06,
+      "loss": 0.9044,
+      "step": 385
+    },
+    {
+      "epoch": 0.6043841336116911,
+      "grad_norm": 0.9507855772972107,
+      "learning_rate": 4.92774186385835e-06,
+      "loss": 0.8359,
+      "step": 386
+    },
+    {
+      "epoch": 0.6059498956158664,
+      "grad_norm": 0.9827553033828735,
+      "learning_rate": 4.9272381485723765e-06,
+      "loss": 0.7938,
+      "step": 387
+    },
+    {
+      "epoch": 0.6075156576200418,
+      "grad_norm": 0.8896475434303284,
+      "learning_rate": 4.926732709591879e-06,
+      "loss": 0.8545,
+      "step": 388
+    },
+    {
+      "epoch": 0.6090814196242171,
+      "grad_norm": 0.9817705750465393,
+      "learning_rate": 4.926225547275795e-06,
+      "loss": 0.8456,
+      "step": 389
+    },
+    {
+      "epoch": 0.6106471816283925,
+      "grad_norm": 0.9646008610725403,
+      "learning_rate": 4.925716661984283e-06,
+      "loss": 0.7906,
+      "step": 390
+    },
+    {
+      "epoch": 0.6122129436325678,
+      "grad_norm": 0.8940444588661194,
+      "learning_rate": 4.9252060540787275e-06,
+      "loss": 0.8371,
+      "step": 391
+    },
+    {
+      "epoch": 0.6137787056367432,
+      "grad_norm": 1.0812394618988037,
+      "learning_rate": 4.924693723921734e-06,
+      "loss": 0.8056,
+      "step": 392
+    },
+    {
+      "epoch": 0.6153444676409185,
+      "grad_norm": 0.9953856468200684,
+      "learning_rate": 4.924179671877132e-06,
+      "loss": 0.8651,
+      "step": 393
+    },
+    {
+      "epoch": 0.6169102296450939,
+      "grad_norm": 0.9760703444480896,
+      "learning_rate": 4.9236638983099735e-06,
+      "loss": 0.845,
+      "step": 394
+    },
+    {
+      "epoch": 0.6184759916492694,
+      "grad_norm": 1.0127615928649902,
+      "learning_rate": 4.923146403586533e-06,
+      "loss": 0.7987,
+      "step": 395
+    },
+    {
+      "epoch": 0.6200417536534447,
+      "grad_norm": 0.8771396279335022,
+      "learning_rate": 4.9226271880743086e-06,
+      "loss": 0.8092,
+      "step": 396
+    },
+    {
+      "epoch": 0.6216075156576201,
+      "grad_norm": 0.8906831741333008,
+      "learning_rate": 4.9221062521420196e-06,
+      "loss": 0.8039,
+      "step": 397
+    },
+    {
+      "epoch": 0.6231732776617954,
+      "grad_norm": 1.0218392610549927,
+      "learning_rate": 4.9215835961596045e-06,
+      "loss": 0.8354,
+      "step": 398
+    },
+    {
+      "epoch": 0.6247390396659708,
+      "grad_norm": 1.0382038354873657,
+      "learning_rate": 4.921059220498227e-06,
+      "loss": 0.8021,
+      "step": 399
+    },
+    {
+      "epoch": 0.6263048016701461,
+      "grad_norm": 0.8997588157653809,
+      "learning_rate": 4.920533125530272e-06,
+      "loss": 0.8716,
+      "step": 400
+    },
+    {
+      "epoch": 0.6278705636743215,
+      "grad_norm": 0.8725724220275879,
+      "learning_rate": 4.920005311629342e-06,
+      "loss": 0.7934,
+      "step": 401
+    },
+    {
+      "epoch": 0.6294363256784968,
+      "grad_norm": 0.9289800524711609,
+      "learning_rate": 4.9194757791702625e-06,
+      "loss": 0.812,
+      "step": 402
+    },
+    {
+      "epoch": 0.6310020876826722,
+      "grad_norm": 0.9969078898429871,
+      "learning_rate": 4.918944528529079e-06,
+      "loss": 0.866,
+      "step": 403
+    },
+    {
+      "epoch": 0.6325678496868476,
+      "grad_norm": 0.9544941186904907,
+      "learning_rate": 4.918411560083058e-06,
+      "loss": 0.8035,
+      "step": 404
+    },
+    {
+      "epoch": 0.634133611691023,
+      "grad_norm": 0.9678053855895996,
+      "learning_rate": 4.917876874210686e-06,
+      "loss": 0.8254,
+      "step": 405
+    },
+    {
+      "epoch": 0.6356993736951984,
+      "grad_norm": 0.9782965779304504,
+      "learning_rate": 4.917340471291666e-06,
+      "loss": 0.7954,
+      "step": 406
+    },
+    {
+      "epoch": 0.6372651356993737,
+      "grad_norm": 0.9380600452423096,
+      "learning_rate": 4.916802351706923e-06,
+      "loss": 0.8287,
+      "step": 407
+    },
+    {
+      "epoch": 0.6388308977035491,
+      "grad_norm": 0.9455891847610474,
+      "learning_rate": 4.916262515838603e-06,
+      "loss": 0.8714,
+      "step": 408
+    },
+    {
+      "epoch": 0.6403966597077244,
+      "grad_norm": 0.8937395811080933,
+      "learning_rate": 4.915720964070065e-06,
+      "loss": 0.7618,
+      "step": 409
+    },
+    {
+      "epoch": 0.6419624217118998,
+      "grad_norm": 1.0304689407348633,
+      "learning_rate": 4.915177696785894e-06,
+      "loss": 0.8436,
+      "step": 410
+    },
+    {
+      "epoch": 0.6435281837160751,
+      "grad_norm": 0.9623110294342041,
+      "learning_rate": 4.914632714371885e-06,
+      "loss": 0.865,
+      "step": 411
+    },
+    {
+      "epoch": 0.6450939457202505,
+      "grad_norm": 0.9212867021560669,
+      "learning_rate": 4.914086017215059e-06,
+      "loss": 0.8232,
+      "step": 412
+    },
+    {
+      "epoch": 0.6466597077244259,
+      "grad_norm": 0.9817205667495728,
+      "learning_rate": 4.9135376057036485e-06,
+      "loss": 0.7926,
+      "step": 413
+    },
+    {
+      "epoch": 0.6482254697286013,
+      "grad_norm": 0.9576719403266907,
+      "learning_rate": 4.912987480227108e-06,
+      "loss": 0.8812,
+      "step": 414
+    },
+    {
+      "epoch": 0.6497912317327766,
+      "grad_norm": 0.9415407180786133,
+      "learning_rate": 4.912435641176106e-06,
+      "loss": 0.8985,
+      "step": 415
+    },
+    {
+      "epoch": 0.651356993736952,
+      "grad_norm": 0.9084275364875793,
+      "learning_rate": 4.91188208894253e-06,
+      "loss": 0.8085,
+      "step": 416
+    },
+    {
+      "epoch": 0.6529227557411273,
+      "grad_norm": 0.8890969753265381,
+      "learning_rate": 4.911326823919482e-06,
+      "loss": 0.8127,
+      "step": 417
+    },
+    {
+      "epoch": 0.6544885177453027,
+      "grad_norm": 0.8727678656578064,
+      "learning_rate": 4.9107698465012815e-06,
+      "loss": 0.8656,
+      "step": 418
+    },
+    {
+      "epoch": 0.656054279749478,
+      "grad_norm": 1.0174667835235596,
+      "learning_rate": 4.910211157083465e-06,
+      "loss": 0.8358,
+      "step": 419
+    },
+    {
+      "epoch": 0.6576200417536534,
+      "grad_norm": 1.045121431350708,
+      "learning_rate": 4.909650756062782e-06,
+      "loss": 0.8239,
+      "step": 420
+    },
+    {
+      "epoch": 0.6591858037578288,
+      "grad_norm": 1.0688676834106445,
+      "learning_rate": 4.909088643837202e-06,
+      "loss": 0.9246,
+      "step": 421
+    },
+    {
+      "epoch": 0.6607515657620042,
+      "grad_norm": 0.9371877908706665,
+      "learning_rate": 4.908524820805905e-06,
+      "loss": 0.7862,
+      "step": 422
+    },
+    {
+      "epoch": 0.6623173277661796,
+      "grad_norm": 0.9567264318466187,
+      "learning_rate": 4.907959287369288e-06,
+      "loss": 0.8164,
+      "step": 423
+    },
+    {
+      "epoch": 0.6638830897703549,
+      "grad_norm": 0.9109166860580444,
+      "learning_rate": 4.907392043928964e-06,
+      "loss": 0.7016,
+      "step": 424
+    },
+    {
+      "epoch": 0.6654488517745303,
+      "grad_norm": 1.0378203392028809,
+      "learning_rate": 4.906823090887757e-06,
+      "loss": 0.8986,
+      "step": 425
+    },
+    {
+      "epoch": 0.6670146137787056,
+      "grad_norm": 0.9871951341629028,
+      "learning_rate": 4.906252428649708e-06,
+      "loss": 0.8366,
+      "step": 426
+    },
+    {
+      "epoch": 0.668580375782881,
+      "grad_norm": 0.8682780265808105,
+      "learning_rate": 4.905680057620072e-06,
+      "loss": 0.738,
+      "step": 427
+    },
+    {
+      "epoch": 0.6701461377870563,
+      "grad_norm": 0.8834647536277771,
+      "learning_rate": 4.9051059782053125e-06,
+      "loss": 0.8304,
+      "step": 428
+    },
+    {
+      "epoch": 0.6717118997912317,
+      "grad_norm": 1.0133378505706787,
+      "learning_rate": 4.904530190813113e-06,
+      "loss": 0.9369,
+      "step": 429
+    },
+    {
+      "epoch": 0.673277661795407,
+      "grad_norm": 0.9967373013496399,
+      "learning_rate": 4.9039526958523676e-06,
+      "loss": 0.782,
+      "step": 430
+    },
+    {
+      "epoch": 0.6748434237995825,
+      "grad_norm": 1.0258032083511353,
+      "learning_rate": 4.90337349373318e-06,
+      "loss": 0.8214,
+      "step": 431
+    },
+    {
+      "epoch": 0.6764091858037579,
+      "grad_norm": 0.9814366102218628,
+      "learning_rate": 4.90279258486687e-06,
+      "loss": 0.8173,
+      "step": 432
+    },
+    {
+      "epoch": 0.6779749478079332,
+      "grad_norm": 0.9351440668106079,
+      "learning_rate": 4.902209969665966e-06,
+      "loss": 0.8815,
+      "step": 433
+    },
+    {
+      "epoch": 0.6795407098121086,
+      "grad_norm": 0.9490398168563843,
+      "learning_rate": 4.901625648544212e-06,
+      "loss": 0.8295,
+      "step": 434
+    },
+    {
+      "epoch": 0.6811064718162839,
+      "grad_norm": 1.0183534622192383,
+      "learning_rate": 4.9010396219165614e-06,
+      "loss": 0.8538,
+      "step": 435
+    },
+    {
+      "epoch": 0.6826722338204593,
+      "grad_norm": 0.9546058773994446,
+      "learning_rate": 4.900451890199179e-06,
+      "loss": 0.8385,
+      "step": 436
+    },
+    {
+      "epoch": 0.6842379958246346,
+      "grad_norm": 1.076120138168335,
+      "learning_rate": 4.8998624538094394e-06,
+      "loss": 0.8958,
+      "step": 437
+    },
+    {
+      "epoch": 0.68580375782881,
+      "grad_norm": 0.8935732841491699,
+      "learning_rate": 4.899271313165929e-06,
+      "loss": 0.8674,
+      "step": 438
+    },
+    {
+      "epoch": 0.6873695198329853,
+      "grad_norm": 0.9574604630470276,
+      "learning_rate": 4.898678468688445e-06,
+      "loss": 0.8245,
+      "step": 439
+    },
+    {
+      "epoch": 0.6889352818371608,
+      "grad_norm": 0.931751549243927,
+      "learning_rate": 4.898083920797993e-06,
+      "loss": 0.8409,
+      "step": 440
+    },
+    {
+      "epoch": 0.6905010438413361,
+      "grad_norm": 0.937874436378479,
+      "learning_rate": 4.89748766991679e-06,
+      "loss": 0.8222,
+      "step": 441
+    },
+    {
+      "epoch": 0.6920668058455115,
+      "grad_norm": 0.9606375694274902,
+      "learning_rate": 4.896889716468259e-06,
+      "loss": 0.8123,
+      "step": 442
+    },
+    {
+      "epoch": 0.6936325678496869,
+      "grad_norm": 0.8692249059677124,
+      "learning_rate": 4.896290060877038e-06,
+      "loss": 0.8532,
+      "step": 443
+    },
+    {
+      "epoch": 0.6951983298538622,
+      "grad_norm": 0.9597892165184021,
+      "learning_rate": 4.895688703568968e-06,
+      "loss": 0.8365,
+      "step": 444
+    },
+    {
+      "epoch": 0.6967640918580376,
+      "grad_norm": 0.958342432975769,
+      "learning_rate": 4.895085644971102e-06,
+      "loss": 0.8535,
+      "step": 445
+    },
+    {
+      "epoch": 0.6983298538622129,
+      "grad_norm": 0.9603601098060608,
+      "learning_rate": 4.894480885511699e-06,
+      "loss": 0.8227,
+      "step": 446
+    },
+    {
+      "epoch": 0.6998956158663883,
+      "grad_norm": 1.0010526180267334,
+      "learning_rate": 4.893874425620226e-06,
+      "loss": 0.8785,
+      "step": 447
+    },
+    {
+      "epoch": 0.7014613778705637,
+      "grad_norm": 1.0061219930648804,
+      "learning_rate": 4.893266265727361e-06,
+      "loss": 0.8336,
+      "step": 448
+    },
+    {
+      "epoch": 0.7030271398747391,
+      "grad_norm": 0.8967173099517822,
+      "learning_rate": 4.892656406264984e-06,
+      "loss": 0.8305,
+      "step": 449
+    },
+    {
+      "epoch": 0.7045929018789144,
+      "grad_norm": 0.962729275226593,
+      "learning_rate": 4.892044847666186e-06,
+      "loss": 0.7843,
+      "step": 450
+    },
+    {
+      "epoch": 0.7061586638830898,
+      "grad_norm": 1.0067342519760132,
+      "learning_rate": 4.891431590365262e-06,
+      "loss": 0.8111,
+      "step": 451
+    },
+    {
+      "epoch": 0.7077244258872651,
+      "grad_norm": 1.0252268314361572,
+      "learning_rate": 4.890816634797716e-06,
+      "loss": 0.8253,
+      "step": 452
+    },
+    {
+      "epoch": 0.7092901878914405,
+      "grad_norm": 0.9098564386367798,
+      "learning_rate": 4.890199981400255e-06,
+      "loss": 0.8526,
+      "step": 453
+    },
+    {
+      "epoch": 0.7108559498956158,
+      "grad_norm": 0.893135130405426,
+      "learning_rate": 4.889581630610793e-06,
+      "loss": 0.8217,
+      "step": 454
+    },
+    {
+      "epoch": 0.7124217118997912,
+      "grad_norm": 0.9225817322731018,
+      "learning_rate": 4.888961582868453e-06,
+      "loss": 0.8296,
+      "step": 455
+    },
+    {
+      "epoch": 0.7139874739039666,
+      "grad_norm": 0.973386824131012,
+      "learning_rate": 4.888339838613555e-06,
+      "loss": 0.7429,
+      "step": 456
+    },
+    {
+      "epoch": 0.715553235908142,
+      "grad_norm": 0.9829993844032288,
+      "learning_rate": 4.8877163982876306e-06,
+      "loss": 0.7405,
+      "step": 457
+    },
+    {
+      "epoch": 0.7171189979123174,
+      "grad_norm": 0.927728533744812,
+      "learning_rate": 4.887091262333414e-06,
+      "loss": 0.8257,
+      "step": 458
+    },
+    {
+      "epoch": 0.7186847599164927,
+      "grad_norm": 0.9419897794723511,
+      "learning_rate": 4.886464431194844e-06,
+      "loss": 0.7762,
+      "step": 459
+    },
+    {
+      "epoch": 0.7202505219206681,
+      "grad_norm": 0.9358823299407959,
+      "learning_rate": 4.885835905317061e-06,
+      "loss": 0.7794,
+      "step": 460
+    },
+    {
+      "epoch": 0.7218162839248434,
+      "grad_norm": 1.0290783643722534,
+      "learning_rate": 4.88520568514641e-06,
+      "loss": 0.8723,
+      "step": 461
+    },
+    {
+      "epoch": 0.7233820459290188,
+      "grad_norm": 0.9909068942070007,
+      "learning_rate": 4.884573771130442e-06,
+      "loss": 0.8302,
+      "step": 462
+    },
+    {
+      "epoch": 0.7249478079331941,
+      "grad_norm": 0.9972653985023499,
+      "learning_rate": 4.883940163717906e-06,
+      "loss": 0.7961,
+      "step": 463
+    },
+    {
+      "epoch": 0.7265135699373695,
+      "grad_norm": 0.9245638251304626,
+      "learning_rate": 4.883304863358757e-06,
+      "loss": 0.8057,
+      "step": 464
+    },
+    {
+      "epoch": 0.7280793319415448,
+      "grad_norm": 1.0097148418426514,
+      "learning_rate": 4.882667870504153e-06,
+      "loss": 0.7296,
+      "step": 465
+    },
+    {
+      "epoch": 0.7296450939457203,
+      "grad_norm": 0.9792670607566833,
+      "learning_rate": 4.882029185606448e-06,
+      "loss": 0.8006,
+      "step": 466
+    },
+    {
+      "epoch": 0.7312108559498957,
+      "grad_norm": 0.9308120608329773,
+      "learning_rate": 4.881388809119206e-06,
+      "loss": 0.7957,
+      "step": 467
+    },
+    {
+      "epoch": 0.732776617954071,
+      "grad_norm": 0.9472301602363586,
+      "learning_rate": 4.880746741497187e-06,
+      "loss": 0.8984,
+      "step": 468
+    },
+    {
+      "epoch": 0.7343423799582464,
+      "grad_norm": 0.9898937344551086,
+      "learning_rate": 4.880102983196351e-06,
+      "loss": 0.8365,
+      "step": 469
+    },
+    {
+      "epoch": 0.7359081419624217,
+      "grad_norm": 0.9376612901687622,
+      "learning_rate": 4.879457534673863e-06,
+      "loss": 0.7838,
+      "step": 470
+    },
+    {
+      "epoch": 0.7374739039665971,
+      "grad_norm": 1.030616283416748,
+      "learning_rate": 4.878810396388085e-06,
+      "loss": 0.7698,
+      "step": 471
+    },
+    {
+      "epoch": 0.7390396659707724,
+      "grad_norm": 0.9583387970924377,
+      "learning_rate": 4.87816156879858e-06,
+      "loss": 0.7381,
+      "step": 472
+    },
+    {
+      "epoch": 0.7406054279749478,
+      "grad_norm": 1.00180184841156,
+      "learning_rate": 4.87751105236611e-06,
+      "loss": 0.8497,
+      "step": 473
+    },
+    {
+      "epoch": 0.7421711899791231,
+      "grad_norm": 0.9382562637329102,
+      "learning_rate": 4.876858847552638e-06,
+      "loss": 0.8698,
+      "step": 474
+    },
+    {
+      "epoch": 0.7437369519832986,
+      "grad_norm": 1.0104063749313354,
+      "learning_rate": 4.8762049548213255e-06,
+      "loss": 0.8444,
+      "step": 475
+    },
+    {
+      "epoch": 0.7453027139874739,
+      "grad_norm": 0.9321765303611755,
+      "learning_rate": 4.87554937463653e-06,
+      "loss": 0.8027,
+      "step": 476
+    },
+    {
+      "epoch": 0.7468684759916493,
+      "grad_norm": 0.9527666568756104,
+      "learning_rate": 4.874892107463811e-06,
+      "loss": 0.8069,
+      "step": 477
+    },
+    {
+      "epoch": 0.7484342379958246,
+      "grad_norm": 0.9576907157897949,
+      "learning_rate": 4.874233153769924e-06,
+      "loss": 0.774,
+      "step": 478
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0516161918640137,
+      "learning_rate": 4.873572514022822e-06,
+      "loss": 0.8176,
+      "step": 479
+    },
+    {
+      "epoch": 0.7515657620041754,
+      "grad_norm": 1.026928186416626,
+      "learning_rate": 4.872910188691657e-06,
+      "loss": 0.8797,
+      "step": 480
+    },
+    {
+      "epoch": 0.7531315240083507,
+      "grad_norm": 0.9312179088592529,
+      "learning_rate": 4.872246178246776e-06,
+      "loss": 0.7747,
+      "step": 481
+    },
+    {
+      "epoch": 0.7546972860125261,
+      "grad_norm": 1.0676255226135254,
+      "learning_rate": 4.871580483159725e-06,
+      "loss": 0.7801,
+      "step": 482
+    },
+    {
+      "epoch": 0.7562630480167014,
+      "grad_norm": 1.066056489944458,
+      "learning_rate": 4.870913103903245e-06,
+      "loss": 0.7837,
+      "step": 483
+    },
+    {
+      "epoch": 0.7578288100208769,
+      "grad_norm": 0.9388124346733093,
+      "learning_rate": 4.87024404095127e-06,
+      "loss": 0.7801,
+      "step": 484
+    },
+    {
+      "epoch": 0.7593945720250522,
+      "grad_norm": 0.9879791140556335,
+      "learning_rate": 4.8695732947789356e-06,
+      "loss": 0.7849,
+      "step": 485
+    },
+    {
+      "epoch": 0.7609603340292276,
+      "grad_norm": 1.0002318620681763,
+      "learning_rate": 4.868900865862569e-06,
+      "loss": 0.8219,
+      "step": 486
+    },
+    {
+      "epoch": 0.7625260960334029,
+      "grad_norm": 0.9599125385284424,
+      "learning_rate": 4.8682267546796945e-06,
+      "loss": 0.8888,
+      "step": 487
+    },
+    {
+      "epoch": 0.7640918580375783,
+      "grad_norm": 0.9099973440170288,
+      "learning_rate": 4.867550961709027e-06,
+      "loss": 0.7585,
+      "step": 488
+    },
+    {
+      "epoch": 0.7656576200417536,
+      "grad_norm": 0.9715955257415771,
+      "learning_rate": 4.866873487430481e-06,
+      "loss": 0.8237,
+      "step": 489
+    },
+    {
+      "epoch": 0.767223382045929,
+      "grad_norm": 1.0194706916809082,
+      "learning_rate": 4.86619433232516e-06,
+      "loss": 0.7708,
+      "step": 490
+    },
+    {
+      "epoch": 0.7687891440501043,
+      "grad_norm": 1.044005036354065,
+      "learning_rate": 4.865513496875367e-06,
+      "loss": 0.8,
+      "step": 491
+    },
+    {
+      "epoch": 0.7703549060542797,
+      "grad_norm": 0.9776361584663391,
+      "learning_rate": 4.8648309815645915e-06,
+      "loss": 0.7696,
+      "step": 492
+    },
+    {
+      "epoch": 0.7719206680584552,
+      "grad_norm": 0.9213030934333801,
+      "learning_rate": 4.864146786877521e-06,
+      "loss": 0.8558,
+      "step": 493
+    },
+    {
+      "epoch": 0.7734864300626305,
+      "grad_norm": 0.9361440539360046,
+      "learning_rate": 4.8634609133000336e-06,
+      "loss": 0.7851,
+      "step": 494
+    },
+    {
+      "epoch": 0.7750521920668059,
+      "grad_norm": 1.0199540853500366,
+      "learning_rate": 4.8627733613192e-06,
+      "loss": 0.8229,
+      "step": 495
+    },
+    {
+      "epoch": 0.7766179540709812,
+      "grad_norm": 1.0244604349136353,
+      "learning_rate": 4.862084131423283e-06,
+      "loss": 0.772,
+      "step": 496
+    },
+    {
+      "epoch": 0.7781837160751566,
+      "grad_norm": 0.9414368271827698,
+      "learning_rate": 4.861393224101736e-06,
+      "loss": 0.7579,
+      "step": 497
+    },
+    {
+      "epoch": 0.7797494780793319,
+      "grad_norm": 0.9198965430259705,
+      "learning_rate": 4.860700639845205e-06,
+      "loss": 0.855,
+      "step": 498
+    },
+    {
+      "epoch": 0.7813152400835073,
+      "grad_norm": 0.9398712515830994,
+      "learning_rate": 4.8600063791455275e-06,
+      "loss": 0.7824,
+      "step": 499
+    },
+    {
+      "epoch": 0.7828810020876826,
+      "grad_norm": 0.9544969201087952,
+      "learning_rate": 4.8593104424957275e-06,
+      "loss": 0.8104,
+      "step": 500
+    },
+    {
+      "epoch": 0.784446764091858,
+      "grad_norm": 0.9522370100021362,
+      "learning_rate": 4.858612830390024e-06,
+      "loss": 0.813,
+      "step": 501
+    },
+    {
+      "epoch": 0.7860125260960334,
+      "grad_norm": 0.9830234050750732,
+      "learning_rate": 4.857913543323823e-06,
+      "loss": 0.8656,
+      "step": 502
+    },
+    {
+      "epoch": 0.7875782881002088,
+      "grad_norm": 1.0527008771896362,
+      "learning_rate": 4.8572125817937195e-06,
+      "loss": 0.8417,
+      "step": 503
+    },
+    {
+      "epoch": 0.7891440501043842,
+      "grad_norm": 0.9772782921791077,
+      "learning_rate": 4.856509946297501e-06,
+      "loss": 0.7983,
+      "step": 504
+    },
+    {
+      "epoch": 0.7907098121085595,
+      "grad_norm": 0.9250578880310059,
+      "learning_rate": 4.855805637334141e-06,
+      "loss": 0.7391,
+      "step": 505
+    },
+    {
+      "epoch": 0.7922755741127349,
+      "grad_norm": 0.9792859554290771,
+      "learning_rate": 4.855099655403802e-06,
+      "loss": 0.7753,
+      "step": 506
+    },
+    {
+      "epoch": 0.7938413361169102,
+      "grad_norm": 0.8876479864120483,
+      "learning_rate": 4.854392001007834e-06,
+      "loss": 0.8249,
+      "step": 507
+    },
+    {
+      "epoch": 0.7954070981210856,
+      "grad_norm": 0.9336771368980408,
+      "learning_rate": 4.853682674648775e-06,
+      "loss": 0.8462,
+      "step": 508
+    },
+    {
+      "epoch": 0.7969728601252609,
+      "grad_norm": 0.9732388854026794,
+      "learning_rate": 4.852971676830352e-06,
+      "loss": 0.8791,
+      "step": 509
+    },
+    {
+      "epoch": 0.7985386221294363,
+      "grad_norm": 0.9135870337486267,
+      "learning_rate": 4.852259008057476e-06,
+      "loss": 0.7279,
+      "step": 510
+    },
+    {
+      "epoch": 0.8001043841336117,
+      "grad_norm": 0.9782464504241943,
+      "learning_rate": 4.8515446688362476e-06,
+      "loss": 0.7797,
+      "step": 511
+    },
+    {
+      "epoch": 0.8016701461377871,
+      "grad_norm": 0.9571824073791504,
+      "learning_rate": 4.85082865967395e-06,
+      "loss": 0.8349,
+      "step": 512
+    },
+    {
+      "epoch": 0.8032359081419624,
+      "grad_norm": 1.011786937713623,
+      "learning_rate": 4.850110981079057e-06,
+      "loss": 0.8057,
+      "step": 513
+    },
+    {
+      "epoch": 0.8048016701461378,
+      "grad_norm": 0.9418787956237793,
+      "learning_rate": 4.8493916335612255e-06,
+      "loss": 0.8592,
+      "step": 514
+    },
+    {
+      "epoch": 0.8063674321503131,
+      "grad_norm": 0.9292604923248291,
+      "learning_rate": 4.848670617631296e-06,
+      "loss": 0.7471,
+      "step": 515
+    },
+    {
+      "epoch": 0.8079331941544885,
+      "grad_norm": 0.9303188323974609,
+      "learning_rate": 4.847947933801296e-06,
+      "loss": 0.8768,
+      "step": 516
+    },
+    {
+      "epoch": 0.8094989561586639,
+      "grad_norm": 0.9897359609603882,
+      "learning_rate": 4.847223582584437e-06,
+      "loss": 0.8343,
+      "step": 517
+    },
+    {
+      "epoch": 0.8110647181628392,
+      "grad_norm": 0.9695891737937927,
+      "learning_rate": 4.846497564495114e-06,
+      "loss": 0.8318,
+      "step": 518
+    },
+    {
+      "epoch": 0.8126304801670147,
+      "grad_norm": 1.0175273418426514,
+      "learning_rate": 4.845769880048906e-06,
+      "loss": 0.9137,
+      "step": 519
+    },
+    {
+      "epoch": 0.81419624217119,
+      "grad_norm": 0.9497922658920288,
+      "learning_rate": 4.845040529762577e-06,
+      "loss": 0.6549,
+      "step": 520
+    },
+    {
+      "epoch": 0.8157620041753654,
+      "grad_norm": 1.0647574663162231,
+      "learning_rate": 4.844309514154071e-06,
+      "loss": 0.8243,
+      "step": 521
+    },
+    {
+      "epoch": 0.8173277661795407,
+      "grad_norm": 0.9710965156555176,
+      "learning_rate": 4.843576833742517e-06,
+      "loss": 0.7864,
+      "step": 522
+    },
+    {
+      "epoch": 0.8188935281837161,
+      "grad_norm": 0.9881441593170166,
+      "learning_rate": 4.842842489048225e-06,
+      "loss": 0.7799,
+      "step": 523
+    },
+    {
+      "epoch": 0.8204592901878914,
+      "grad_norm": 0.9455597400665283,
+      "learning_rate": 4.842106480592687e-06,
+      "loss": 0.7943,
+      "step": 524
+    },
+    {
+      "epoch": 0.8220250521920668,
+      "grad_norm": 0.9948442578315735,
+      "learning_rate": 4.841368808898577e-06,
+      "loss": 0.8587,
+      "step": 525
+    },
+    {
+      "epoch": 0.8235908141962421,
+      "grad_norm": 0.9959186315536499,
+      "learning_rate": 4.84062947448975e-06,
+      "loss": 0.8477,
+      "step": 526
+    },
+    {
+      "epoch": 0.8251565762004175,
+      "grad_norm": 0.9791848659515381,
+      "learning_rate": 4.839888477891243e-06,
+      "loss": 0.7568,
+      "step": 527
+    },
+    {
+      "epoch": 0.826722338204593,
+      "grad_norm": 1.1499426364898682,
+      "learning_rate": 4.839145819629269e-06,
+      "loss": 0.8296,
+      "step": 528
+    },
+    {
+      "epoch": 0.8282881002087683,
+      "grad_norm": 0.9269506335258484,
+      "learning_rate": 4.8384015002312265e-06,
+      "loss": 0.7935,
+      "step": 529
+    },
+    {
+      "epoch": 0.8298538622129437,
+      "grad_norm": 0.970565140247345,
+      "learning_rate": 4.837655520225692e-06,
+      "loss": 0.8138,
+      "step": 530
+    },
+    {
+      "epoch": 0.831419624217119,
+      "grad_norm": 0.9852119088172913,
+      "learning_rate": 4.836907880142418e-06,
+      "loss": 0.7909,
+      "step": 531
+    },
+    {
+      "epoch": 0.8329853862212944,
+      "grad_norm": 0.927129328250885,
+      "learning_rate": 4.836158580512339e-06,
+      "loss": 0.8105,
+      "step": 532
+    },
+    {
+      "epoch": 0.8345511482254697,
+      "grad_norm": 1.0711781978607178,
+      "learning_rate": 4.835407621867569e-06,
+      "loss": 0.8241,
+      "step": 533
+    },
+    {
+      "epoch": 0.8361169102296451,
+      "grad_norm": 1.0472421646118164,
+      "learning_rate": 4.834655004741397e-06,
+      "loss": 0.7923,
+      "step": 534
+    },
+    {
+      "epoch": 0.8376826722338204,
+      "grad_norm": 0.9761579632759094,
+      "learning_rate": 4.833900729668293e-06,
+      "loss": 0.8203,
+      "step": 535
+    },
+    {
+      "epoch": 0.8392484342379958,
+      "grad_norm": 1.0559637546539307,
+      "learning_rate": 4.833144797183902e-06,
+      "loss": 0.8915,
+      "step": 536
+    },
+    {
+      "epoch": 0.8408141962421712,
+      "grad_norm": 0.9951052069664001,
+      "learning_rate": 4.832387207825045e-06,
+      "loss": 0.8253,
+      "step": 537
+    },
+    {
+      "epoch": 0.8423799582463466,
+      "grad_norm": 0.894079864025116,
+      "learning_rate": 4.831627962129725e-06,
+      "loss": 0.7479,
+      "step": 538
+    },
+    {
+      "epoch": 0.843945720250522,
+      "grad_norm": 0.9832531213760376,
+      "learning_rate": 4.830867060637115e-06,
+      "loss": 0.8484,
+      "step": 539
+    },
+    {
+      "epoch": 0.8455114822546973,
+      "grad_norm": 1.0477591753005981,
+      "learning_rate": 4.8301045038875665e-06,
+      "loss": 0.8052,
+      "step": 540
+    },
+    {
+      "epoch": 0.8470772442588727,
+      "grad_norm": 0.9388875365257263,
+      "learning_rate": 4.829340292422607e-06,
+      "loss": 0.7998,
+      "step": 541
+    },
+    {
+      "epoch": 0.848643006263048,
+      "grad_norm": 0.8975532054901123,
+      "learning_rate": 4.8285744267849395e-06,
+      "loss": 0.848,
+      "step": 542
+    },
+    {
+      "epoch": 0.8502087682672234,
+      "grad_norm": 0.9211076498031616,
+      "learning_rate": 4.8278069075184396e-06,
+      "loss": 0.8337,
+      "step": 543
+    },
+    {
+      "epoch": 0.8517745302713987,
+      "grad_norm": 0.970134437084198,
+      "learning_rate": 4.827037735168159e-06,
+      "loss": 0.7843,
+      "step": 544
+    },
+    {
+      "epoch": 0.8533402922755741,
+      "grad_norm": 0.9714353084564209,
+      "learning_rate": 4.826266910280322e-06,
+      "loss": 0.6991,
+      "step": 545
+    },
+    {
+      "epoch": 0.8549060542797495,
+      "grad_norm": 0.8607689142227173,
+      "learning_rate": 4.8254944334023265e-06,
+      "loss": 0.7314,
+      "step": 546
+    },
+    {
+      "epoch": 0.8564718162839249,
+      "grad_norm": 0.9843030571937561,
+      "learning_rate": 4.824720305082747e-06,
+      "loss": 0.7497,
+      "step": 547
+    },
+    {
+      "epoch": 0.8580375782881002,
+      "grad_norm": 0.9638474583625793,
+      "learning_rate": 4.823944525871324e-06,
+      "loss": 0.776,
+      "step": 548
+    },
+    {
+      "epoch": 0.8596033402922756,
+      "grad_norm": 0.9342365860939026,
+      "learning_rate": 4.823167096318979e-06,
+      "loss": 0.77,
+      "step": 549
+    },
+    {
+      "epoch": 0.8611691022964509,
+      "grad_norm": 0.9608579874038696,
+      "learning_rate": 4.822388016977797e-06,
+      "loss": 0.8281,
+      "step": 550
+    },
+    {
+      "epoch": 0.8627348643006263,
+      "grad_norm": 0.8996383547782898,
+      "learning_rate": 4.8216072884010395e-06,
+      "loss": 0.7519,
+      "step": 551
+    },
+    {
+      "epoch": 0.8643006263048016,
+      "grad_norm": 0.9922810196876526,
+      "learning_rate": 4.820824911143139e-06,
+      "loss": 0.8714,
+      "step": 552
+    },
+    {
+      "epoch": 0.865866388308977,
+      "grad_norm": 0.9861842393875122,
+      "learning_rate": 4.820040885759697e-06,
+      "loss": 0.7998,
+      "step": 553
+    },
+    {
+      "epoch": 0.8674321503131524,
+      "grad_norm": 1.013168215751648,
+      "learning_rate": 4.819255212807486e-06,
+      "loss": 0.8291,
+      "step": 554
+    },
+    {
+      "epoch": 0.8689979123173278,
+      "grad_norm": 0.9695484638214111,
+      "learning_rate": 4.818467892844451e-06,
+      "loss": 0.8257,
+      "step": 555
+    },
+    {
+      "epoch": 0.8705636743215032,
+      "grad_norm": 1.0018948316574097,
+      "learning_rate": 4.817678926429702e-06,
+      "loss": 0.7551,
+      "step": 556
+    },
+    {
+      "epoch": 0.8721294363256785,
+      "grad_norm": 1.0259422063827515,
+      "learning_rate": 4.816888314123521e-06,
+      "loss": 0.8582,
+      "step": 557
+    },
+    {
+      "epoch": 0.8736951983298539,
+      "grad_norm": 0.9045076966285706,
+      "learning_rate": 4.81609605648736e-06,
+      "loss": 0.811,
+      "step": 558
+    },
+    {
+      "epoch": 0.8752609603340292,
+      "grad_norm": 0.9112316370010376,
+      "learning_rate": 4.815302154083837e-06,
+      "loss": 0.818,
+      "step": 559
+    },
+    {
+      "epoch": 0.8768267223382046,
+      "grad_norm": 1.0060032606124878,
+      "learning_rate": 4.814506607476739e-06,
+      "loss": 0.766,
+      "step": 560
+    },
+    {
+      "epoch": 0.8783924843423799,
+      "grad_norm": 0.9126970171928406,
+      "learning_rate": 4.813709417231021e-06,
+      "loss": 0.8903,
+      "step": 561
+    },
+    {
+      "epoch": 0.8799582463465553,
+      "grad_norm": 1.0825906991958618,
+      "learning_rate": 4.812910583912804e-06,
+      "loss": 0.8466,
+      "step": 562
+    },
+    {
+      "epoch": 0.8815240083507306,
+      "grad_norm": 0.9530975818634033,
+      "learning_rate": 4.812110108089377e-06,
+      "loss": 0.8163,
+      "step": 563
+    },
+    {
+      "epoch": 0.8830897703549061,
+      "grad_norm": 0.9566405415534973,
+      "learning_rate": 4.8113079903291955e-06,
+      "loss": 0.8244,
+      "step": 564
+    },
+    {
+      "epoch": 0.8846555323590815,
+      "grad_norm": 0.919980525970459,
+      "learning_rate": 4.81050423120188e-06,
+      "loss": 0.8335,
+      "step": 565
+    },
+    {
+      "epoch": 0.8862212943632568,
+      "grad_norm": 0.9123868346214294,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.7712,
+      "step": 566
+    },
+    {
+      "epoch": 0.8877870563674322,
+      "grad_norm": 0.9589663743972778,
+      "learning_rate": 4.8088917911301595e-06,
+      "loss": 0.8467,
+      "step": 567
+    },
+    {
+      "epoch": 0.8893528183716075,
+      "grad_norm": 0.9052798748016357,
+      "learning_rate": 4.808083111330823e-06,
+      "loss": 0.8078,
+      "step": 568
+    },
+    {
+      "epoch": 0.8909185803757829,
+      "grad_norm": 0.9250494241714478,
+      "learning_rate": 4.807272792454489e-06,
+      "loss": 0.8536,
+      "step": 569
+    },
+    {
+      "epoch": 0.8924843423799582,
+      "grad_norm": 0.8451298475265503,
+      "learning_rate": 4.806460835076603e-06,
+      "loss": 0.7752,
+      "step": 570
+    },
+    {
+      "epoch": 0.8940501043841336,
+      "grad_norm": 1.0324240922927856,
+      "learning_rate": 4.805647239773773e-06,
+      "loss": 0.7871,
+      "step": 571
+    },
+    {
+      "epoch": 0.8956158663883089,
+      "grad_norm": 0.9338528513908386,
+      "learning_rate": 4.804832007123771e-06,
+      "loss": 0.8082,
+      "step": 572
+    },
+    {
+      "epoch": 0.8971816283924844,
+      "grad_norm": 1.0080739259719849,
+      "learning_rate": 4.804015137705531e-06,
+      "loss": 0.7636,
+      "step": 573
+    },
+    {
+      "epoch": 0.8987473903966597,
+      "grad_norm": 0.9817036986351013,
+      "learning_rate": 4.803196632099152e-06,
+      "loss": 0.8599,
+      "step": 574
+    },
+    {
+      "epoch": 0.9003131524008351,
+      "grad_norm": 0.9284219145774841,
+      "learning_rate": 4.80237649088589e-06,
+      "loss": 0.8152,
+      "step": 575
+    },
+    {
+      "epoch": 0.9018789144050104,
+      "grad_norm": 0.8922317028045654,
+      "learning_rate": 4.801554714648166e-06,
+      "loss": 0.8146,
+      "step": 576
+    },
+    {
+      "epoch": 0.9034446764091858,
+      "grad_norm": 0.9465014338493347,
+      "learning_rate": 4.800731303969565e-06,
+      "loss": 0.7663,
+      "step": 577
+    },
+    {
+      "epoch": 0.9050104384133612,
+      "grad_norm": 1.0325525999069214,
+      "learning_rate": 4.799906259434824e-06,
+      "loss": 0.8458,
+      "step": 578
+    },
+    {
+      "epoch": 0.9065762004175365,
+      "grad_norm": 0.9366307258605957,
+      "learning_rate": 4.7990795816298485e-06,
+      "loss": 0.7992,
+      "step": 579
+    },
+    {
+      "epoch": 0.9081419624217119,
+      "grad_norm": 0.9923847317695618,
+      "learning_rate": 4.7982512711416995e-06,
+      "loss": 0.7256,
+      "step": 580
+    },
+    {
+      "epoch": 0.9097077244258872,
+      "grad_norm": 1.0387133359909058,
+      "learning_rate": 4.7974213285586e-06,
+      "loss": 0.7766,
+      "step": 581
+    },
+    {
+      "epoch": 0.9112734864300627,
+      "grad_norm": 0.9617043733596802,
+      "learning_rate": 4.79658975446993e-06,
+      "loss": 0.7602,
+      "step": 582
+    },
+    {
+      "epoch": 0.912839248434238,
+      "grad_norm": 0.9530353546142578,
+      "learning_rate": 4.795756549466229e-06,
+      "loss": 0.7922,
+      "step": 583
+    },
+    {
+      "epoch": 0.9144050104384134,
+      "grad_norm": 0.9305918216705322,
+      "learning_rate": 4.794921714139194e-06,
+      "loss": 0.8377,
+      "step": 584
+    },
+    {
+      "epoch": 0.9159707724425887,
+      "grad_norm": 0.899860143661499,
+      "learning_rate": 4.794085249081682e-06,
+      "loss": 0.8343,
+      "step": 585
+    },
+    {
+      "epoch": 0.9175365344467641,
+      "grad_norm": 0.9024940729141235,
+      "learning_rate": 4.793247154887703e-06,
+      "loss": 0.8271,
+      "step": 586
+    },
+    {
+      "epoch": 0.9191022964509394,
+      "grad_norm": 0.980869472026825,
+      "learning_rate": 4.7924074321524286e-06,
+      "loss": 0.7468,
+      "step": 587
+    },
+    {
+      "epoch": 0.9206680584551148,
+      "grad_norm": 1.0506541728973389,
+      "learning_rate": 4.791566081472185e-06,
+      "loss": 0.9207,
+      "step": 588
+    },
+    {
+      "epoch": 0.9222338204592901,
+      "grad_norm": 1.0723917484283447,
+      "learning_rate": 4.790723103444454e-06,
+      "loss": 0.7615,
+      "step": 589
+    },
+    {
+      "epoch": 0.9237995824634656,
+      "grad_norm": 1.0455081462860107,
+      "learning_rate": 4.789878498667873e-06,
+      "loss": 0.8115,
+      "step": 590
+    },
+    {
+      "epoch": 0.925365344467641,
+      "grad_norm": 0.9729639291763306,
+      "learning_rate": 4.789032267742234e-06,
+      "loss": 0.8622,
+      "step": 591
+    },
+    {
+      "epoch": 0.9269311064718163,
+      "grad_norm": 0.9695239067077637,
+      "learning_rate": 4.788184411268488e-06,
+      "loss": 0.8165,
+      "step": 592
+    },
+    {
+      "epoch": 0.9284968684759917,
+      "grad_norm": 1.0099278688430786,
+      "learning_rate": 4.7873349298487345e-06,
+      "loss": 0.759,
+      "step": 593
+    },
+    {
+      "epoch": 0.930062630480167,
+      "grad_norm": 0.9377149939537048,
+      "learning_rate": 4.786483824086231e-06,
+      "loss": 0.8323,
+      "step": 594
+    },
+    {
+      "epoch": 0.9316283924843424,
+      "grad_norm": 0.8885690569877625,
+      "learning_rate": 4.785631094585387e-06,
+      "loss": 0.8448,
+      "step": 595
+    },
+    {
+      "epoch": 0.9331941544885177,
+      "grad_norm": 0.9133430123329163,
+      "learning_rate": 4.784776741951766e-06,
+      "loss": 0.8135,
+      "step": 596
+    },
+    {
+      "epoch": 0.9347599164926931,
+      "grad_norm": 1.015851378440857,
+      "learning_rate": 4.783920766792082e-06,
+      "loss": 0.8464,
+      "step": 597
+    },
+    {
+      "epoch": 0.9363256784968684,
+      "grad_norm": 0.8883825540542603,
+      "learning_rate": 4.783063169714203e-06,
+      "loss": 0.7445,
+      "step": 598
+    },
+    {
+      "epoch": 0.9378914405010439,
+      "grad_norm": 0.9872002005577087,
+      "learning_rate": 4.782203951327149e-06,
+      "loss": 0.8293,
+      "step": 599
+    },
+    {
+      "epoch": 0.9394572025052192,
+      "grad_norm": 1.0305813550949097,
+      "learning_rate": 4.781343112241091e-06,
+      "loss": 0.8606,
+      "step": 600
+    },
+    {
+      "epoch": 0.9410229645093946,
+      "grad_norm": 0.9667104482650757,
+      "learning_rate": 4.780480653067351e-06,
+      "loss": 0.8057,
+      "step": 601
+    },
+    {
+      "epoch": 0.94258872651357,
+      "grad_norm": 1.067185640335083,
+      "learning_rate": 4.7796165744184e-06,
+      "loss": 0.849,
+      "step": 602
+    },
+    {
+      "epoch": 0.9441544885177453,
+      "grad_norm": 0.9017406105995178,
+      "learning_rate": 4.778750876907862e-06,
+      "loss": 0.7927,
+      "step": 603
+    },
+    {
+      "epoch": 0.9457202505219207,
+      "grad_norm": 0.9827666878700256,
+      "learning_rate": 4.777883561150508e-06,
+      "loss": 0.8087,
+      "step": 604
+    },
+    {
+      "epoch": 0.947286012526096,
+      "grad_norm": 0.9627811908721924,
+      "learning_rate": 4.77701462776226e-06,
+      "loss": 0.8488,
+      "step": 605
+    },
+    {
+      "epoch": 0.9488517745302714,
+      "grad_norm": 1.0321338176727295,
+      "learning_rate": 4.776144077360186e-06,
+      "loss": 0.8289,
+      "step": 606
+    },
+    {
+      "epoch": 0.9504175365344467,
+      "grad_norm": 1.0012633800506592,
+      "learning_rate": 4.775271910562508e-06,
+      "loss": 0.7184,
+      "step": 607
+    },
+    {
+      "epoch": 0.9519832985386222,
+      "grad_norm": 0.9186056852340698,
+      "learning_rate": 4.7743981279885875e-06,
+      "loss": 0.8727,
+      "step": 608
+    },
+    {
+      "epoch": 0.9535490605427975,
+      "grad_norm": 0.910044252872467,
+      "learning_rate": 4.773522730258942e-06,
+      "loss": 0.8455,
+      "step": 609
+    },
+    {
+      "epoch": 0.9551148225469729,
+      "grad_norm": 0.9342944025993347,
+      "learning_rate": 4.77264571799523e-06,
+      "loss": 0.7548,
+      "step": 610
+    },
+    {
+      "epoch": 0.9566805845511482,
+      "grad_norm": 0.9609628319740295,
+      "learning_rate": 4.771767091820259e-06,
+      "loss": 0.8294,
+      "step": 611
+    },
+    {
+      "epoch": 0.9582463465553236,
+      "grad_norm": 0.9708508253097534,
+      "learning_rate": 4.770886852357983e-06,
+      "loss": 0.7857,
+      "step": 612
+    },
+    {
+      "epoch": 0.9598121085594989,
+      "grad_norm": 1.0910807847976685,
+      "learning_rate": 4.7700050002334995e-06,
+      "loss": 0.8162,
+      "step": 613
+    },
+    {
+      "epoch": 0.9613778705636743,
+      "grad_norm": 0.9703771471977234,
+      "learning_rate": 4.769121536073054e-06,
+      "loss": 0.7205,
+      "step": 614
+    },
+    {
+      "epoch": 0.9629436325678496,
+      "grad_norm": 1.027938961982727,
+      "learning_rate": 4.768236460504035e-06,
+      "loss": 0.7641,
+      "step": 615
+    },
+    {
+      "epoch": 0.964509394572025,
+      "grad_norm": 1.0003989934921265,
+      "learning_rate": 4.767349774154974e-06,
+      "loss": 0.7588,
+      "step": 616
+    },
+    {
+      "epoch": 0.9660751565762005,
+      "grad_norm": 0.9572570323944092,
+      "learning_rate": 4.766461477655552e-06,
+      "loss": 0.8274,
+      "step": 617
+    },
+    {
+      "epoch": 0.9676409185803758,
+      "grad_norm": 0.977287769317627,
+      "learning_rate": 4.765571571636586e-06,
+      "loss": 0.7856,
+      "step": 618
+    },
+    {
+      "epoch": 0.9692066805845512,
+      "grad_norm": 0.9582562446594238,
+      "learning_rate": 4.764680056730042e-06,
+      "loss": 0.8102,
+      "step": 619
+    },
+    {
+      "epoch": 0.9707724425887265,
+      "grad_norm": 0.9842986464500427,
+      "learning_rate": 4.763786933569025e-06,
+      "loss": 0.7515,
+      "step": 620
+    },
+    {
+      "epoch": 0.9723382045929019,
+      "grad_norm": 0.9955743551254272,
+      "learning_rate": 4.762892202787782e-06,
+      "loss": 0.7398,
+      "step": 621
+    },
+    {
+      "epoch": 0.9739039665970772,
+      "grad_norm": 1.0011107921600342,
+      "learning_rate": 4.761995865021706e-06,
+      "loss": 0.8182,
+      "step": 622
+    },
+    {
+      "epoch": 0.9754697286012526,
+      "grad_norm": 0.9586448669433594,
+      "learning_rate": 4.761097920907325e-06,
+      "loss": 0.8434,
+      "step": 623
+    },
+    {
+      "epoch": 0.9770354906054279,
+      "grad_norm": 0.979123055934906,
+      "learning_rate": 4.760198371082312e-06,
+      "loss": 0.7954,
+      "step": 624
+    },
+    {
+      "epoch": 0.9786012526096033,
+      "grad_norm": 1.0081250667572021,
+      "learning_rate": 4.75929721618548e-06,
+      "loss": 0.7931,
+      "step": 625
+    },
+    {
+      "epoch": 0.9801670146137788,
+      "grad_norm": 1.027185320854187,
+      "learning_rate": 4.75839445685678e-06,
+      "loss": 0.7339,
+      "step": 626
+    },
+    {
+      "epoch": 0.9817327766179541,
+      "grad_norm": 1.0553215742111206,
+      "learning_rate": 4.757490093737303e-06,
+      "loss": 0.7639,
+      "step": 627
+    },
+    {
+      "epoch": 0.9832985386221295,
+      "grad_norm": 0.8578981757164001,
+      "learning_rate": 4.75658412746928e-06,
+      "loss": 0.8128,
+      "step": 628
+    },
+    {
+      "epoch": 0.9848643006263048,
+      "grad_norm": 1.0143071413040161,
+      "learning_rate": 4.75567655869608e-06,
+      "loss": 0.8169,
+      "step": 629
+    },
+    {
+      "epoch": 0.9864300626304802,
+      "grad_norm": 0.9743554592132568,
+      "learning_rate": 4.754767388062208e-06,
+      "loss": 0.8179,
+      "step": 630
+    },
+    {
+      "epoch": 0.9879958246346555,
+      "grad_norm": 1.0276602506637573,
+      "learning_rate": 4.7538566162133105e-06,
+      "loss": 0.8009,
+      "step": 631
+    },
+    {
+      "epoch": 0.9895615866388309,
+      "grad_norm": 0.9296634793281555,
+      "learning_rate": 4.752944243796167e-06,
+      "loss": 0.8262,
+      "step": 632
+    },
+    {
+      "epoch": 0.9911273486430062,
+      "grad_norm": 0.9216791987419128,
+      "learning_rate": 4.752030271458696e-06,
+      "loss": 0.8929,
+      "step": 633
+    },
+    {
+      "epoch": 0.9926931106471816,
+      "grad_norm": 0.9272832274436951,
+      "learning_rate": 4.751114699849954e-06,
+      "loss": 0.7396,
+      "step": 634
+    },
+    {
+      "epoch": 0.994258872651357,
+      "grad_norm": 0.9121203422546387,
+      "learning_rate": 4.750197529620127e-06,
+      "loss": 0.8311,
+      "step": 635
+    },
+    {
+      "epoch": 0.9958246346555324,
+      "grad_norm": 0.999600887298584,
+      "learning_rate": 4.7492787614205425e-06,
+      "loss": 0.8125,
+      "step": 636
+    },
+    {
+      "epoch": 0.9973903966597077,
+      "grad_norm": 1.012978196144104,
+      "learning_rate": 4.748358395903661e-06,
+      "loss": 0.7966,
+      "step": 637
+    },
+    {
+      "epoch": 0.9989561586638831,
+      "grad_norm": 0.9767918586730957,
+      "learning_rate": 4.747436433723075e-06,
+      "loss": 0.8024,
+      "step": 638
+    },
+    {
+      "epoch": 1.0015657620041754,
+      "grad_norm": 1.8302738666534424,
+      "learning_rate": 4.7465128755335135e-06,
+      "loss": 1.6427,
+      "step": 639
+    },
+    {
+      "epoch": 1.0031315240083507,
+      "grad_norm": 0.8704890012741089,
+      "learning_rate": 4.7455877219908386e-06,
+      "loss": 0.7766,
+      "step": 640
+    },
+    {
+      "epoch": 1.004697286012526,
+      "grad_norm": 0.9848434925079346,
+      "learning_rate": 4.744660973752044e-06,
+      "loss": 0.8314,
+      "step": 641
+    },
+    {
+      "epoch": 1.0062630480167014,
+      "grad_norm": 1.054824709892273,
+      "learning_rate": 4.743732631475258e-06,
+      "loss": 0.797,
+      "step": 642
+    },
+    {
+      "epoch": 1.0078288100208768,
+      "grad_norm": 0.9286114573478699,
+      "learning_rate": 4.742802695819737e-06,
+      "loss": 0.764,
+      "step": 643
+    },
+    {
+      "epoch": 1.0093945720250521,
+      "grad_norm": 0.8625781536102295,
+      "learning_rate": 4.7418711674458735e-06,
+      "loss": 0.6691,
+      "step": 644
+    },
+    {
+      "epoch": 1.0109603340292275,
+      "grad_norm": 0.8710452914237976,
+      "learning_rate": 4.740938047015188e-06,
+      "loss": 0.7277,
+      "step": 645
+    },
+    {
+      "epoch": 1.0125260960334028,
+      "grad_norm": 0.8873130679130554,
+      "learning_rate": 4.740003335190334e-06,
+      "loss": 0.6531,
+      "step": 646
+    },
+    {
+      "epoch": 1.0140918580375784,
+      "grad_norm": 0.9777164459228516,
+      "learning_rate": 4.7390670326350925e-06,
+      "loss": 0.8047,
+      "step": 647
+    },
+    {
+      "epoch": 1.0156576200417538,
+      "grad_norm": 0.886025071144104,
+      "learning_rate": 4.738129140014377e-06,
+      "loss": 0.715,
+      "step": 648
+    },
+    {
+      "epoch": 1.017223382045929,
+      "grad_norm": 0.9825332164764404,
+      "learning_rate": 4.737189657994226e-06,
+      "loss": 0.7783,
+      "step": 649
+    },
+    {
+      "epoch": 1.0187891440501045,
+      "grad_norm": 0.9331428408622742,
+      "learning_rate": 4.736248587241812e-06,
+      "loss": 0.7598,
+      "step": 650
+    },
+    {
+      "epoch": 1.0203549060542798,
+      "grad_norm": 0.9406535625457764,
+      "learning_rate": 4.735305928425431e-06,
+      "loss": 0.7989,
+      "step": 651
+    },
+    {
+      "epoch": 1.0219206680584552,
+      "grad_norm": 1.0644745826721191,
+      "learning_rate": 4.734361682214511e-06,
+      "loss": 0.8135,
+      "step": 652
+    },
+    {
+      "epoch": 1.0234864300626305,
+      "grad_norm": 0.9037052392959595,
+      "learning_rate": 4.733415849279603e-06,
+      "loss": 0.7051,
+      "step": 653
+    },
+    {
+      "epoch": 1.0250521920668059,
+      "grad_norm": 1.000563144683838,
+      "learning_rate": 4.7324684302923865e-06,
+      "loss": 0.7427,
+      "step": 654
+    },
+    {
+      "epoch": 1.0266179540709812,
+      "grad_norm": 0.9181354641914368,
+      "learning_rate": 4.731519425925671e-06,
+      "loss": 0.7236,
+      "step": 655
+    },
+    {
+      "epoch": 1.0281837160751566,
+      "grad_norm": 1.0283046960830688,
+      "learning_rate": 4.730568836853384e-06,
+      "loss": 0.7998,
+      "step": 656
+    },
+    {
+      "epoch": 1.029749478079332,
+      "grad_norm": 0.8692315220832825,
+      "learning_rate": 4.729616663750587e-06,
+      "loss": 0.7974,
+      "step": 657
+    },
+    {
+      "epoch": 1.0313152400835073,
+      "grad_norm": 0.928920567035675,
+      "learning_rate": 4.72866290729346e-06,
+      "loss": 0.7199,
+      "step": 658
+    },
+    {
+      "epoch": 1.0328810020876826,
+      "grad_norm": 0.9066957235336304,
+      "learning_rate": 4.727707568159311e-06,
+      "loss": 0.791,
+      "step": 659
+    },
+    {
+      "epoch": 1.034446764091858,
+      "grad_norm": 1.0086729526519775,
+      "learning_rate": 4.726750647026569e-06,
+      "loss": 0.7113,
+      "step": 660
+    },
+    {
+      "epoch": 1.0360125260960333,
+      "grad_norm": 0.931856632232666,
+      "learning_rate": 4.72579214457479e-06,
+      "loss": 0.7977,
+      "step": 661
+    },
+    {
+      "epoch": 1.0375782881002087,
+      "grad_norm": 0.9191203713417053,
+      "learning_rate": 4.724832061484651e-06,
+      "loss": 0.7979,
+      "step": 662
+    },
+    {
+      "epoch": 1.039144050104384,
+      "grad_norm": 0.9614707827568054,
+      "learning_rate": 4.72387039843795e-06,
+      "loss": 0.6994,
+      "step": 663
+    },
+    {
+      "epoch": 1.0407098121085594,
+      "grad_norm": 1.014823317527771,
+      "learning_rate": 4.72290715611761e-06,
+      "loss": 0.7539,
+      "step": 664
+    },
+    {
+      "epoch": 1.042275574112735,
+      "grad_norm": 0.9177424311637878,
+      "learning_rate": 4.721942335207673e-06,
+      "loss": 0.6625,
+      "step": 665
+    },
+    {
+      "epoch": 1.0438413361169103,
+      "grad_norm": 1.0253031253814697,
+      "learning_rate": 4.720975936393305e-06,
+      "loss": 0.7747,
+      "step": 666
+    },
+    {
+      "epoch": 1.0454070981210857,
+      "grad_norm": 1.0004568099975586,
+      "learning_rate": 4.720007960360788e-06,
+      "loss": 0.7008,
+      "step": 667
+    },
+    {
+      "epoch": 1.046972860125261,
+      "grad_norm": 1.1344349384307861,
+      "learning_rate": 4.719038407797529e-06,
+      "loss": 0.7493,
+      "step": 668
+    },
+    {
+      "epoch": 1.0485386221294364,
+      "grad_norm": 0.9944799542427063,
+      "learning_rate": 4.718067279392052e-06,
+      "loss": 0.7823,
+      "step": 669
+    },
+    {
+      "epoch": 1.0501043841336117,
+      "grad_norm": 0.9469346404075623,
+      "learning_rate": 4.717094575833999e-06,
+      "loss": 0.7419,
+      "step": 670
+    },
+    {
+      "epoch": 1.051670146137787,
+      "grad_norm": 1.015339732170105,
+      "learning_rate": 4.716120297814133e-06,
+      "loss": 0.7168,
+      "step": 671
+    },
+    {
+      "epoch": 1.0532359081419624,
+      "grad_norm": 0.9544686675071716,
+      "learning_rate": 4.715144446024333e-06,
+      "loss": 0.791,
+      "step": 672
+    },
+    {
+      "epoch": 1.0548016701461378,
+      "grad_norm": 1.0285652875900269,
+      "learning_rate": 4.714167021157599e-06,
+      "loss": 0.7649,
+      "step": 673
+    },
+    {
+      "epoch": 1.0563674321503131,
+      "grad_norm": 0.9366872906684875,
+      "learning_rate": 4.713188023908044e-06,
+      "loss": 0.7524,
+      "step": 674
+    },
+    {
+      "epoch": 1.0579331941544885,
+      "grad_norm": 1.0348883867263794,
+      "learning_rate": 4.712207454970899e-06,
+      "loss": 0.6965,
+      "step": 675
+    },
+    {
+      "epoch": 1.0594989561586639,
+      "grad_norm": 0.8716043829917908,
+      "learning_rate": 4.711225315042513e-06,
+      "loss": 0.6264,
+      "step": 676
+    },
+    {
+      "epoch": 1.0610647181628392,
+      "grad_norm": 1.0421313047409058,
+      "learning_rate": 4.710241604820348e-06,
+      "loss": 0.7548,
+      "step": 677
+    },
+    {
+      "epoch": 1.0626304801670146,
+      "grad_norm": 1.0886270999908447,
+      "learning_rate": 4.709256325002983e-06,
+      "loss": 0.6352,
+      "step": 678
+    },
+    {
+      "epoch": 1.06419624217119,
+      "grad_norm": 0.9752539396286011,
+      "learning_rate": 4.708269476290111e-06,
+      "loss": 0.7764,
+      "step": 679
+    },
+    {
+      "epoch": 1.0657620041753653,
+      "grad_norm": 1.0658339262008667,
+      "learning_rate": 4.7072810593825385e-06,
+      "loss": 0.766,
+      "step": 680
+    },
+    {
+      "epoch": 1.0673277661795406,
+      "grad_norm": 1.0299617052078247,
+      "learning_rate": 4.706291074982187e-06,
+      "loss": 0.7558,
+      "step": 681
+    },
+    {
+      "epoch": 1.068893528183716,
+      "grad_norm": 0.9171326160430908,
+      "learning_rate": 4.705299523792089e-06,
+      "loss": 0.7084,
+      "step": 682
+    },
+    {
+      "epoch": 1.0704592901878915,
+      "grad_norm": 0.9924243688583374,
+      "learning_rate": 4.7043064065163925e-06,
+      "loss": 0.8393,
+      "step": 683
+    },
+    {
+      "epoch": 1.072025052192067,
+      "grad_norm": 1.0243877172470093,
+      "learning_rate": 4.703311723860356e-06,
+      "loss": 0.7755,
+      "step": 684
+    },
+    {
+      "epoch": 1.0735908141962422,
+      "grad_norm": 1.0019137859344482,
+      "learning_rate": 4.702315476530349e-06,
+      "loss": 0.7254,
+      "step": 685
+    },
+    {
+      "epoch": 1.0751565762004176,
+      "grad_norm": 0.8541996479034424,
+      "learning_rate": 4.701317665233853e-06,
+      "loss": 0.7389,
+      "step": 686
+    },
+    {
+      "epoch": 1.076722338204593,
+      "grad_norm": 1.1119922399520874,
+      "learning_rate": 4.700318290679462e-06,
+      "loss": 0.7803,
+      "step": 687
+    },
+    {
+      "epoch": 1.0782881002087683,
+      "grad_norm": 1.078452706336975,
+      "learning_rate": 4.699317353576876e-06,
+      "loss": 0.756,
+      "step": 688
+    },
+    {
+      "epoch": 1.0798538622129437,
+      "grad_norm": 1.0079457759857178,
+      "learning_rate": 4.698314854636908e-06,
+      "loss": 0.7365,
+      "step": 689
+    },
+    {
+      "epoch": 1.081419624217119,
+      "grad_norm": 0.931423544883728,
+      "learning_rate": 4.6973107945714795e-06,
+      "loss": 0.7881,
+      "step": 690
+    },
+    {
+      "epoch": 1.0829853862212944,
+      "grad_norm": 1.0135142803192139,
+      "learning_rate": 4.696305174093619e-06,
+      "loss": 0.7884,
+      "step": 691
+    },
+    {
+      "epoch": 1.0845511482254697,
+      "grad_norm": 1.1194610595703125,
+      "learning_rate": 4.695297993917465e-06,
+      "loss": 0.7969,
+      "step": 692
+    },
+    {
+      "epoch": 1.086116910229645,
+      "grad_norm": 0.9972780346870422,
+      "learning_rate": 4.6942892547582625e-06,
+      "loss": 0.5873,
+      "step": 693
+    },
+    {
+      "epoch": 1.0876826722338204,
+      "grad_norm": 1.0711641311645508,
+      "learning_rate": 4.693278957332364e-06,
+      "loss": 0.7273,
+      "step": 694
+    },
+    {
+      "epoch": 1.0892484342379958,
+      "grad_norm": 0.9708715081214905,
+      "learning_rate": 4.6922671023572305e-06,
+      "loss": 0.6786,
+      "step": 695
+    },
+    {
+      "epoch": 1.0908141962421711,
+      "grad_norm": 0.9773644804954529,
+      "learning_rate": 4.691253690551424e-06,
+      "loss": 0.7839,
+      "step": 696
+    },
+    {
+      "epoch": 1.0923799582463465,
+      "grad_norm": 1.039009690284729,
+      "learning_rate": 4.690238722634618e-06,
+      "loss": 0.6807,
+      "step": 697
+    },
+    {
+      "epoch": 1.0939457202505218,
+      "grad_norm": 1.140777349472046,
+      "learning_rate": 4.689222199327586e-06,
+      "loss": 0.7506,
+      "step": 698
+    },
+    {
+      "epoch": 1.0955114822546972,
+      "grad_norm": 0.9779882431030273,
+      "learning_rate": 4.688204121352211e-06,
+      "loss": 0.7732,
+      "step": 699
+    },
+    {
+      "epoch": 1.0970772442588728,
+      "grad_norm": 1.0412248373031616,
+      "learning_rate": 4.687184489431476e-06,
+      "loss": 0.7034,
+      "step": 700
+    },
+    {
+      "epoch": 1.0986430062630481,
+      "grad_norm": 0.9802400469779968,
+      "learning_rate": 4.686163304289469e-06,
+      "loss": 0.729,
+      "step": 701
+    },
+    {
+      "epoch": 1.1002087682672235,
+      "grad_norm": 1.0531584024429321,
+      "learning_rate": 4.6851405666513804e-06,
+      "loss": 0.7635,
+      "step": 702
+    },
+    {
+      "epoch": 1.1017745302713988,
+      "grad_norm": 0.983923077583313,
+      "learning_rate": 4.684116277243505e-06,
+      "loss": 0.7062,
+      "step": 703
+    },
+    {
+      "epoch": 1.1033402922755742,
+      "grad_norm": 0.9886301159858704,
+      "learning_rate": 4.683090436793238e-06,
+      "loss": 0.7299,
+      "step": 704
+    },
+    {
+      "epoch": 1.1049060542797495,
+      "grad_norm": 1.0247817039489746,
+      "learning_rate": 4.682063046029075e-06,
+      "loss": 0.729,
+      "step": 705
+    },
+    {
+      "epoch": 1.1064718162839249,
+      "grad_norm": 0.9860950708389282,
+      "learning_rate": 4.681034105680616e-06,
+      "loss": 0.7392,
+      "step": 706
+    },
+    {
+      "epoch": 1.1080375782881002,
+      "grad_norm": 0.9543380737304688,
+      "learning_rate": 4.680003616478556e-06,
+      "loss": 0.7061,
+      "step": 707
+    },
+    {
+      "epoch": 1.1096033402922756,
+      "grad_norm": 1.0253225564956665,
+      "learning_rate": 4.678971579154698e-06,
+      "loss": 0.6577,
+      "step": 708
+    },
+    {
+      "epoch": 1.111169102296451,
+      "grad_norm": 0.9013616442680359,
+      "learning_rate": 4.677937994441935e-06,
+      "loss": 0.7422,
+      "step": 709
+    },
+    {
+      "epoch": 1.1127348643006263,
+      "grad_norm": 0.9716471433639526,
+      "learning_rate": 4.676902863074268e-06,
+      "loss": 0.7353,
+      "step": 710
+    },
+    {
+      "epoch": 1.1143006263048016,
+      "grad_norm": 1.040898084640503,
+      "learning_rate": 4.675866185786787e-06,
+      "loss": 0.7396,
+      "step": 711
+    },
+    {
+      "epoch": 1.115866388308977,
+      "grad_norm": 0.9931808114051819,
+      "learning_rate": 4.674827963315688e-06,
+      "loss": 0.7354,
+      "step": 712
+    },
+    {
+      "epoch": 1.1174321503131524,
+      "grad_norm": 1.020512580871582,
+      "learning_rate": 4.6737881963982605e-06,
+      "loss": 0.7366,
+      "step": 713
+    },
+    {
+      "epoch": 1.1189979123173277,
+      "grad_norm": 1.0201045274734497,
+      "learning_rate": 4.67274688577289e-06,
+      "loss": 0.7907,
+      "step": 714
+    },
+    {
+      "epoch": 1.120563674321503,
+      "grad_norm": 1.0163429975509644,
+      "learning_rate": 4.671704032179061e-06,
+      "loss": 0.7452,
+      "step": 715
+    },
+    {
+      "epoch": 1.1221294363256784,
+      "grad_norm": 0.9667377471923828,
+      "learning_rate": 4.670659636357352e-06,
+      "loss": 0.7983,
+      "step": 716
+    },
+    {
+      "epoch": 1.1236951983298538,
+      "grad_norm": 1.0788301229476929,
+      "learning_rate": 4.669613699049436e-06,
+      "loss": 0.7861,
+      "step": 717
+    },
+    {
+      "epoch": 1.1252609603340291,
+      "grad_norm": 1.0512142181396484,
+      "learning_rate": 4.668566220998082e-06,
+      "loss": 0.7582,
+      "step": 718
+    },
+    {
+      "epoch": 1.1268267223382047,
+      "grad_norm": 0.9225794076919556,
+      "learning_rate": 4.667517202947154e-06,
+      "loss": 0.7801,
+      "step": 719
+    },
+    {
+      "epoch": 1.12839248434238,
+      "grad_norm": 1.1193747520446777,
+      "learning_rate": 4.666466645641607e-06,
+      "loss": 0.8073,
+      "step": 720
+    },
+    {
+      "epoch": 1.1299582463465554,
+      "grad_norm": 0.934438943862915,
+      "learning_rate": 4.6654145498274915e-06,
+      "loss": 0.774,
+      "step": 721
+    },
+    {
+      "epoch": 1.1315240083507307,
+      "grad_norm": 1.0614968538284302,
+      "learning_rate": 4.664360916251949e-06,
+      "loss": 0.7216,
+      "step": 722
+    },
+    {
+      "epoch": 1.133089770354906,
+      "grad_norm": 1.0807424783706665,
+      "learning_rate": 4.663305745663212e-06,
+      "loss": 0.6434,
+      "step": 723
+    },
+    {
+      "epoch": 1.1346555323590815,
+      "grad_norm": 0.9515354633331299,
+      "learning_rate": 4.66224903881061e-06,
+      "loss": 0.7336,
+      "step": 724
+    },
+    {
+      "epoch": 1.1362212943632568,
+      "grad_norm": 0.9852489233016968,
+      "learning_rate": 4.661190796444557e-06,
+      "loss": 0.8156,
+      "step": 725
+    },
+    {
+      "epoch": 1.1377870563674322,
+      "grad_norm": 1.0071501731872559,
+      "learning_rate": 4.66013101931656e-06,
+      "loss": 0.8836,
+      "step": 726
+    },
+    {
+      "epoch": 1.1393528183716075,
+      "grad_norm": 0.9622960686683655,
+      "learning_rate": 4.6590697081792176e-06,
+      "loss": 0.7631,
+      "step": 727
+    },
+    {
+      "epoch": 1.1409185803757829,
+      "grad_norm": 1.0588105916976929,
+      "learning_rate": 4.6580068637862144e-06,
+      "loss": 0.7874,
+      "step": 728
+    },
+    {
+      "epoch": 1.1424843423799582,
+      "grad_norm": 0.9920447468757629,
+      "learning_rate": 4.656942486892326e-06,
+      "loss": 0.7161,
+      "step": 729
+    },
+    {
+      "epoch": 1.1440501043841336,
+      "grad_norm": 1.0036561489105225,
+      "learning_rate": 4.655876578253416e-06,
+      "loss": 0.7116,
+      "step": 730
+    },
+    {
+      "epoch": 1.145615866388309,
+      "grad_norm": 1.0208396911621094,
+      "learning_rate": 4.654809138626436e-06,
+      "loss": 0.818,
+      "step": 731
+    },
+    {
+      "epoch": 1.1471816283924843,
+      "grad_norm": 1.1095491647720337,
+      "learning_rate": 4.653740168769424e-06,
+      "loss": 0.7319,
+      "step": 732
+    },
+    {
+      "epoch": 1.1487473903966596,
+      "grad_norm": 0.8151328563690186,
+      "learning_rate": 4.652669669441505e-06,
+      "loss": 0.774,
+      "step": 733
+    },
+    {
+      "epoch": 1.150313152400835,
+      "grad_norm": 1.0747915506362915,
+      "learning_rate": 4.651597641402891e-06,
+      "loss": 0.8275,
+      "step": 734
+    },
+    {
+      "epoch": 1.1518789144050103,
+      "grad_norm": 0.9265272617340088,
+      "learning_rate": 4.650524085414878e-06,
+      "loss": 0.6946,
+      "step": 735
+    },
+    {
+      "epoch": 1.153444676409186,
+      "grad_norm": 0.9531672596931458,
+      "learning_rate": 4.649449002239849e-06,
+      "loss": 0.7899,
+      "step": 736
+    },
+    {
+      "epoch": 1.1550104384133613,
+      "grad_norm": 1.0496262311935425,
+      "learning_rate": 4.648372392641269e-06,
+      "loss": 0.8237,
+      "step": 737
+    },
+    {
+      "epoch": 1.1565762004175366,
+      "grad_norm": 0.8967193961143494,
+      "learning_rate": 4.64729425738369e-06,
+      "loss": 0.736,
+      "step": 738
+    },
+    {
+      "epoch": 1.158141962421712,
+      "grad_norm": 1.061038851737976,
+      "learning_rate": 4.646214597232745e-06,
+      "loss": 0.7308,
+      "step": 739
+    },
+    {
+      "epoch": 1.1597077244258873,
+      "grad_norm": 0.9556939601898193,
+      "learning_rate": 4.64513341295515e-06,
+      "loss": 0.7526,
+      "step": 740
+    },
+    {
+      "epoch": 1.1612734864300627,
+      "grad_norm": 1.000093936920166,
+      "learning_rate": 4.6440507053187075e-06,
+      "loss": 0.7488,
+      "step": 741
+    },
+    {
+      "epoch": 1.162839248434238,
+      "grad_norm": 0.9552261233329773,
+      "learning_rate": 4.642966475092294e-06,
+      "loss": 0.8291,
+      "step": 742
+    },
+    {
+      "epoch": 1.1644050104384134,
+      "grad_norm": 1.0215847492218018,
+      "learning_rate": 4.641880723045875e-06,
+      "loss": 0.756,
+      "step": 743
+    },
+    {
+      "epoch": 1.1659707724425887,
+      "grad_norm": 1.0543124675750732,
+      "learning_rate": 4.640793449950493e-06,
+      "loss": 0.7668,
+      "step": 744
+    },
+    {
+      "epoch": 1.167536534446764,
+      "grad_norm": 0.9648609161376953,
+      "learning_rate": 4.639704656578269e-06,
+      "loss": 0.6531,
+      "step": 745
+    },
+    {
+      "epoch": 1.1691022964509394,
+      "grad_norm": 1.0117427110671997,
+      "learning_rate": 4.638614343702408e-06,
+      "loss": 0.7084,
+      "step": 746
+    },
+    {
+      "epoch": 1.1706680584551148,
+      "grad_norm": 0.9475041627883911,
+      "learning_rate": 4.6375225120971915e-06,
+      "loss": 0.8578,
+      "step": 747
+    },
+    {
+      "epoch": 1.1722338204592901,
+      "grad_norm": 1.0122299194335938,
+      "learning_rate": 4.6364291625379785e-06,
+      "loss": 0.7541,
+      "step": 748
+    },
+    {
+      "epoch": 1.1737995824634655,
+      "grad_norm": 0.9487535357475281,
+      "learning_rate": 4.635334295801208e-06,
+      "loss": 0.7203,
+      "step": 749
+    },
+    {
+      "epoch": 1.1753653444676408,
+      "grad_norm": 1.0628844499588013,
+      "learning_rate": 4.6342379126643965e-06,
+      "loss": 0.7532,
+      "step": 750
+    },
+    {
+      "epoch": 1.1769311064718162,
+      "grad_norm": 0.9647781848907471,
+      "learning_rate": 4.633140013906136e-06,
+      "loss": 0.5918,
+      "step": 751
+    },
+    {
+      "epoch": 1.1784968684759916,
+      "grad_norm": 0.9011932611465454,
+      "learning_rate": 4.632040600306095e-06,
+      "loss": 0.7395,
+      "step": 752
+    },
+    {
+      "epoch": 1.1800626304801671,
+      "grad_norm": 1.0524142980575562,
+      "learning_rate": 4.630939672645018e-06,
+      "loss": 0.7487,
+      "step": 753
+    },
+    {
+      "epoch": 1.1816283924843423,
+      "grad_norm": 1.165077567100525,
+      "learning_rate": 4.6298372317047265e-06,
+      "loss": 0.6817,
+      "step": 754
+    },
+    {
+      "epoch": 1.1831941544885178,
+      "grad_norm": 0.9301269054412842,
+      "learning_rate": 4.628733278268113e-06,
+      "loss": 0.8446,
+      "step": 755
+    },
+    {
+      "epoch": 1.1847599164926932,
+      "grad_norm": 0.9762471318244934,
+      "learning_rate": 4.627627813119147e-06,
+      "loss": 0.7527,
+      "step": 756
+    },
+    {
+      "epoch": 1.1863256784968685,
+      "grad_norm": 0.8634734153747559,
+      "learning_rate": 4.626520837042871e-06,
+      "loss": 0.7177,
+      "step": 757
+    },
+    {
+      "epoch": 1.187891440501044,
+      "grad_norm": 1.0592188835144043,
+      "learning_rate": 4.625412350825399e-06,
+      "loss": 0.7293,
+      "step": 758
+    },
+    {
+      "epoch": 1.1894572025052192,
+      "grad_norm": 1.1318559646606445,
+      "learning_rate": 4.62430235525392e-06,
+      "loss": 0.6592,
+      "step": 759
+    },
+    {
+      "epoch": 1.1910229645093946,
+      "grad_norm": 0.9689339399337769,
+      "learning_rate": 4.623190851116692e-06,
+      "loss": 0.7667,
+      "step": 760
+    },
+    {
+      "epoch": 1.19258872651357,
+      "grad_norm": 0.9965885281562805,
+      "learning_rate": 4.6220778392030464e-06,
+      "loss": 0.6982,
+      "step": 761
+    },
+    {
+      "epoch": 1.1941544885177453,
+      "grad_norm": 1.1428629159927368,
+      "learning_rate": 4.6209633203033846e-06,
+      "loss": 0.7342,
+      "step": 762
+    },
+    {
+      "epoch": 1.1957202505219207,
+      "grad_norm": 0.901360809803009,
+      "learning_rate": 4.6198472952091785e-06,
+      "loss": 0.8371,
+      "step": 763
+    },
+    {
+      "epoch": 1.197286012526096,
+      "grad_norm": 0.9882062077522278,
+      "learning_rate": 4.618729764712969e-06,
+      "loss": 0.6882,
+      "step": 764
+    },
+    {
+      "epoch": 1.1988517745302714,
+      "grad_norm": 1.0692517757415771,
+      "learning_rate": 4.617610729608366e-06,
+      "loss": 0.7327,
+      "step": 765
+    },
+    {
+      "epoch": 1.2004175365344467,
+      "grad_norm": 1.1056455373764038,
+      "learning_rate": 4.616490190690049e-06,
+      "loss": 0.781,
+      "step": 766
+    },
+    {
+      "epoch": 1.201983298538622,
+      "grad_norm": 0.9928885698318481,
+      "learning_rate": 4.615368148753765e-06,
+      "loss": 0.7463,
+      "step": 767
+    },
+    {
+      "epoch": 1.2035490605427974,
+      "grad_norm": 1.0006698369979858,
+      "learning_rate": 4.614244604596327e-06,
+      "loss": 0.6994,
+      "step": 768
+    },
+    {
+      "epoch": 1.2051148225469728,
+      "grad_norm": 0.907508909702301,
+      "learning_rate": 4.613119559015616e-06,
+      "loss": 0.7117,
+      "step": 769
+    },
+    {
+      "epoch": 1.2066805845511483,
+      "grad_norm": 1.1037627458572388,
+      "learning_rate": 4.61199301281058e-06,
+      "loss": 0.772,
+      "step": 770
+    },
+    {
+      "epoch": 1.2082463465553235,
+      "grad_norm": 0.9474150538444519,
+      "learning_rate": 4.610864966781231e-06,
+      "loss": 0.7601,
+      "step": 771
+    },
+    {
+      "epoch": 1.209812108559499,
+      "grad_norm": 1.003667950630188,
+      "learning_rate": 4.609735421728647e-06,
+      "loss": 0.7141,
+      "step": 772
+    },
+    {
+      "epoch": 1.2113778705636744,
+      "grad_norm": 0.9724787473678589,
+      "learning_rate": 4.608604378454971e-06,
+      "loss": 0.7841,
+      "step": 773
+    },
+    {
+      "epoch": 1.2129436325678498,
+      "grad_norm": 1.0460760593414307,
+      "learning_rate": 4.607471837763409e-06,
+      "loss": 0.7563,
+      "step": 774
+    },
+    {
+      "epoch": 1.2145093945720251,
+      "grad_norm": 0.9798458218574524,
+      "learning_rate": 4.606337800458231e-06,
+      "loss": 0.6812,
+      "step": 775
+    },
+    {
+      "epoch": 1.2160751565762005,
+      "grad_norm": 1.1637914180755615,
+      "learning_rate": 4.605202267344768e-06,
+      "loss": 0.7571,
+      "step": 776
+    },
+    {
+      "epoch": 1.2176409185803758,
+      "grad_norm": 0.9384608864784241,
+      "learning_rate": 4.6040652392294175e-06,
+      "loss": 0.7842,
+      "step": 777
+    },
+    {
+      "epoch": 1.2192066805845512,
+      "grad_norm": 1.0490702390670776,
+      "learning_rate": 4.602926716919633e-06,
+      "loss": 0.7157,
+      "step": 778
+    },
+    {
+      "epoch": 1.2207724425887265,
+      "grad_norm": 1.131402611732483,
+      "learning_rate": 4.601786701223934e-06,
+      "loss": 0.7724,
+      "step": 779
+    },
+    {
+      "epoch": 1.2223382045929019,
+      "grad_norm": 1.0159722566604614,
+      "learning_rate": 4.600645192951898e-06,
+      "loss": 0.7666,
+      "step": 780
+    },
+    {
+      "epoch": 1.2239039665970772,
+      "grad_norm": 1.039806604385376,
+      "learning_rate": 4.599502192914164e-06,
+      "loss": 0.7845,
+      "step": 781
+    },
+    {
+      "epoch": 1.2254697286012526,
+      "grad_norm": 1.0722614526748657,
+      "learning_rate": 4.598357701922428e-06,
+      "loss": 0.782,
+      "step": 782
+    },
+    {
+      "epoch": 1.227035490605428,
+      "grad_norm": 0.9443033933639526,
+      "learning_rate": 4.5972117207894465e-06,
+      "loss": 0.6573,
+      "step": 783
+    },
+    {
+      "epoch": 1.2286012526096033,
+      "grad_norm": 0.9738026857376099,
+      "learning_rate": 4.596064250329035e-06,
+      "loss": 0.7475,
+      "step": 784
+    },
+    {
+      "epoch": 1.2301670146137786,
+      "grad_norm": 0.9932008385658264,
+      "learning_rate": 4.594915291356063e-06,
+      "loss": 0.757,
+      "step": 785
+    },
+    {
+      "epoch": 1.231732776617954,
+      "grad_norm": 1.0612564086914062,
+      "learning_rate": 4.5937648446864625e-06,
+      "loss": 0.7824,
+      "step": 786
+    },
+    {
+      "epoch": 1.2332985386221293,
+      "grad_norm": 1.0500893592834473,
+      "learning_rate": 4.592612911137218e-06,
+      "loss": 0.7646,
+      "step": 787
+    },
+    {
+      "epoch": 1.2348643006263047,
+      "grad_norm": 0.9607855081558228,
+      "learning_rate": 4.591459491526371e-06,
+      "loss": 0.754,
+      "step": 788
+    },
+    {
+      "epoch": 1.2364300626304803,
+      "grad_norm": 1.0871951580047607,
+      "learning_rate": 4.590304586673018e-06,
+      "loss": 0.7226,
+      "step": 789
+    },
+    {
+      "epoch": 1.2379958246346556,
+      "grad_norm": 1.0030092000961304,
+      "learning_rate": 4.589148197397311e-06,
+      "loss": 0.7357,
+      "step": 790
+    },
+    {
+      "epoch": 1.239561586638831,
+      "grad_norm": 1.0297991037368774,
+      "learning_rate": 4.587990324520455e-06,
+      "loss": 0.7724,
+      "step": 791
+    },
+    {
+      "epoch": 1.2411273486430063,
+      "grad_norm": 1.0192303657531738,
+      "learning_rate": 4.586830968864711e-06,
+      "loss": 0.6921,
+      "step": 792
+    },
+    {
+      "epoch": 1.2426931106471817,
+      "grad_norm": 0.9617307186126709,
+      "learning_rate": 4.585670131253389e-06,
+      "loss": 0.7443,
+      "step": 793
+    },
+    {
+      "epoch": 1.244258872651357,
+      "grad_norm": 1.0355238914489746,
+      "learning_rate": 4.584507812510856e-06,
+      "loss": 0.7741,
+      "step": 794
+    },
+    {
+      "epoch": 1.2458246346555324,
+      "grad_norm": 0.9938539862632751,
+      "learning_rate": 4.583344013462526e-06,
+      "loss": 0.7372,
+      "step": 795
+    },
+    {
+      "epoch": 1.2473903966597077,
+      "grad_norm": 1.0573335886001587,
+      "learning_rate": 4.582178734934869e-06,
+      "loss": 0.7191,
+      "step": 796
+    },
+    {
+      "epoch": 1.248956158663883,
+      "grad_norm": 0.963642954826355,
+      "learning_rate": 4.581011977755402e-06,
+      "loss": 0.6283,
+      "step": 797
+    },
+    {
+      "epoch": 1.2505219206680585,
+      "grad_norm": 1.0119287967681885,
+      "learning_rate": 4.579843742752694e-06,
+      "loss": 0.7599,
+      "step": 798
+    },
+    {
+      "epoch": 1.2520876826722338,
+      "grad_norm": 1.0628082752227783,
+      "learning_rate": 4.578674030756364e-06,
+      "loss": 0.7741,
+      "step": 799
+    },
+    {
+      "epoch": 1.2536534446764092,
+      "grad_norm": 1.0510241985321045,
+      "learning_rate": 4.577502842597077e-06,
+      "loss": 0.7615,
+      "step": 800
+    },
+    {
+      "epoch": 1.2552192066805845,
+      "grad_norm": 1.1180559396743774,
+      "learning_rate": 4.5763301791065474e-06,
+      "loss": 0.7119,
+      "step": 801
+    },
+    {
+      "epoch": 1.2567849686847599,
+      "grad_norm": 1.0198049545288086,
+      "learning_rate": 4.575156041117541e-06,
+      "loss": 0.7487,
+      "step": 802
+    },
+    {
+      "epoch": 1.2583507306889352,
+      "grad_norm": 1.018902063369751,
+      "learning_rate": 4.5739804294638655e-06,
+      "loss": 0.7461,
+      "step": 803
+    },
+    {
+      "epoch": 1.2599164926931106,
+      "grad_norm": 0.9230141639709473,
+      "learning_rate": 4.572803344980378e-06,
+      "loss": 0.7295,
+      "step": 804
+    },
+    {
+      "epoch": 1.261482254697286,
+      "grad_norm": 1.0011619329452515,
+      "learning_rate": 4.571624788502981e-06,
+      "loss": 0.7325,
+      "step": 805
+    },
+    {
+      "epoch": 1.2630480167014615,
+      "grad_norm": 0.9569756388664246,
+      "learning_rate": 4.570444760868622e-06,
+      "loss": 0.7494,
+      "step": 806
+    },
+    {
+      "epoch": 1.2646137787056366,
+      "grad_norm": 0.9860491156578064,
+      "learning_rate": 4.569263262915293e-06,
+      "loss": 0.6536,
+      "step": 807
+    },
+    {
+      "epoch": 1.2661795407098122,
+      "grad_norm": 1.0986074209213257,
+      "learning_rate": 4.568080295482032e-06,
+      "loss": 0.7337,
+      "step": 808
+    },
+    {
+      "epoch": 1.2677453027139876,
+      "grad_norm": 1.0782687664031982,
+      "learning_rate": 4.566895859408916e-06,
+      "loss": 0.7193,
+      "step": 809
+    },
+    {
+      "epoch": 1.269311064718163,
+      "grad_norm": 1.1424410343170166,
+      "learning_rate": 4.565709955537071e-06,
+      "loss": 0.6153,
+      "step": 810
+    },
+    {
+      "epoch": 1.2708768267223383,
+      "grad_norm": 0.9178900718688965,
+      "learning_rate": 4.564522584708662e-06,
+      "loss": 0.7802,
+      "step": 811
+    },
+    {
+      "epoch": 1.2724425887265136,
+      "grad_norm": 1.0006638765335083,
+      "learning_rate": 4.563333747766896e-06,
+      "loss": 0.7882,
+      "step": 812
+    },
+    {
+      "epoch": 1.274008350730689,
+      "grad_norm": 1.0139516592025757,
+      "learning_rate": 4.56214344555602e-06,
+      "loss": 0.6562,
+      "step": 813
+    },
+    {
+      "epoch": 1.2755741127348643,
+      "grad_norm": 1.0360203981399536,
+      "learning_rate": 4.560951678921324e-06,
+      "loss": 0.7951,
+      "step": 814
+    },
+    {
+      "epoch": 1.2771398747390397,
+      "grad_norm": 1.042457103729248,
+      "learning_rate": 4.559758448709138e-06,
+      "loss": 0.7572,
+      "step": 815
+    },
+    {
+      "epoch": 1.278705636743215,
+      "grad_norm": 1.0056653022766113,
+      "learning_rate": 4.558563755766827e-06,
+      "loss": 0.7679,
+      "step": 816
+    },
+    {
+      "epoch": 1.2802713987473904,
+      "grad_norm": 0.9838577508926392,
+      "learning_rate": 4.5573676009428e-06,
+      "loss": 0.7245,
+      "step": 817
+    },
+    {
+      "epoch": 1.2818371607515657,
+      "grad_norm": 1.0474631786346436,
+      "learning_rate": 4.556169985086503e-06,
+      "loss": 0.7823,
+      "step": 818
+    },
+    {
+      "epoch": 1.283402922755741,
+      "grad_norm": 1.0177569389343262,
+      "learning_rate": 4.554970909048417e-06,
+      "loss": 0.7599,
+      "step": 819
+    },
+    {
+      "epoch": 1.2849686847599164,
+      "grad_norm": 1.0120493173599243,
+      "learning_rate": 4.553770373680062e-06,
+      "loss": 0.7842,
+      "step": 820
+    },
+    {
+      "epoch": 1.2865344467640918,
+      "grad_norm": 1.1045918464660645,
+      "learning_rate": 4.5525683798339946e-06,
+      "loss": 0.7075,
+      "step": 821
+    },
+    {
+      "epoch": 1.2881002087682671,
+      "grad_norm": 1.097714900970459,
+      "learning_rate": 4.551364928363805e-06,
+      "loss": 0.7571,
+      "step": 822
+    },
+    {
+      "epoch": 1.2896659707724427,
+      "grad_norm": 0.9125490188598633,
+      "learning_rate": 4.550160020124121e-06,
+      "loss": 0.7626,
+      "step": 823
+    },
+    {
+      "epoch": 1.2912317327766178,
+      "grad_norm": 1.1127262115478516,
+      "learning_rate": 4.548953655970604e-06,
+      "loss": 0.6618,
+      "step": 824
+    },
+    {
+      "epoch": 1.2927974947807934,
+      "grad_norm": 1.0372540950775146,
+      "learning_rate": 4.547745836759949e-06,
+      "loss": 0.7467,
+      "step": 825
+    },
+    {
+      "epoch": 1.2943632567849686,
+      "grad_norm": 1.0158213376998901,
+      "learning_rate": 4.546536563349885e-06,
+      "loss": 0.6421,
+      "step": 826
+    },
+    {
+      "epoch": 1.2959290187891441,
+      "grad_norm": 1.0057018995285034,
+      "learning_rate": 4.545325836599171e-06,
+      "loss": 0.7252,
+      "step": 827
+    },
+    {
+      "epoch": 1.2974947807933195,
+      "grad_norm": 1.0552566051483154,
+      "learning_rate": 4.544113657367604e-06,
+      "loss": 0.6645,
+      "step": 828
+    },
+    {
+      "epoch": 1.2990605427974948,
+      "grad_norm": 1.0798696279525757,
+      "learning_rate": 4.542900026516006e-06,
+      "loss": 0.7627,
+      "step": 829
+    },
+    {
+      "epoch": 1.3006263048016702,
+      "grad_norm": 1.0450047254562378,
+      "learning_rate": 4.541684944906233e-06,
+      "loss": 0.705,
+      "step": 830
+    },
+    {
+      "epoch": 1.3021920668058455,
+      "grad_norm": 0.9975817203521729,
+      "learning_rate": 4.540468413401171e-06,
+      "loss": 0.7639,
+      "step": 831
+    },
+    {
+      "epoch": 1.303757828810021,
+      "grad_norm": 0.9081487059593201,
+      "learning_rate": 4.539250432864736e-06,
+      "loss": 0.8129,
+      "step": 832
+    },
+    {
+      "epoch": 1.3053235908141962,
+      "grad_norm": 1.228684902191162,
+      "learning_rate": 4.538031004161872e-06,
+      "loss": 0.6984,
+      "step": 833
+    },
+    {
+      "epoch": 1.3068893528183716,
+      "grad_norm": 1.0577794313430786,
+      "learning_rate": 4.536810128158552e-06,
+      "loss": 0.7436,
+      "step": 834
+    },
+    {
+      "epoch": 1.308455114822547,
+      "grad_norm": 1.1270350217819214,
+      "learning_rate": 4.535587805721777e-06,
+      "loss": 0.7589,
+      "step": 835
+    },
+    {
+      "epoch": 1.3100208768267223,
+      "grad_norm": 1.1473027467727661,
+      "learning_rate": 4.5343640377195766e-06,
+      "loss": 0.7988,
+      "step": 836
+    },
+    {
+      "epoch": 1.3115866388308977,
+      "grad_norm": 0.9700750112533569,
+      "learning_rate": 4.533138825021001e-06,
+      "loss": 0.7388,
+      "step": 837
+    },
+    {
+      "epoch": 1.313152400835073,
+      "grad_norm": 0.9596260786056519,
+      "learning_rate": 4.531912168496135e-06,
+      "loss": 0.8177,
+      "step": 838
+    },
+    {
+      "epoch": 1.3147181628392484,
+      "grad_norm": 0.9888957738876343,
+      "learning_rate": 4.530684069016081e-06,
+      "loss": 0.809,
+      "step": 839
+    },
+    {
+      "epoch": 1.316283924843424,
+      "grad_norm": 1.219224214553833,
+      "learning_rate": 4.529454527452972e-06,
+      "loss": 0.7761,
+      "step": 840
+    },
+    {
+      "epoch": 1.317849686847599,
+      "grad_norm": 0.9730296730995178,
+      "learning_rate": 4.5282235446799624e-06,
+      "loss": 0.8044,
+      "step": 841
+    },
+    {
+      "epoch": 1.3194154488517746,
+      "grad_norm": 1.1046907901763916,
+      "learning_rate": 4.526991121571228e-06,
+      "loss": 0.7452,
+      "step": 842
+    },
+    {
+      "epoch": 1.3209812108559498,
+      "grad_norm": 0.9707220196723938,
+      "learning_rate": 4.525757259001972e-06,
+      "loss": 0.7112,
+      "step": 843
+    },
+    {
+      "epoch": 1.3225469728601253,
+      "grad_norm": 1.0085593461990356,
+      "learning_rate": 4.524521957848416e-06,
+      "loss": 0.6266,
+      "step": 844
+    },
+    {
+      "epoch": 1.3241127348643007,
+      "grad_norm": 1.131439447402954,
+      "learning_rate": 4.5232852189878055e-06,
+      "loss": 0.7393,
+      "step": 845
+    },
+    {
+      "epoch": 1.325678496868476,
+      "grad_norm": 1.1238782405853271,
+      "learning_rate": 4.522047043298406e-06,
+      "loss": 0.7541,
+      "step": 846
+    },
+    {
+      "epoch": 1.3272442588726514,
+      "grad_norm": 0.9149152636528015,
+      "learning_rate": 4.520807431659504e-06,
+      "loss": 0.8162,
+      "step": 847
+    },
+    {
+      "epoch": 1.3288100208768268,
+      "grad_norm": 1.0428732633590698,
+      "learning_rate": 4.519566384951403e-06,
+      "loss": 0.7511,
+      "step": 848
+    },
+    {
+      "epoch": 1.3303757828810021,
+      "grad_norm": 1.0263090133666992,
+      "learning_rate": 4.518323904055432e-06,
+      "loss": 0.7482,
+      "step": 849
+    },
+    {
+      "epoch": 1.3319415448851775,
+      "grad_norm": 1.113810420036316,
+      "learning_rate": 4.517079989853931e-06,
+      "loss": 0.797,
+      "step": 850
+    },
+    {
+      "epoch": 1.3335073068893528,
+      "grad_norm": 1.0090736150741577,
+      "learning_rate": 4.515834643230263e-06,
+      "loss": 0.688,
+      "step": 851
+    },
+    {
+      "epoch": 1.3350730688935282,
+      "grad_norm": 1.012824296951294,
+      "learning_rate": 4.514587865068806e-06,
+      "loss": 0.6683,
+      "step": 852
+    },
+    {
+      "epoch": 1.3366388308977035,
+      "grad_norm": 0.9442676305770874,
+      "learning_rate": 4.513339656254954e-06,
+      "loss": 0.7812,
+      "step": 853
+    },
+    {
+      "epoch": 1.3382045929018789,
+      "grad_norm": 1.0870879888534546,
+      "learning_rate": 4.512090017675119e-06,
+      "loss": 0.6943,
+      "step": 854
+    },
+    {
+      "epoch": 1.3397703549060542,
+      "grad_norm": 0.9352210164070129,
+      "learning_rate": 4.510838950216727e-06,
+      "loss": 0.7571,
+      "step": 855
+    },
+    {
+      "epoch": 1.3413361169102296,
+      "grad_norm": 1.173621654510498,
+      "learning_rate": 4.509586454768221e-06,
+      "loss": 0.6985,
+      "step": 856
+    },
+    {
+      "epoch": 1.342901878914405,
+      "grad_norm": 0.9790180921554565,
+      "learning_rate": 4.508332532219053e-06,
+      "loss": 0.7132,
+      "step": 857
+    },
+    {
+      "epoch": 1.3444676409185803,
+      "grad_norm": 1.0999163389205933,
+      "learning_rate": 4.507077183459694e-06,
+      "loss": 0.7858,
+      "step": 858
+    },
+    {
+      "epoch": 1.3460334029227559,
+      "grad_norm": 1.018182635307312,
+      "learning_rate": 4.505820409381625e-06,
+      "loss": 0.712,
+      "step": 859
+    },
+    {
+      "epoch": 1.347599164926931,
+      "grad_norm": 1.1015701293945312,
+      "learning_rate": 4.504562210877338e-06,
+      "loss": 0.8004,
+      "step": 860
+    },
+    {
+      "epoch": 1.3491649269311066,
+      "grad_norm": 1.0872273445129395,
+      "learning_rate": 4.50330258884034e-06,
+      "loss": 0.7478,
+      "step": 861
+    },
+    {
+      "epoch": 1.350730688935282,
+      "grad_norm": 1.1098262071609497,
+      "learning_rate": 4.502041544165148e-06,
+      "loss": 0.7894,
+      "step": 862
+    },
+    {
+      "epoch": 1.3522964509394573,
+      "grad_norm": 1.0535573959350586,
+      "learning_rate": 4.500779077747285e-06,
+      "loss": 0.7185,
+      "step": 863
+    },
+    {
+      "epoch": 1.3538622129436326,
+      "grad_norm": 1.0424474477767944,
+      "learning_rate": 4.499515190483291e-06,
+      "loss": 0.748,
+      "step": 864
+    },
+    {
+      "epoch": 1.355427974947808,
+      "grad_norm": 1.005065679550171,
+      "learning_rate": 4.498249883270708e-06,
+      "loss": 0.7948,
+      "step": 865
+    },
+    {
+      "epoch": 1.3569937369519833,
+      "grad_norm": 1.0052721500396729,
+      "learning_rate": 4.49698315700809e-06,
+      "loss": 0.7619,
+      "step": 866
+    },
+    {
+      "epoch": 1.3585594989561587,
+      "grad_norm": 1.0953692197799683,
+      "learning_rate": 4.495715012594999e-06,
+      "loss": 0.7363,
+      "step": 867
+    },
+    {
+      "epoch": 1.360125260960334,
+      "grad_norm": 1.1238093376159668,
+      "learning_rate": 4.494445450932003e-06,
+      "loss": 0.6576,
+      "step": 868
+    },
+    {
+      "epoch": 1.3616910229645094,
+      "grad_norm": 0.8546008467674255,
+      "learning_rate": 4.493174472920676e-06,
+      "loss": 0.7513,
+      "step": 869
+    },
+    {
+      "epoch": 1.3632567849686847,
+      "grad_norm": 1.0863126516342163,
+      "learning_rate": 4.4919020794635995e-06,
+      "loss": 0.6886,
+      "step": 870
+    },
+    {
+      "epoch": 1.36482254697286,
+      "grad_norm": 0.9943820834159851,
+      "learning_rate": 4.490628271464359e-06,
+      "loss": 0.7253,
+      "step": 871
+    },
+    {
+      "epoch": 1.3663883089770354,
+      "grad_norm": 0.9377979636192322,
+      "learning_rate": 4.4893530498275436e-06,
+      "loss": 0.7026,
+      "step": 872
+    },
+    {
+      "epoch": 1.3679540709812108,
+      "grad_norm": 1.100799798965454,
+      "learning_rate": 4.488076415458749e-06,
+      "loss": 0.719,
+      "step": 873
+    },
+    {
+      "epoch": 1.3695198329853862,
+      "grad_norm": 1.015074610710144,
+      "learning_rate": 4.486798369264572e-06,
+      "loss": 0.7513,
+      "step": 874
+    },
+    {
+      "epoch": 1.3710855949895615,
+      "grad_norm": 1.047530174255371,
+      "learning_rate": 4.485518912152611e-06,
+      "loss": 0.7757,
+      "step": 875
+    },
+    {
+      "epoch": 1.372651356993737,
+      "grad_norm": 1.0264610052108765,
+      "learning_rate": 4.484238045031471e-06,
+      "loss": 0.7712,
+      "step": 876
+    },
+    {
+      "epoch": 1.3742171189979122,
+      "grad_norm": 0.9848430156707764,
+      "learning_rate": 4.482955768810752e-06,
+      "loss": 0.7569,
+      "step": 877
+    },
+    {
+      "epoch": 1.3757828810020878,
+      "grad_norm": 1.070816159248352,
+      "learning_rate": 4.48167208440106e-06,
+      "loss": 0.7857,
+      "step": 878
+    },
+    {
+      "epoch": 1.377348643006263,
+      "grad_norm": 1.0134745836257935,
+      "learning_rate": 4.480386992713998e-06,
+      "loss": 0.7372,
+      "step": 879
+    },
+    {
+      "epoch": 1.3789144050104385,
+      "grad_norm": 1.0736140012741089,
+      "learning_rate": 4.47910049466217e-06,
+      "loss": 0.7794,
+      "step": 880
+    },
+    {
+      "epoch": 1.3804801670146138,
+      "grad_norm": 1.0163054466247559,
+      "learning_rate": 4.477812591159176e-06,
+      "loss": 0.7341,
+      "step": 881
+    },
+    {
+      "epoch": 1.3820459290187892,
+      "grad_norm": 1.074578881263733,
+      "learning_rate": 4.476523283119619e-06,
+      "loss": 0.6924,
+      "step": 882
+    },
+    {
+      "epoch": 1.3836116910229646,
+      "grad_norm": 1.183667540550232,
+      "learning_rate": 4.475232571459095e-06,
+      "loss": 0.7866,
+      "step": 883
+    },
+    {
+      "epoch": 1.38517745302714,
+      "grad_norm": 1.030827522277832,
+      "learning_rate": 4.473940457094199e-06,
+      "loss": 0.7845,
+      "step": 884
+    },
+    {
+      "epoch": 1.3867432150313153,
+      "grad_norm": 1.0043658018112183,
+      "learning_rate": 4.47264694094252e-06,
+      "loss": 0.7399,
+      "step": 885
+    },
+    {
+      "epoch": 1.3883089770354906,
+      "grad_norm": 1.1135472059249878,
+      "learning_rate": 4.471352023922645e-06,
+      "loss": 0.771,
+      "step": 886
+    },
+    {
+      "epoch": 1.389874739039666,
+      "grad_norm": 1.0338670015335083,
+      "learning_rate": 4.470055706954154e-06,
+      "loss": 0.7282,
+      "step": 887
+    },
+    {
+      "epoch": 1.3914405010438413,
+      "grad_norm": 0.918855607509613,
+      "learning_rate": 4.468757990957623e-06,
+      "loss": 0.728,
+      "step": 888
+    },
+    {
+      "epoch": 1.3930062630480167,
+      "grad_norm": 1.06833016872406,
+      "learning_rate": 4.467458876854619e-06,
+      "loss": 0.7012,
+      "step": 889
+    },
+    {
+      "epoch": 1.394572025052192,
+      "grad_norm": 1.0103005170822144,
+      "learning_rate": 4.4661583655677045e-06,
+      "loss": 0.7074,
+      "step": 890
+    },
+    {
+      "epoch": 1.3961377870563674,
+      "grad_norm": 1.0418543815612793,
+      "learning_rate": 4.464856458020433e-06,
+      "loss": 0.6858,
+      "step": 891
+    },
+    {
+      "epoch": 1.3977035490605427,
+      "grad_norm": 0.9720668196678162,
+      "learning_rate": 4.463553155137348e-06,
+      "loss": 0.7024,
+      "step": 892
+    },
+    {
+      "epoch": 1.399269311064718,
+      "grad_norm": 1.1870561838150024,
+      "learning_rate": 4.462248457843986e-06,
+      "loss": 0.8069,
+      "step": 893
+    },
+    {
+      "epoch": 1.4008350730688934,
+      "grad_norm": 0.9888094663619995,
+      "learning_rate": 4.460942367066874e-06,
+      "loss": 0.7281,
+      "step": 894
+    },
+    {
+      "epoch": 1.402400835073069,
+      "grad_norm": 0.9120670557022095,
+      "learning_rate": 4.459634883733527e-06,
+      "loss": 0.7992,
+      "step": 895
+    },
+    {
+      "epoch": 1.4039665970772441,
+      "grad_norm": 1.0239481925964355,
+      "learning_rate": 4.45832600877245e-06,
+      "loss": 0.7312,
+      "step": 896
+    },
+    {
+      "epoch": 1.4055323590814197,
+      "grad_norm": 0.9379130005836487,
+      "learning_rate": 4.4570157431131355e-06,
+      "loss": 0.7019,
+      "step": 897
+    },
+    {
+      "epoch": 1.407098121085595,
+      "grad_norm": 1.0789109468460083,
+      "learning_rate": 4.455704087686063e-06,
+      "loss": 0.8149,
+      "step": 898
+    },
+    {
+      "epoch": 1.4086638830897704,
+      "grad_norm": 1.0521917343139648,
+      "learning_rate": 4.454391043422703e-06,
+      "loss": 0.7873,
+      "step": 899
+    },
+    {
+      "epoch": 1.4102296450939458,
+      "grad_norm": 0.9582643508911133,
+      "learning_rate": 4.453076611255507e-06,
+      "loss": 0.7815,
+      "step": 900
+    },
+    {
+      "epoch": 1.4117954070981211,
+      "grad_norm": 1.1096739768981934,
+      "learning_rate": 4.451760792117914e-06,
+      "loss": 0.7636,
+      "step": 901
+    },
+    {
+      "epoch": 1.4133611691022965,
+      "grad_norm": 1.043574333190918,
+      "learning_rate": 4.450443586944349e-06,
+      "loss": 0.7411,
+      "step": 902
+    },
+    {
+      "epoch": 1.4149269311064718,
+      "grad_norm": 1.0280033349990845,
+      "learning_rate": 4.449124996670221e-06,
+      "loss": 0.6525,
+      "step": 903
+    },
+    {
+      "epoch": 1.4164926931106472,
+      "grad_norm": 1.0329992771148682,
+      "learning_rate": 4.44780502223192e-06,
+      "loss": 0.7133,
+      "step": 904
+    },
+    {
+      "epoch": 1.4180584551148225,
+      "grad_norm": 1.016168475151062,
+      "learning_rate": 4.446483664566825e-06,
+      "loss": 0.7098,
+      "step": 905
+    },
+    {
+      "epoch": 1.4196242171189979,
+      "grad_norm": 1.2022786140441895,
+      "learning_rate": 4.4451609246132895e-06,
+      "loss": 0.7303,
+      "step": 906
+    },
+    {
+      "epoch": 1.4211899791231732,
+      "grad_norm": 1.0311683416366577,
+      "learning_rate": 4.443836803310655e-06,
+      "loss": 0.7475,
+      "step": 907
+    },
+    {
+      "epoch": 1.4227557411273486,
+      "grad_norm": 1.0073823928833008,
+      "learning_rate": 4.442511301599241e-06,
+      "loss": 0.7513,
+      "step": 908
+    },
+    {
+      "epoch": 1.424321503131524,
+      "grad_norm": 1.0702322721481323,
+      "learning_rate": 4.441184420420347e-06,
+      "loss": 0.6918,
+      "step": 909
+    },
+    {
+      "epoch": 1.4258872651356993,
+      "grad_norm": 1.1416171789169312,
+      "learning_rate": 4.439856160716254e-06,
+      "loss": 0.6324,
+      "step": 910
+    },
+    {
+      "epoch": 1.4274530271398747,
+      "grad_norm": 1.0560253858566284,
+      "learning_rate": 4.438526523430221e-06,
+      "loss": 0.749,
+      "step": 911
+    },
+    {
+      "epoch": 1.4290187891440502,
+      "grad_norm": 1.021008849143982,
+      "learning_rate": 4.4371955095064835e-06,
+      "loss": 0.7666,
+      "step": 912
+    },
+    {
+      "epoch": 1.4305845511482254,
+      "grad_norm": 1.151904821395874,
+      "learning_rate": 4.435863119890258e-06,
+      "loss": 0.6904,
+      "step": 913
+    },
+    {
+      "epoch": 1.432150313152401,
+      "grad_norm": 1.1216702461242676,
+      "learning_rate": 4.434529355527736e-06,
+      "loss": 0.7068,
+      "step": 914
+    },
+    {
+      "epoch": 1.433716075156576,
+      "grad_norm": 1.1130588054656982,
+      "learning_rate": 4.433194217366085e-06,
+      "loss": 0.6524,
+      "step": 915
+    },
+    {
+      "epoch": 1.4352818371607516,
+      "grad_norm": 1.07612144947052,
+      "learning_rate": 4.431857706353449e-06,
+      "loss": 0.8078,
+      "step": 916
+    },
+    {
+      "epoch": 1.436847599164927,
+      "grad_norm": 1.0220447778701782,
+      "learning_rate": 4.430519823438946e-06,
+      "loss": 0.6736,
+      "step": 917
+    },
+    {
+      "epoch": 1.4384133611691023,
+      "grad_norm": 1.0657141208648682,
+      "learning_rate": 4.429180569572669e-06,
+      "loss": 0.7457,
+      "step": 918
+    },
+    {
+      "epoch": 1.4399791231732777,
+      "grad_norm": 1.0646854639053345,
+      "learning_rate": 4.427839945705684e-06,
+      "loss": 0.7925,
+      "step": 919
+    },
+    {
+      "epoch": 1.441544885177453,
+      "grad_norm": 1.0619502067565918,
+      "learning_rate": 4.426497952790031e-06,
+      "loss": 0.6968,
+      "step": 920
+    },
+    {
+      "epoch": 1.4431106471816284,
+      "grad_norm": 0.9684338569641113,
+      "learning_rate": 4.425154591778722e-06,
+      "loss": 0.7411,
+      "step": 921
+    },
+    {
+      "epoch": 1.4446764091858038,
+      "grad_norm": 1.1225167512893677,
+      "learning_rate": 4.423809863625739e-06,
+      "loss": 0.7623,
+      "step": 922
+    },
+    {
+      "epoch": 1.446242171189979,
+      "grad_norm": 0.9310746788978577,
+      "learning_rate": 4.422463769286036e-06,
+      "loss": 0.8219,
+      "step": 923
+    },
+    {
+      "epoch": 1.4478079331941545,
+      "grad_norm": 1.1107957363128662,
+      "learning_rate": 4.4211163097155375e-06,
+      "loss": 0.7755,
+      "step": 924
+    },
+    {
+      "epoch": 1.4493736951983298,
+      "grad_norm": 0.9827596545219421,
+      "learning_rate": 4.419767485871136e-06,
+      "loss": 0.7102,
+      "step": 925
+    },
+    {
+      "epoch": 1.4509394572025052,
+      "grad_norm": 1.1075087785720825,
+      "learning_rate": 4.418417298710695e-06,
+      "loss": 0.6399,
+      "step": 926
+    },
+    {
+      "epoch": 1.4525052192066805,
+      "grad_norm": 0.9575802087783813,
+      "learning_rate": 4.417065749193045e-06,
+      "loss": 0.721,
+      "step": 927
+    },
+    {
+      "epoch": 1.4540709812108559,
+      "grad_norm": 1.0016006231307983,
+      "learning_rate": 4.415712838277985e-06,
+      "loss": 0.762,
+      "step": 928
+    },
+    {
+      "epoch": 1.4556367432150314,
+      "grad_norm": 1.1907106637954712,
+      "learning_rate": 4.414358566926279e-06,
+      "loss": 0.7904,
+      "step": 929
+    },
+    {
+      "epoch": 1.4572025052192066,
+      "grad_norm": 1.0011543035507202,
+      "learning_rate": 4.413002936099656e-06,
+      "loss": 0.7179,
+      "step": 930
+    },
+    {
+      "epoch": 1.4587682672233822,
+      "grad_norm": 0.980684757232666,
+      "learning_rate": 4.411645946760817e-06,
+      "loss": 0.6845,
+      "step": 931
+    },
+    {
+      "epoch": 1.4603340292275573,
+      "grad_norm": 0.9458135366439819,
+      "learning_rate": 4.4102875998734176e-06,
+      "loss": 0.6911,
+      "step": 932
+    },
+    {
+      "epoch": 1.4618997912317329,
+      "grad_norm": 1.020966649055481,
+      "learning_rate": 4.408927896402087e-06,
+      "loss": 0.7284,
+      "step": 933
+    },
+    {
+      "epoch": 1.4634655532359082,
+      "grad_norm": 1.041833519935608,
+      "learning_rate": 4.407566837312413e-06,
+      "loss": 0.7451,
+      "step": 934
+    },
+    {
+      "epoch": 1.4650313152400836,
+      "grad_norm": 0.9076465964317322,
+      "learning_rate": 4.406204423570946e-06,
+      "loss": 0.7553,
+      "step": 935
+    },
+    {
+      "epoch": 1.466597077244259,
+      "grad_norm": 1.1963952779769897,
+      "learning_rate": 4.404840656145199e-06,
+      "loss": 0.7437,
+      "step": 936
+    },
+    {
+      "epoch": 1.4681628392484343,
+      "grad_norm": 1.0181536674499512,
+      "learning_rate": 4.403475536003648e-06,
+      "loss": 0.7283,
+      "step": 937
+    },
+    {
+      "epoch": 1.4697286012526096,
+      "grad_norm": 0.9563003778457642,
+      "learning_rate": 4.402109064115727e-06,
+      "loss": 0.7375,
+      "step": 938
+    },
+    {
+      "epoch": 1.471294363256785,
+      "grad_norm": 1.05129873752594,
+      "learning_rate": 4.400741241451832e-06,
+      "loss": 0.6805,
+      "step": 939
+    },
+    {
+      "epoch": 1.4728601252609603,
+      "grad_norm": 1.0912166833877563,
+      "learning_rate": 4.399372068983317e-06,
+      "loss": 0.7076,
+      "step": 940
+    },
+    {
+      "epoch": 1.4744258872651357,
+      "grad_norm": 1.108500599861145,
+      "learning_rate": 4.398001547682494e-06,
+      "loss": 0.6836,
+      "step": 941
+    },
+    {
+      "epoch": 1.475991649269311,
+      "grad_norm": 1.0741651058197021,
+      "learning_rate": 4.396629678522636e-06,
+      "loss": 0.7405,
+      "step": 942
+    },
+    {
+      "epoch": 1.4775574112734864,
+      "grad_norm": 1.0603930950164795,
+      "learning_rate": 4.39525646247797e-06,
+      "loss": 0.7602,
+      "step": 943
+    },
+    {
+      "epoch": 1.4791231732776617,
+      "grad_norm": 0.9814990162849426,
+      "learning_rate": 4.39388190052368e-06,
+      "loss": 0.7324,
+      "step": 944
+    },
+    {
+      "epoch": 1.480688935281837,
+      "grad_norm": 0.9643695950508118,
+      "learning_rate": 4.392505993635906e-06,
+      "loss": 0.7034,
+      "step": 945
+    },
+    {
+      "epoch": 1.4822546972860124,
+      "grad_norm": 1.0490994453430176,
+      "learning_rate": 4.391128742791743e-06,
+      "loss": 0.7358,
+      "step": 946
+    },
+    {
+      "epoch": 1.4838204592901878,
+      "grad_norm": 0.9223467111587524,
+      "learning_rate": 4.389750148969244e-06,
+      "loss": 0.698,
+      "step": 947
+    },
+    {
+      "epoch": 1.4853862212943634,
+      "grad_norm": 1.1107758283615112,
+      "learning_rate": 4.388370213147409e-06,
+      "loss": 0.7264,
+      "step": 948
+    },
+    {
+      "epoch": 1.4869519832985385,
+      "grad_norm": 1.05939781665802,
+      "learning_rate": 4.386988936306196e-06,
+      "loss": 0.7209,
+      "step": 949
+    },
+    {
+      "epoch": 1.488517745302714,
+      "grad_norm": 1.024686574935913,
+      "learning_rate": 4.385606319426512e-06,
+      "loss": 0.7798,
+      "step": 950
+    },
+    {
+      "epoch": 1.4900835073068894,
+      "grad_norm": 0.9920977354049683,
+      "learning_rate": 4.3842223634902205e-06,
+      "loss": 0.8045,
+      "step": 951
+    },
+    {
+      "epoch": 1.4916492693110648,
+      "grad_norm": 0.950927197933197,
+      "learning_rate": 4.382837069480131e-06,
+      "loss": 0.6777,
+      "step": 952
+    },
+    {
+      "epoch": 1.4932150313152401,
+      "grad_norm": 1.0072081089019775,
+      "learning_rate": 4.3814504383800056e-06,
+      "loss": 0.7458,
+      "step": 953
+    },
+    {
+      "epoch": 1.4947807933194155,
+      "grad_norm": 0.9982030987739563,
+      "learning_rate": 4.3800624711745546e-06,
+      "loss": 0.8215,
+      "step": 954
+    },
+    {
+      "epoch": 1.4963465553235908,
+      "grad_norm": 0.9682126641273499,
+      "learning_rate": 4.3786731688494396e-06,
+      "loss": 0.7157,
+      "step": 955
+    },
+    {
+      "epoch": 1.4979123173277662,
+      "grad_norm": 1.1599830389022827,
+      "learning_rate": 4.377282532391267e-06,
+      "loss": 0.8091,
+      "step": 956
+    },
+    {
+      "epoch": 1.4994780793319415,
+      "grad_norm": 1.0624948740005493,
+      "learning_rate": 4.3758905627875934e-06,
+      "loss": 0.7428,
+      "step": 957
+    },
+    {
+      "epoch": 1.501043841336117,
+      "grad_norm": 1.14694082736969,
+      "learning_rate": 4.374497261026921e-06,
+      "loss": 0.7752,
+      "step": 958
+    },
+    {
+      "epoch": 1.5026096033402923,
+      "grad_norm": 1.0355825424194336,
+      "learning_rate": 4.373102628098697e-06,
+      "loss": 0.7348,
+      "step": 959
+    },
+    {
+      "epoch": 1.5041753653444676,
+      "grad_norm": 1.036026954650879,
+      "learning_rate": 4.371706664993316e-06,
+      "loss": 0.7536,
+      "step": 960
+    },
+    {
+      "epoch": 1.505741127348643,
+      "grad_norm": 1.2319790124893188,
+      "learning_rate": 4.370309372702116e-06,
+      "loss": 0.6985,
+      "step": 961
+    },
+    {
+      "epoch": 1.5073068893528183,
+      "grad_norm": 0.908687949180603,
+      "learning_rate": 4.36891075221738e-06,
+      "loss": 0.7173,
+      "step": 962
+    },
+    {
+      "epoch": 1.5088726513569939,
+      "grad_norm": 1.0482099056243896,
+      "learning_rate": 4.367510804532331e-06,
+      "loss": 0.7514,
+      "step": 963
+    },
+    {
+      "epoch": 1.510438413361169,
+      "grad_norm": 1.0434144735336304,
+      "learning_rate": 4.36610953064114e-06,
+      "loss": 0.754,
+      "step": 964
+    },
+    {
+      "epoch": 1.5120041753653446,
+      "grad_norm": 1.0122450590133667,
+      "learning_rate": 4.364706931538914e-06,
+      "loss": 0.7767,
+      "step": 965
+    },
+    {
+      "epoch": 1.5135699373695197,
+      "grad_norm": 1.0471900701522827,
+      "learning_rate": 4.363303008221705e-06,
+      "loss": 0.7337,
+      "step": 966
+    },
+    {
+      "epoch": 1.5151356993736953,
+      "grad_norm": 1.0691189765930176,
+      "learning_rate": 4.3618977616865045e-06,
+      "loss": 0.8047,
+      "step": 967
+    },
+    {
+      "epoch": 1.5167014613778704,
+      "grad_norm": 1.015114188194275,
+      "learning_rate": 4.360491192931242e-06,
+      "loss": 0.726,
+      "step": 968
+    },
+    {
+      "epoch": 1.518267223382046,
+      "grad_norm": 0.9770560264587402,
+      "learning_rate": 4.359083302954788e-06,
+      "loss": 0.7692,
+      "step": 969
+    },
+    {
+      "epoch": 1.5198329853862211,
+      "grad_norm": 1.079991340637207,
+      "learning_rate": 4.357674092756951e-06,
+      "loss": 0.8424,
+      "step": 970
+    },
+    {
+      "epoch": 1.5213987473903967,
+      "grad_norm": 1.068708896636963,
+      "learning_rate": 4.356263563338476e-06,
+      "loss": 0.758,
+      "step": 971
+    },
+    {
+      "epoch": 1.522964509394572,
+      "grad_norm": 1.1005477905273438,
+      "learning_rate": 4.354851715701046e-06,
+      "loss": 0.7209,
+      "step": 972
+    },
+    {
+      "epoch": 1.5245302713987474,
+      "grad_norm": 0.9870902299880981,
+      "learning_rate": 4.3534385508472786e-06,
+      "loss": 0.7272,
+      "step": 973
+    },
+    {
+      "epoch": 1.5260960334029228,
+      "grad_norm": 0.9616238474845886,
+      "learning_rate": 4.352024069780729e-06,
+      "loss": 0.8124,
+      "step": 974
+    },
+    {
+      "epoch": 1.5276617954070981,
+      "grad_norm": 1.0341622829437256,
+      "learning_rate": 4.3506082735058855e-06,
+      "loss": 0.7434,
+      "step": 975
+    },
+    {
+      "epoch": 1.5292275574112735,
+      "grad_norm": 1.0157302618026733,
+      "learning_rate": 4.349191163028171e-06,
+      "loss": 0.7534,
+      "step": 976
+    },
+    {
+      "epoch": 1.5307933194154488,
+      "grad_norm": 1.041379690170288,
+      "learning_rate": 4.347772739353939e-06,
+      "loss": 0.6942,
+      "step": 977
+    },
+    {
+      "epoch": 1.5323590814196242,
+      "grad_norm": 1.016205072402954,
+      "learning_rate": 4.346353003490481e-06,
+      "loss": 0.6707,
+      "step": 978
+    },
+    {
+      "epoch": 1.5339248434237995,
+      "grad_norm": 1.0034714937210083,
+      "learning_rate": 4.344931956446017e-06,
+      "loss": 0.6358,
+      "step": 979
+    },
+    {
+      "epoch": 1.535490605427975,
+      "grad_norm": 0.9671767354011536,
+      "learning_rate": 4.343509599229697e-06,
+      "loss": 0.6333,
+      "step": 980
+    },
+    {
+      "epoch": 1.5370563674321502,
+      "grad_norm": 1.1384506225585938,
+      "learning_rate": 4.342085932851605e-06,
+      "loss": 0.803,
+      "step": 981
+    },
+    {
+      "epoch": 1.5386221294363258,
+      "grad_norm": 1.023031234741211,
+      "learning_rate": 4.34066095832275e-06,
+      "loss": 0.7801,
+      "step": 982
+    },
+    {
+      "epoch": 1.540187891440501,
+      "grad_norm": 1.0675241947174072,
+      "learning_rate": 4.339234676655075e-06,
+      "loss": 0.7335,
+      "step": 983
+    },
+    {
+      "epoch": 1.5417536534446765,
+      "grad_norm": 1.0397543907165527,
+      "learning_rate": 4.337807088861447e-06,
+      "loss": 0.7722,
+      "step": 984
+    },
+    {
+      "epoch": 1.5433194154488517,
+      "grad_norm": 1.0892809629440308,
+      "learning_rate": 4.3363781959556635e-06,
+      "loss": 0.7603,
+      "step": 985
+    },
+    {
+      "epoch": 1.5448851774530272,
+      "grad_norm": 0.9335037469863892,
+      "learning_rate": 4.334947998952448e-06,
+      "loss": 0.7678,
+      "step": 986
+    },
+    {
+      "epoch": 1.5464509394572024,
+      "grad_norm": 1.0991342067718506,
+      "learning_rate": 4.333516498867449e-06,
+      "loss": 0.7461,
+      "step": 987
+    },
+    {
+      "epoch": 1.548016701461378,
+      "grad_norm": 1.112265944480896,
+      "learning_rate": 4.332083696717242e-06,
+      "loss": 0.7593,
+      "step": 988
+    },
+    {
+      "epoch": 1.5495824634655533,
+      "grad_norm": 0.9525822997093201,
+      "learning_rate": 4.330649593519325e-06,
+      "loss": 0.7629,
+      "step": 989
+    },
+    {
+      "epoch": 1.5511482254697286,
+      "grad_norm": 1.0624464750289917,
+      "learning_rate": 4.329214190292122e-06,
+      "loss": 0.7261,
+      "step": 990
+    },
+    {
+      "epoch": 1.552713987473904,
+      "grad_norm": 1.1969516277313232,
+      "learning_rate": 4.327777488054979e-06,
+      "loss": 0.7483,
+      "step": 991
+    },
+    {
+      "epoch": 1.5542797494780793,
+      "grad_norm": 1.0239863395690918,
+      "learning_rate": 4.326339487828167e-06,
+      "loss": 0.7265,
+      "step": 992
+    },
+    {
+      "epoch": 1.5558455114822547,
+      "grad_norm": 1.0881083011627197,
+      "learning_rate": 4.324900190632874e-06,
+      "loss": 0.724,
+      "step": 993
+    },
+    {
+      "epoch": 1.55741127348643,
+      "grad_norm": 1.098827838897705,
+      "learning_rate": 4.323459597491214e-06,
+      "loss": 0.7222,
+      "step": 994
+    },
+    {
+      "epoch": 1.5589770354906054,
+      "grad_norm": 1.105057716369629,
+      "learning_rate": 4.322017709426217e-06,
+      "loss": 0.7827,
+      "step": 995
+    },
+    {
+      "epoch": 1.5605427974947808,
+      "grad_norm": 0.9830679297447205,
+      "learning_rate": 4.3205745274618365e-06,
+      "loss": 0.7316,
+      "step": 996
+    },
+    {
+      "epoch": 1.562108559498956,
+      "grad_norm": 0.9951156973838806,
+      "learning_rate": 4.319130052622942e-06,
+      "loss": 0.7181,
+      "step": 997
+    },
+    {
+      "epoch": 1.5636743215031315,
+      "grad_norm": 1.0374284982681274,
+      "learning_rate": 4.317684285935323e-06,
+      "loss": 0.7355,
+      "step": 998
+    },
+    {
+      "epoch": 1.565240083507307,
+      "grad_norm": 1.0139906406402588,
+      "learning_rate": 4.316237228425683e-06,
+      "loss": 0.7249,
+      "step": 999
+    },
+    {
+      "epoch": 1.5668058455114822,
+      "grad_norm": 0.9555441737174988,
+      "learning_rate": 4.3147888811216496e-06,
+      "loss": 0.7387,
+      "step": 1000
+    },
+    {
+      "epoch": 1.5683716075156577,
+      "grad_norm": 1.0600265264511108,
+      "learning_rate": 4.313339245051758e-06,
+      "loss": 0.7445,
+      "step": 1001
+    },
+    {
+      "epoch": 1.5699373695198329,
+      "grad_norm": 1.0070478916168213,
+      "learning_rate": 4.311888321245461e-06,
+      "loss": 0.7605,
+      "step": 1002
+    },
+    {
+      "epoch": 1.5715031315240084,
+      "grad_norm": 1.019710898399353,
+      "learning_rate": 4.31043611073313e-06,
+      "loss": 0.7262,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5730688935281836,
+      "grad_norm": 1.0476088523864746,
+      "learning_rate": 4.308982614546045e-06,
+      "loss": 0.6766,
+      "step": 1004
+    },
+    {
+      "epoch": 1.5746346555323592,
+      "grad_norm": 1.0670830011367798,
+      "learning_rate": 4.307527833716403e-06,
+      "loss": 0.7465,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5762004175365343,
+      "grad_norm": 1.1807290315628052,
+      "learning_rate": 4.30607176927731e-06,
+      "loss": 0.7032,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5777661795407099,
+      "grad_norm": 1.0482438802719116,
+      "learning_rate": 4.3046144222627855e-06,
+      "loss": 0.7156,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5793319415448852,
+      "grad_norm": 0.9434391856193542,
+      "learning_rate": 4.303155793707759e-06,
+      "loss": 0.7816,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5808977035490606,
+      "grad_norm": 1.0497010946273804,
+      "learning_rate": 4.3016958846480706e-06,
+      "loss": 0.735,
+      "step": 1009
+    },
+    {
+      "epoch": 1.582463465553236,
+      "grad_norm": 0.955949068069458,
+      "learning_rate": 4.300234696120469e-06,
+      "loss": 0.6414,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5840292275574113,
+      "grad_norm": 1.0462027788162231,
+      "learning_rate": 4.298772229162613e-06,
+      "loss": 0.6949,
+      "step": 1011
+    },
+    {
+      "epoch": 1.5855949895615866,
+      "grad_norm": 1.0634411573410034,
+      "learning_rate": 4.297308484813067e-06,
+      "loss": 0.7736,
+      "step": 1012
+    },
+    {
+      "epoch": 1.587160751565762,
+      "grad_norm": 1.0463027954101562,
+      "learning_rate": 4.295843464111306e-06,
+      "loss": 0.6812,
+      "step": 1013
+    },
+    {
+      "epoch": 1.5887265135699373,
+      "grad_norm": 1.0409252643585205,
+      "learning_rate": 4.294377168097706e-06,
+      "loss": 0.7319,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5902922755741127,
+      "grad_norm": 0.9714667201042175,
+      "learning_rate": 4.292909597813555e-06,
+      "loss": 0.7473,
+      "step": 1015
+    },
+    {
+      "epoch": 1.5918580375782883,
+      "grad_norm": 1.03997802734375,
+      "learning_rate": 4.291440754301041e-06,
+      "loss": 0.7261,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5934237995824634,
+      "grad_norm": 1.063293218612671,
+      "learning_rate": 4.289970638603258e-06,
+      "loss": 0.7909,
+      "step": 1017
+    },
+    {
+      "epoch": 1.594989561586639,
+      "grad_norm": 1.110007643699646,
+      "learning_rate": 4.288499251764205e-06,
+      "loss": 0.7,
+      "step": 1018
+    },
+    {
+      "epoch": 1.596555323590814,
+      "grad_norm": 0.981261134147644,
+      "learning_rate": 4.2870265948287806e-06,
+      "loss": 0.7516,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5981210855949897,
+      "grad_norm": 1.0653973817825317,
+      "learning_rate": 4.2855526688427875e-06,
+      "loss": 0.7278,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5996868475991648,
+      "grad_norm": 0.9212517738342285,
+      "learning_rate": 4.284077474852928e-06,
+      "loss": 0.7251,
+      "step": 1021
+    },
+    {
+      "epoch": 1.6012526096033404,
+      "grad_norm": 1.0887659788131714,
+      "learning_rate": 4.282601013906808e-06,
+      "loss": 0.7271,
+      "step": 1022
+    },
+    {
+      "epoch": 1.6028183716075155,
+      "grad_norm": 1.0753779411315918,
+      "learning_rate": 4.281123287052928e-06,
+      "loss": 0.7679,
+      "step": 1023
+    },
+    {
+      "epoch": 1.604384133611691,
+      "grad_norm": 1.0282115936279297,
+      "learning_rate": 4.2796442953406935e-06,
+      "loss": 0.695,
+      "step": 1024
+    },
+    {
+      "epoch": 1.6059498956158664,
+      "grad_norm": 0.9782281517982483,
+      "learning_rate": 4.2781640398204036e-06,
+      "loss": 0.7717,
+      "step": 1025
+    },
+    {
+      "epoch": 1.6075156576200418,
+      "grad_norm": 1.06228768825531,
+      "learning_rate": 4.276682521543256e-06,
+      "loss": 0.8069,
+      "step": 1026
+    },
+    {
+      "epoch": 1.6090814196242171,
+      "grad_norm": 0.8853397965431213,
+      "learning_rate": 4.2751997415613465e-06,
+      "loss": 0.8174,
+      "step": 1027
+    },
+    {
+      "epoch": 1.6106471816283925,
+      "grad_norm": 1.1525455713272095,
+      "learning_rate": 4.273715700927666e-06,
+      "loss": 0.7149,
+      "step": 1028
+    },
+    {
+      "epoch": 1.6122129436325678,
+      "grad_norm": 0.8743720650672913,
+      "learning_rate": 4.272230400696101e-06,
+      "loss": 0.7296,
+      "step": 1029
+    },
+    {
+      "epoch": 1.6137787056367432,
+      "grad_norm": 1.1136301755905151,
+      "learning_rate": 4.270743841921431e-06,
+      "loss": 0.7539,
+      "step": 1030
+    },
+    {
+      "epoch": 1.6153444676409185,
+      "grad_norm": 0.9830031991004944,
+      "learning_rate": 4.269256025659332e-06,
+      "loss": 0.6545,
+      "step": 1031
+    },
+    {
+      "epoch": 1.616910229645094,
+      "grad_norm": 1.1221113204956055,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.6144,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6184759916492695,
+      "grad_norm": 0.975713849067688,
+      "learning_rate": 4.266276624900004e-06,
+      "loss": 0.746,
+      "step": 1033
+    },
+    {
+      "epoch": 1.6200417536534446,
+      "grad_norm": 1.1131149530410767,
+      "learning_rate": 4.264785042518587e-06,
+      "loss": 0.705,
+      "step": 1034
+    },
+    {
+      "epoch": 1.6216075156576202,
+      "grad_norm": 1.0574990510940552,
+      "learning_rate": 4.263292206881361e-06,
+      "loss": 0.7681,
+      "step": 1035
+    },
+    {
+      "epoch": 1.6231732776617953,
+      "grad_norm": 0.9931074976921082,
+      "learning_rate": 4.261798119048456e-06,
+      "loss": 0.7854,
+      "step": 1036
+    },
+    {
+      "epoch": 1.6247390396659709,
+      "grad_norm": 1.1598386764526367,
+      "learning_rate": 4.260302780080896e-06,
+      "loss": 0.7406,
+      "step": 1037
+    },
+    {
+      "epoch": 1.626304801670146,
+      "grad_norm": 0.9910677671432495,
+      "learning_rate": 4.258806191040587e-06,
+      "loss": 0.7107,
+      "step": 1038
+    },
+    {
+      "epoch": 1.6278705636743216,
+      "grad_norm": 1.1023072004318237,
+      "learning_rate": 4.2573083529903294e-06,
+      "loss": 0.7964,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6294363256784967,
+      "grad_norm": 0.991082489490509,
+      "learning_rate": 4.255809266993806e-06,
+      "loss": 0.762,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6310020876826723,
+      "grad_norm": 0.9281193017959595,
+      "learning_rate": 4.254308934115589e-06,
+      "loss": 0.7272,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6325678496868476,
+      "grad_norm": 1.20292067527771,
+      "learning_rate": 4.252807355421133e-06,
+      "loss": 0.7535,
+      "step": 1042
+    },
+    {
+      "epoch": 1.634133611691023,
+      "grad_norm": 1.0362178087234497,
+      "learning_rate": 4.251304531976778e-06,
+      "loss": 0.7524,
+      "step": 1043
+    },
+    {
+      "epoch": 1.6356993736951984,
+      "grad_norm": 1.0267268419265747,
+      "learning_rate": 4.249800464849751e-06,
+      "loss": 0.6588,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6372651356993737,
+      "grad_norm": 1.0531187057495117,
+      "learning_rate": 4.248295155108158e-06,
+      "loss": 0.7328,
+      "step": 1045
+    },
+    {
+      "epoch": 1.638830897703549,
+      "grad_norm": 1.0008069276809692,
+      "learning_rate": 4.246788603820989e-06,
+      "loss": 0.719,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6403966597077244,
+      "grad_norm": 1.064780831336975,
+      "learning_rate": 4.245280812058119e-06,
+      "loss": 0.7388,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6419624217118998,
+      "grad_norm": 1.06492018699646,
+      "learning_rate": 4.243771780890298e-06,
+      "loss": 0.7495,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6435281837160751,
+      "grad_norm": 0.9472903609275818,
+      "learning_rate": 4.24226151138916e-06,
+      "loss": 0.7409,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6450939457202505,
+      "grad_norm": 1.0612895488739014,
+      "learning_rate": 4.240750004627217e-06,
+      "loss": 0.7605,
+      "step": 1050
+    },
+    {
+      "epoch": 1.6466597077244258,
+      "grad_norm": 0.9634467959403992,
+      "learning_rate": 4.239237261677863e-06,
+      "loss": 0.7109,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6482254697286014,
+      "grad_norm": 1.1347044706344604,
+      "learning_rate": 4.2377232836153635e-06,
+      "loss": 0.693,
+      "step": 1052
+    },
+    {
+      "epoch": 1.6497912317327765,
+      "grad_norm": 1.0814287662506104,
+      "learning_rate": 4.236208071514865e-06,
+      "loss": 0.7159,
+      "step": 1053
+    },
+    {
+      "epoch": 1.651356993736952,
+      "grad_norm": 0.9427600502967834,
+      "learning_rate": 4.234691626452392e-06,
+      "loss": 0.7014,
+      "step": 1054
+    },
+    {
+      "epoch": 1.6529227557411272,
+      "grad_norm": 1.0177332162857056,
+      "learning_rate": 4.233173949504842e-06,
+      "loss": 0.7717,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6544885177453028,
+      "grad_norm": 1.0621635913848877,
+      "learning_rate": 4.231655041749985e-06,
+      "loss": 0.7078,
+      "step": 1056
+    },
+    {
+      "epoch": 1.656054279749478,
+      "grad_norm": 1.0474066734313965,
+      "learning_rate": 4.230134904266472e-06,
+      "loss": 0.7869,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6576200417536535,
+      "grad_norm": 0.9170332551002502,
+      "learning_rate": 4.228613538133821e-06,
+      "loss": 0.7806,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6591858037578286,
+      "grad_norm": 1.010465383529663,
+      "learning_rate": 4.227090944432426e-06,
+      "loss": 0.7352,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6607515657620042,
+      "grad_norm": 0.988068163394928,
+      "learning_rate": 4.22556712424355e-06,
+      "loss": 0.7173,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6623173277661796,
+      "grad_norm": 1.0753575563430786,
+      "learning_rate": 4.22404207864933e-06,
+      "loss": 0.6981,
+      "step": 1061
+    },
+    {
+      "epoch": 1.663883089770355,
+      "grad_norm": 1.0075691938400269,
+      "learning_rate": 4.222515808732769e-06,
+      "loss": 0.7279,
+      "step": 1062
+    },
+    {
+      "epoch": 1.6654488517745303,
+      "grad_norm": 1.0954983234405518,
+      "learning_rate": 4.220988315577745e-06,
+      "loss": 0.7675,
+      "step": 1063
+    },
+    {
+      "epoch": 1.6670146137787056,
+      "grad_norm": 1.0368746519088745,
+      "learning_rate": 4.2194596002690016e-06,
+      "loss": 0.7466,
+      "step": 1064
+    },
+    {
+      "epoch": 1.668580375782881,
+      "grad_norm": 1.083906650543213,
+      "learning_rate": 4.217929663892149e-06,
+      "loss": 0.7517,
+      "step": 1065
+    },
+    {
+      "epoch": 1.6701461377870563,
+      "grad_norm": 1.1356595754623413,
+      "learning_rate": 4.216398507533667e-06,
+      "loss": 0.7601,
+      "step": 1066
+    },
+    {
+      "epoch": 1.6717118997912317,
+      "grad_norm": 1.1610008478164673,
+      "learning_rate": 4.2148661322809e-06,
+      "loss": 0.7356,
+      "step": 1067
+    },
+    {
+      "epoch": 1.673277661795407,
+      "grad_norm": 1.019673466682434,
+      "learning_rate": 4.21333253922206e-06,
+      "loss": 0.7618,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6748434237995826,
+      "grad_norm": 1.1470829248428345,
+      "learning_rate": 4.2117977294462235e-06,
+      "loss": 0.8001,
+      "step": 1069
+    },
+    {
+      "epoch": 1.6764091858037578,
+      "grad_norm": 0.9750358462333679,
+      "learning_rate": 4.210261704043327e-06,
+      "loss": 0.6736,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6779749478079333,
+      "grad_norm": 1.061862826347351,
+      "learning_rate": 4.208724464104176e-06,
+      "loss": 0.7751,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6795407098121085,
+      "grad_norm": 1.0311503410339355,
+      "learning_rate": 4.2071860107204346e-06,
+      "loss": 0.7629,
+      "step": 1072
+    },
+    {
+      "epoch": 1.681106471816284,
+      "grad_norm": 1.0023622512817383,
+      "learning_rate": 4.2056463449846305e-06,
+      "loss": 0.6935,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6826722338204592,
+      "grad_norm": 1.1624165773391724,
+      "learning_rate": 4.204105467990153e-06,
+      "loss": 0.7451,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6842379958246347,
+      "grad_norm": 0.9852432608604431,
+      "learning_rate": 4.202563380831249e-06,
+      "loss": 0.7446,
+      "step": 1075
+    },
+    {
+      "epoch": 1.6858037578288099,
+      "grad_norm": 1.0009260177612305,
+      "learning_rate": 4.201020084603027e-06,
+      "loss": 0.726,
+      "step": 1076
+    },
+    {
+      "epoch": 1.6873695198329854,
+      "grad_norm": 1.0374701023101807,
+      "learning_rate": 4.199475580401452e-06,
+      "loss": 0.6687,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6889352818371608,
+      "grad_norm": 1.1437046527862549,
+      "learning_rate": 4.19792986932335e-06,
+      "loss": 0.7802,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6905010438413361,
+      "grad_norm": 0.9866487383842468,
+      "learning_rate": 4.1963829524664e-06,
+      "loss": 0.7117,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6920668058455115,
+      "grad_norm": 0.9720652103424072,
+      "learning_rate": 4.194834830929142e-06,
+      "loss": 0.778,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6936325678496869,
+      "grad_norm": 1.009422779083252,
+      "learning_rate": 4.1932855058109675e-06,
+      "loss": 0.6783,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6951983298538622,
+      "grad_norm": 1.024781346321106,
+      "learning_rate": 4.191734978212124e-06,
+      "loss": 0.6697,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6967640918580376,
+      "grad_norm": 1.060354471206665,
+      "learning_rate": 4.190183249233715e-06,
+      "loss": 0.7151,
+      "step": 1083
+    },
+    {
+      "epoch": 1.698329853862213,
+      "grad_norm": 1.118486762046814,
+      "learning_rate": 4.1886303199776924e-06,
+      "loss": 0.7408,
+      "step": 1084
+    },
+    {
+      "epoch": 1.6998956158663883,
+      "grad_norm": 0.9334209561347961,
+      "learning_rate": 4.187076191546865e-06,
+      "loss": 0.7646,
+      "step": 1085
+    },
+    {
+      "epoch": 1.7014613778705638,
+      "grad_norm": 1.0506161451339722,
+      "learning_rate": 4.185520865044892e-06,
+      "loss": 0.7299,
+      "step": 1086
+    },
+    {
+      "epoch": 1.703027139874739,
+      "grad_norm": 1.1261264085769653,
+      "learning_rate": 4.1839643415762825e-06,
+      "loss": 0.6813,
+      "step": 1087
+    },
+    {
+      "epoch": 1.7045929018789145,
+      "grad_norm": 0.9690954685211182,
+      "learning_rate": 4.182406622246396e-06,
+      "loss": 0.7417,
+      "step": 1088
+    },
+    {
+      "epoch": 1.7061586638830897,
+      "grad_norm": 1.0872795581817627,
+      "learning_rate": 4.180847708161442e-06,
+      "loss": 0.676,
+      "step": 1089
+    },
+    {
+      "epoch": 1.7077244258872653,
+      "grad_norm": 1.054091453552246,
+      "learning_rate": 4.179287600428476e-06,
+      "loss": 0.785,
+      "step": 1090
+    },
+    {
+      "epoch": 1.7092901878914404,
+      "grad_norm": 1.109984278678894,
+      "learning_rate": 4.177726300155404e-06,
+      "loss": 0.8033,
+      "step": 1091
+    },
+    {
+      "epoch": 1.710855949895616,
+      "grad_norm": 1.0984348058700562,
+      "learning_rate": 4.176163808450978e-06,
+      "loss": 0.6964,
+      "step": 1092
+    },
+    {
+      "epoch": 1.712421711899791,
+      "grad_norm": 1.0788499116897583,
+      "learning_rate": 4.174600126424797e-06,
+      "loss": 0.7596,
+      "step": 1093
+    },
+    {
+      "epoch": 1.7139874739039667,
+      "grad_norm": 1.0846409797668457,
+      "learning_rate": 4.173035255187302e-06,
+      "loss": 0.7322,
+      "step": 1094
+    },
+    {
+      "epoch": 1.715553235908142,
+      "grad_norm": 1.1376627683639526,
+      "learning_rate": 4.171469195849782e-06,
+      "loss": 0.6934,
+      "step": 1095
+    },
+    {
+      "epoch": 1.7171189979123174,
+      "grad_norm": 1.003432035446167,
+      "learning_rate": 4.169901949524368e-06,
+      "loss": 0.7663,
+      "step": 1096
+    },
+    {
+      "epoch": 1.7186847599164927,
+      "grad_norm": 0.9824592471122742,
+      "learning_rate": 4.168333517324036e-06,
+      "loss": 0.8253,
+      "step": 1097
+    },
+    {
+      "epoch": 1.720250521920668,
+      "grad_norm": 1.0256636142730713,
+      "learning_rate": 4.1667639003626e-06,
+      "loss": 0.6614,
+      "step": 1098
+    },
+    {
+      "epoch": 1.7218162839248434,
+      "grad_norm": 1.0290439128875732,
+      "learning_rate": 4.16519309975472e-06,
+      "loss": 0.7403,
+      "step": 1099
+    },
+    {
+      "epoch": 1.7233820459290188,
+      "grad_norm": 1.0358202457427979,
+      "learning_rate": 4.163621116615892e-06,
+      "loss": 0.7332,
+      "step": 1100
+    },
+    {
+      "epoch": 1.7249478079331941,
+      "grad_norm": 1.1046626567840576,
+      "learning_rate": 4.162047952062456e-06,
+      "loss": 0.6672,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7265135699373695,
+      "grad_norm": 1.116408109664917,
+      "learning_rate": 4.16047360721159e-06,
+      "loss": 0.6997,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7280793319415448,
+      "grad_norm": 0.98244309425354,
+      "learning_rate": 4.158898083181308e-06,
+      "loss": 0.754,
+      "step": 1103
+    },
+    {
+      "epoch": 1.7296450939457202,
+      "grad_norm": 1.033761978149414,
+      "learning_rate": 4.1573213810904635e-06,
+      "loss": 0.7317,
+      "step": 1104
+    },
+    {
+      "epoch": 1.7312108559498958,
+      "grad_norm": 1.1415889263153076,
+      "learning_rate": 4.155743502058745e-06,
+      "loss": 0.8361,
+      "step": 1105
+    },
+    {
+      "epoch": 1.732776617954071,
+      "grad_norm": 0.9897739291191101,
+      "learning_rate": 4.154164447206679e-06,
+      "loss": 0.7361,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7343423799582465,
+      "grad_norm": 1.0837332010269165,
+      "learning_rate": 4.1525842176556235e-06,
+      "loss": 0.7007,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7359081419624216,
+      "grad_norm": 1.021332025527954,
+      "learning_rate": 4.151002814527774e-06,
+      "loss": 0.723,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7374739039665972,
+      "grad_norm": 0.9808626770973206,
+      "learning_rate": 4.14942023894616e-06,
+      "loss": 0.6121,
+      "step": 1109
+    },
+    {
+      "epoch": 1.7390396659707723,
+      "grad_norm": 1.0266727209091187,
+      "learning_rate": 4.14783649203464e-06,
+      "loss": 0.8221,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7406054279749479,
+      "grad_norm": 1.0880744457244873,
+      "learning_rate": 4.146251574917907e-06,
+      "loss": 0.7159,
+      "step": 1111
+    },
+    {
+      "epoch": 1.742171189979123,
+      "grad_norm": 1.0125631093978882,
+      "learning_rate": 4.144665488721483e-06,
+      "loss": 0.7228,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7437369519832986,
+      "grad_norm": 0.9339308142662048,
+      "learning_rate": 4.143078234571723e-06,
+      "loss": 0.7472,
+      "step": 1113
+    },
+    {
+      "epoch": 1.745302713987474,
+      "grad_norm": 1.1234889030456543,
+      "learning_rate": 4.141489813595809e-06,
+      "loss": 0.7875,
+      "step": 1114
+    },
+    {
+      "epoch": 1.7468684759916493,
+      "grad_norm": 1.0862795114517212,
+      "learning_rate": 4.139900226921753e-06,
+      "loss": 0.7219,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7484342379958246,
+      "grad_norm": 1.027912974357605,
+      "learning_rate": 4.138309475678393e-06,
+      "loss": 0.6616,
+      "step": 1116
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.0271416902542114,
+      "learning_rate": 4.136717560995398e-06,
+      "loss": 0.7245,
+      "step": 1117
+    },
+    {
+      "epoch": 1.7515657620041754,
+      "grad_norm": 0.9430286288261414,
+      "learning_rate": 4.135124484003258e-06,
+      "loss": 0.7347,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7531315240083507,
+      "grad_norm": 1.1191871166229248,
+      "learning_rate": 4.133530245833292e-06,
+      "loss": 0.6713,
+      "step": 1119
+    },
+    {
+      "epoch": 1.754697286012526,
+      "grad_norm": 1.0734976530075073,
+      "learning_rate": 4.131934847617642e-06,
+      "loss": 0.7268,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7562630480167014,
+      "grad_norm": 1.0727618932724,
+      "learning_rate": 4.130338290489276e-06,
+      "loss": 0.7855,
+      "step": 1121
+    },
+    {
+      "epoch": 1.757828810020877,
+      "grad_norm": 1.0351287126541138,
+      "learning_rate": 4.1287405755819805e-06,
+      "loss": 0.7297,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7593945720250521,
+      "grad_norm": 1.046382188796997,
+      "learning_rate": 4.127141704030371e-06,
+      "loss": 0.6508,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7609603340292277,
+      "grad_norm": 1.0040491819381714,
+      "learning_rate": 4.125541676969876e-06,
+      "loss": 0.7707,
+      "step": 1124
+    },
+    {
+      "epoch": 1.7625260960334028,
+      "grad_norm": 1.1626756191253662,
+      "learning_rate": 4.123940495536752e-06,
+      "loss": 0.7366,
+      "step": 1125
+    },
+    {
+      "epoch": 1.7640918580375784,
+      "grad_norm": 0.9720292091369629,
+      "learning_rate": 4.122338160868071e-06,
+      "loss": 0.7641,
+      "step": 1126
+    },
+    {
+      "epoch": 1.7656576200417535,
+      "grad_norm": 0.9957835078239441,
+      "learning_rate": 4.1207346741017265e-06,
+      "loss": 0.7398,
+      "step": 1127
+    },
+    {
+      "epoch": 1.767223382045929,
+      "grad_norm": 0.972172737121582,
+      "learning_rate": 4.1191300363764276e-06,
+      "loss": 0.6565,
+      "step": 1128
+    },
+    {
+      "epoch": 1.7687891440501042,
+      "grad_norm": 1.0128676891326904,
+      "learning_rate": 4.117524248831702e-06,
+      "loss": 0.7433,
+      "step": 1129
+    },
+    {
+      "epoch": 1.7703549060542798,
+      "grad_norm": 1.1161550283432007,
+      "learning_rate": 4.115917312607895e-06,
+      "loss": 0.8241,
+      "step": 1130
+    },
+    {
+      "epoch": 1.7719206680584552,
+      "grad_norm": 1.009810209274292,
+      "learning_rate": 4.114309228846167e-06,
+      "loss": 0.6939,
+      "step": 1131
+    },
+    {
+      "epoch": 1.7734864300626305,
+      "grad_norm": 1.0082894563674927,
+      "learning_rate": 4.112699998688492e-06,
+      "loss": 0.6772,
+      "step": 1132
+    },
+    {
+      "epoch": 1.7750521920668059,
+      "grad_norm": 0.9453067183494568,
+      "learning_rate": 4.111089623277658e-06,
+      "loss": 0.6982,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7766179540709812,
+      "grad_norm": 1.074747920036316,
+      "learning_rate": 4.1094781037572696e-06,
+      "loss": 0.7296,
+      "step": 1134
+    },
+    {
+      "epoch": 1.7781837160751566,
+      "grad_norm": 1.0477592945098877,
+      "learning_rate": 4.1078654412717404e-06,
+      "loss": 0.8162,
+      "step": 1135
+    },
+    {
+      "epoch": 1.779749478079332,
+      "grad_norm": 1.1362730264663696,
+      "learning_rate": 4.106251636966297e-06,
+      "loss": 0.6942,
+      "step": 1136
+    },
+    {
+      "epoch": 1.7813152400835073,
+      "grad_norm": 0.9564191102981567,
+      "learning_rate": 4.104636691986977e-06,
+      "loss": 0.7613,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7828810020876826,
+      "grad_norm": 1.0384364128112793,
+      "learning_rate": 4.103020607480628e-06,
+      "loss": 0.6482,
+      "step": 1138
+    },
+    {
+      "epoch": 1.784446764091858,
+      "grad_norm": 1.1258740425109863,
+      "learning_rate": 4.101403384594905e-06,
+      "loss": 0.7736,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7860125260960333,
+      "grad_norm": 1.0740283727645874,
+      "learning_rate": 4.099785024478276e-06,
+      "loss": 0.7219,
+      "step": 1140
+    },
+    {
+      "epoch": 1.787578288100209,
+      "grad_norm": 1.0377696752548218,
+      "learning_rate": 4.098165528280012e-06,
+      "loss": 0.7132,
+      "step": 1141
+    },
+    {
+      "epoch": 1.789144050104384,
+      "grad_norm": 1.0551276206970215,
+      "learning_rate": 4.096544897150192e-06,
+      "loss": 0.7879,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7907098121085596,
+      "grad_norm": 1.096389651298523,
+      "learning_rate": 4.0949231322397036e-06,
+      "loss": 0.7516,
+      "step": 1143
+    },
+    {
+      "epoch": 1.7922755741127347,
+      "grad_norm": 0.9602596163749695,
+      "learning_rate": 4.093300234700234e-06,
+      "loss": 0.7197,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7938413361169103,
+      "grad_norm": 1.1918690204620361,
+      "learning_rate": 4.09167620568428e-06,
+      "loss": 0.7634,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7954070981210855,
+      "grad_norm": 1.1249059438705444,
+      "learning_rate": 4.09005104634514e-06,
+      "loss": 0.7718,
+      "step": 1146
+    },
+    {
+      "epoch": 1.796972860125261,
+      "grad_norm": 1.0596246719360352,
+      "learning_rate": 4.0884247578369164e-06,
+      "loss": 0.7258,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7985386221294362,
+      "grad_norm": 1.026201844215393,
+      "learning_rate": 4.086797341314509e-06,
+      "loss": 0.6426,
+      "step": 1148
+    },
+    {
+      "epoch": 1.8001043841336117,
+      "grad_norm": 1.0791343450546265,
+      "learning_rate": 4.085168797933625e-06,
+      "loss": 0.7147,
+      "step": 1149
+    },
+    {
+      "epoch": 1.801670146137787,
+      "grad_norm": 0.9660344123840332,
+      "learning_rate": 4.083539128850766e-06,
+      "loss": 0.7335,
+      "step": 1150
+    },
+    {
+      "epoch": 1.8032359081419624,
+      "grad_norm": 1.0022811889648438,
+      "learning_rate": 4.081908335223238e-06,
+      "loss": 0.6824,
+      "step": 1151
+    },
+    {
+      "epoch": 1.8048016701461378,
+      "grad_norm": 0.9261506199836731,
+      "learning_rate": 4.080276418209143e-06,
+      "loss": 0.7538,
+      "step": 1152
+    },
+    {
+      "epoch": 1.8063674321503131,
+      "grad_norm": 1.088552713394165,
+      "learning_rate": 4.078643378967379e-06,
+      "loss": 0.6988,
+      "step": 1153
+    },
+    {
+      "epoch": 1.8079331941544885,
+      "grad_norm": 1.0581461191177368,
+      "learning_rate": 4.077009218657645e-06,
+      "loss": 0.7075,
+      "step": 1154
+    },
+    {
+      "epoch": 1.8094989561586639,
+      "grad_norm": 1.0722301006317139,
+      "learning_rate": 4.0753739384404335e-06,
+      "loss": 0.6974,
+      "step": 1155
+    },
+    {
+      "epoch": 1.8110647181628392,
+      "grad_norm": 0.9519115090370178,
+      "learning_rate": 4.073737539477033e-06,
+      "loss": 0.6898,
+      "step": 1156
+    },
+    {
+      "epoch": 1.8126304801670146,
+      "grad_norm": 1.1201752424240112,
+      "learning_rate": 4.072100022929526e-06,
+      "loss": 0.8572,
+      "step": 1157
+    },
+    {
+      "epoch": 1.8141962421711901,
+      "grad_norm": 1.24104642868042,
+      "learning_rate": 4.070461389960789e-06,
+      "loss": 0.7407,
+      "step": 1158
+    },
+    {
+      "epoch": 1.8157620041753653,
+      "grad_norm": 1.0625057220458984,
+      "learning_rate": 4.068821641734492e-06,
+      "loss": 0.655,
+      "step": 1159
+    },
+    {
+      "epoch": 1.8173277661795408,
+      "grad_norm": 1.1174049377441406,
+      "learning_rate": 4.067180779415097e-06,
+      "loss": 0.7214,
+      "step": 1160
+    },
+    {
+      "epoch": 1.818893528183716,
+      "grad_norm": 1.0683097839355469,
+      "learning_rate": 4.065538804167855e-06,
+      "loss": 0.7863,
+      "step": 1161
+    },
+    {
+      "epoch": 1.8204592901878915,
+      "grad_norm": 1.031198263168335,
+      "learning_rate": 4.063895717158809e-06,
+      "loss": 0.7663,
+      "step": 1162
+    },
+    {
+      "epoch": 1.8220250521920667,
+      "grad_norm": 1.1091554164886475,
+      "learning_rate": 4.062251519554791e-06,
+      "loss": 0.7234,
+      "step": 1163
+    },
+    {
+      "epoch": 1.8235908141962422,
+      "grad_norm": 1.0818198919296265,
+      "learning_rate": 4.060606212523425e-06,
+      "loss": 0.7709,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8251565762004174,
+      "grad_norm": 1.0104292631149292,
+      "learning_rate": 4.058959797233116e-06,
+      "loss": 0.7225,
+      "step": 1165
+    },
+    {
+      "epoch": 1.826722338204593,
+      "grad_norm": 1.1726491451263428,
+      "learning_rate": 4.057312274853063e-06,
+      "loss": 0.6831,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8282881002087683,
+      "grad_norm": 0.9735565185546875,
+      "learning_rate": 4.055663646553248e-06,
+      "loss": 0.7123,
+      "step": 1167
+    },
+    {
+      "epoch": 1.8298538622129437,
+      "grad_norm": 1.011709213256836,
+      "learning_rate": 4.054013913504436e-06,
+      "loss": 0.6885,
+      "step": 1168
+    },
+    {
+      "epoch": 1.831419624217119,
+      "grad_norm": 1.0374364852905273,
+      "learning_rate": 4.05236307687818e-06,
+      "loss": 0.782,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8329853862212944,
+      "grad_norm": 1.011202096939087,
+      "learning_rate": 4.050711137846817e-06,
+      "loss": 0.7381,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8345511482254697,
+      "grad_norm": 1.0984755754470825,
+      "learning_rate": 4.049058097583464e-06,
+      "loss": 0.7472,
+      "step": 1171
+    },
+    {
+      "epoch": 1.836116910229645,
+      "grad_norm": 1.090078353881836,
+      "learning_rate": 4.047403957262024e-06,
+      "loss": 0.6773,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8376826722338204,
+      "grad_norm": 0.9866096377372742,
+      "learning_rate": 4.045748718057176e-06,
+      "loss": 0.6812,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8392484342379958,
+      "grad_norm": 1.1006139516830444,
+      "learning_rate": 4.0440923811443846e-06,
+      "loss": 0.7953,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8408141962421714,
+      "grad_norm": 0.9621697664260864,
+      "learning_rate": 4.04243494769989e-06,
+      "loss": 0.7557,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8423799582463465,
+      "grad_norm": 1.1520715951919556,
+      "learning_rate": 4.040776418900714e-06,
+      "loss": 0.7422,
+      "step": 1176
+    },
+    {
+      "epoch": 1.843945720250522,
+      "grad_norm": 1.0679237842559814,
+      "learning_rate": 4.039116795924653e-06,
+      "loss": 0.6738,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8455114822546972,
+      "grad_norm": 1.026510238647461,
+      "learning_rate": 4.037456079950284e-06,
+      "loss": 0.7741,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8470772442588728,
+      "grad_norm": 1.0111522674560547,
+      "learning_rate": 4.0357942721569585e-06,
+      "loss": 0.6968,
+      "step": 1179
+    },
+    {
+      "epoch": 1.848643006263048,
+      "grad_norm": 1.0290493965148926,
+      "learning_rate": 4.034131373724802e-06,
+      "loss": 0.7131,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8502087682672235,
+      "grad_norm": 0.9878649711608887,
+      "learning_rate": 4.032467385834718e-06,
+      "loss": 0.7477,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8517745302713986,
+      "grad_norm": 1.0317823886871338,
+      "learning_rate": 4.030802309668379e-06,
+      "loss": 0.6318,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8533402922755742,
+      "grad_norm": 1.0163817405700684,
+      "learning_rate": 4.029136146408235e-06,
+      "loss": 0.7027,
+      "step": 1183
+    },
+    {
+      "epoch": 1.8549060542797495,
+      "grad_norm": 0.9590600728988647,
+      "learning_rate": 4.027468897237505e-06,
+      "loss": 0.6988,
+      "step": 1184
+    },
+    {
+      "epoch": 1.8564718162839249,
+      "grad_norm": 1.0645931959152222,
+      "learning_rate": 4.0258005633401795e-06,
+      "loss": 0.7588,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8580375782881002,
+      "grad_norm": 1.0380210876464844,
+      "learning_rate": 4.024131145901022e-06,
+      "loss": 0.6984,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8596033402922756,
+      "grad_norm": 1.0558087825775146,
+      "learning_rate": 4.022460646105561e-06,
+      "loss": 0.6729,
+      "step": 1187
+    },
+    {
+      "epoch": 1.861169102296451,
+      "grad_norm": 1.0865014791488647,
+      "learning_rate": 4.020789065140097e-06,
+      "loss": 0.7208,
+      "step": 1188
+    },
+    {
+      "epoch": 1.8627348643006263,
+      "grad_norm": 1.0456374883651733,
+      "learning_rate": 4.019116404191697e-06,
+      "loss": 0.7759,
+      "step": 1189
+    },
+    {
+      "epoch": 1.8643006263048016,
+      "grad_norm": 0.9724026918411255,
+      "learning_rate": 4.017442664448197e-06,
+      "loss": 0.7475,
+      "step": 1190
+    },
+    {
+      "epoch": 1.865866388308977,
+      "grad_norm": 1.1170289516448975,
+      "learning_rate": 4.015767847098194e-06,
+      "loss": 0.6993,
+      "step": 1191
+    },
+    {
+      "epoch": 1.8674321503131524,
+      "grad_norm": 1.1052978038787842,
+      "learning_rate": 4.014091953331056e-06,
+      "loss": 0.748,
+      "step": 1192
+    },
+    {
+      "epoch": 1.8689979123173277,
+      "grad_norm": 1.1223454475402832,
+      "learning_rate": 4.012414984336912e-06,
+      "loss": 0.6921,
+      "step": 1193
+    },
+    {
+      "epoch": 1.8705636743215033,
+      "grad_norm": 1.0611138343811035,
+      "learning_rate": 4.010736941306655e-06,
+      "loss": 0.7859,
+      "step": 1194
+    },
+    {
+      "epoch": 1.8721294363256784,
+      "grad_norm": 1.169189691543579,
+      "learning_rate": 4.009057825431944e-06,
+      "loss": 0.5874,
+      "step": 1195
+    },
+    {
+      "epoch": 1.873695198329854,
+      "grad_norm": 1.036626935005188,
+      "learning_rate": 4.0073776379051945e-06,
+      "loss": 0.6801,
+      "step": 1196
+    },
+    {
+      "epoch": 1.8752609603340291,
+      "grad_norm": 0.9431620240211487,
+      "learning_rate": 4.005696379919586e-06,
+      "loss": 0.7476,
+      "step": 1197
+    },
+    {
+      "epoch": 1.8768267223382047,
+      "grad_norm": 1.0381059646606445,
+      "learning_rate": 4.0040140526690585e-06,
+      "loss": 0.6691,
+      "step": 1198
+    },
+    {
+      "epoch": 1.8783924843423798,
+      "grad_norm": 1.120033860206604,
+      "learning_rate": 4.0023306573483105e-06,
+      "loss": 0.6669,
+      "step": 1199
+    },
+    {
+      "epoch": 1.8799582463465554,
+      "grad_norm": 1.0298967361450195,
+      "learning_rate": 4.000646195152798e-06,
+      "loss": 0.7355,
+      "step": 1200
+    },
+    {
+      "epoch": 1.8815240083507305,
+      "grad_norm": 1.0036594867706299,
+      "learning_rate": 3.99896066727874e-06,
+      "loss": 0.7545,
+      "step": 1201
+    },
+    {
+      "epoch": 1.883089770354906,
+      "grad_norm": 1.1166342496871948,
+      "learning_rate": 3.997274074923103e-06,
+      "loss": 0.7468,
+      "step": 1202
+    },
+    {
+      "epoch": 1.8846555323590815,
+      "grad_norm": 1.0512094497680664,
+      "learning_rate": 3.995586419283617e-06,
+      "loss": 0.7647,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8862212943632568,
+      "grad_norm": 1.0115973949432373,
+      "learning_rate": 3.993897701558764e-06,
+      "loss": 0.7893,
+      "step": 1204
+    },
+    {
+      "epoch": 1.8877870563674322,
+      "grad_norm": 1.0839310884475708,
+      "learning_rate": 3.992207922947783e-06,
+      "loss": 0.6824,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8893528183716075,
+      "grad_norm": 0.9987578988075256,
+      "learning_rate": 3.99051708465066e-06,
+      "loss": 0.8045,
+      "step": 1206
+    },
+    {
+      "epoch": 1.8909185803757829,
+      "grad_norm": 1.0759059190750122,
+      "learning_rate": 3.9888251878681425e-06,
+      "loss": 0.7756,
+      "step": 1207
+    },
+    {
+      "epoch": 1.8924843423799582,
+      "grad_norm": 1.1055301427841187,
+      "learning_rate": 3.987132233801722e-06,
+      "loss": 0.6947,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8940501043841336,
+      "grad_norm": 1.042575478553772,
+      "learning_rate": 3.985438223653644e-06,
+      "loss": 0.6669,
+      "step": 1209
+    },
+    {
+      "epoch": 1.895615866388309,
+      "grad_norm": 1.0595242977142334,
+      "learning_rate": 3.983743158626904e-06,
+      "loss": 0.6937,
+      "step": 1210
+    },
+    {
+      "epoch": 1.8971816283924845,
+      "grad_norm": 1.0498789548873901,
+      "learning_rate": 3.982047039925247e-06,
+      "loss": 0.7244,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8987473903966596,
+      "grad_norm": 1.0960590839385986,
+      "learning_rate": 3.980349868753166e-06,
+      "loss": 0.7661,
+      "step": 1212
+    },
+    {
+      "epoch": 1.9003131524008352,
+      "grad_norm": 1.108823299407959,
+      "learning_rate": 3.9786516463159e-06,
+      "loss": 0.6937,
+      "step": 1213
+    },
+    {
+      "epoch": 1.9018789144050103,
+      "grad_norm": 1.051416039466858,
+      "learning_rate": 3.976952373819436e-06,
+      "loss": 0.7263,
+      "step": 1214
+    },
+    {
+      "epoch": 1.903444676409186,
+      "grad_norm": 0.9884061217308044,
+      "learning_rate": 3.9752520524705065e-06,
+      "loss": 0.6577,
+      "step": 1215
+    },
+    {
+      "epoch": 1.905010438413361,
+      "grad_norm": 1.0950369834899902,
+      "learning_rate": 3.9735506834765885e-06,
+      "loss": 0.7376,
+      "step": 1216
+    },
+    {
+      "epoch": 1.9065762004175366,
+      "grad_norm": 1.1140044927597046,
+      "learning_rate": 3.971848268045902e-06,
+      "loss": 0.7125,
+      "step": 1217
+    },
+    {
+      "epoch": 1.9081419624217117,
+      "grad_norm": 1.0908716917037964,
+      "learning_rate": 3.970144807387414e-06,
+      "loss": 0.7796,
+      "step": 1218
+    },
+    {
+      "epoch": 1.9097077244258873,
+      "grad_norm": 1.1068549156188965,
+      "learning_rate": 3.968440302710829e-06,
+      "loss": 0.7454,
+      "step": 1219
+    },
+    {
+      "epoch": 1.9112734864300627,
+      "grad_norm": 1.102594017982483,
+      "learning_rate": 3.9667347552265945e-06,
+      "loss": 0.757,
+      "step": 1220
+    },
+    {
+      "epoch": 1.912839248434238,
+      "grad_norm": 1.1588326692581177,
+      "learning_rate": 3.9650281661459015e-06,
+      "loss": 0.6491,
+      "step": 1221
+    },
+    {
+      "epoch": 1.9144050104384134,
+      "grad_norm": 0.9896461367607117,
+      "learning_rate": 3.963320536680675e-06,
+      "loss": 0.773,
+      "step": 1222
+    },
+    {
+      "epoch": 1.9159707724425887,
+      "grad_norm": 1.0855952501296997,
+      "learning_rate": 3.961611868043583e-06,
+      "loss": 0.6906,
+      "step": 1223
+    },
+    {
+      "epoch": 1.917536534446764,
+      "grad_norm": 1.0556972026824951,
+      "learning_rate": 3.959902161448032e-06,
+      "loss": 0.7577,
+      "step": 1224
+    },
+    {
+      "epoch": 1.9191022964509394,
+      "grad_norm": 1.0284528732299805,
+      "learning_rate": 3.958191418108162e-06,
+      "loss": 0.6844,
+      "step": 1225
+    },
+    {
+      "epoch": 1.9206680584551148,
+      "grad_norm": 0.9578842520713806,
+      "learning_rate": 3.956479639238852e-06,
+      "loss": 0.7216,
+      "step": 1226
+    },
+    {
+      "epoch": 1.9222338204592901,
+      "grad_norm": 1.0366833209991455,
+      "learning_rate": 3.954766826055717e-06,
+      "loss": 0.7365,
+      "step": 1227
+    },
+    {
+      "epoch": 1.9237995824634657,
+      "grad_norm": 1.0498946905136108,
+      "learning_rate": 3.953052979775103e-06,
+      "loss": 0.6788,
+      "step": 1228
+    },
+    {
+      "epoch": 1.9253653444676408,
+      "grad_norm": 1.010860562324524,
+      "learning_rate": 3.951338101614094e-06,
+      "loss": 0.7181,
+      "step": 1229
+    },
+    {
+      "epoch": 1.9269311064718164,
+      "grad_norm": 1.0203677415847778,
+      "learning_rate": 3.949622192790502e-06,
+      "loss": 0.7277,
+      "step": 1230
+    },
+    {
+      "epoch": 1.9284968684759916,
+      "grad_norm": 0.9907971024513245,
+      "learning_rate": 3.947905254522876e-06,
+      "loss": 0.6858,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9300626304801671,
+      "grad_norm": 0.9824337363243103,
+      "learning_rate": 3.946187288030493e-06,
+      "loss": 0.7517,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9316283924843423,
+      "grad_norm": 1.0206327438354492,
+      "learning_rate": 3.944468294533359e-06,
+      "loss": 0.7225,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9331941544885178,
+      "grad_norm": 1.0529900789260864,
+      "learning_rate": 3.942748275252214e-06,
+      "loss": 0.7965,
+      "step": 1234
+    },
+    {
+      "epoch": 1.934759916492693,
+      "grad_norm": 1.0501983165740967,
+      "learning_rate": 3.941027231408521e-06,
+      "loss": 0.7646,
+      "step": 1235
+    },
+    {
+      "epoch": 1.9363256784968685,
+      "grad_norm": 0.9482662081718445,
+      "learning_rate": 3.939305164224474e-06,
+      "loss": 0.7449,
+      "step": 1236
+    },
+    {
+      "epoch": 1.937891440501044,
+      "grad_norm": 1.096723198890686,
+      "learning_rate": 3.937582074922993e-06,
+      "loss": 0.7464,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9394572025052192,
+      "grad_norm": 1.0417059659957886,
+      "learning_rate": 3.935857964727725e-06,
+      "loss": 0.6586,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9410229645093946,
+      "grad_norm": 1.1008849143981934,
+      "learning_rate": 3.934132834863039e-06,
+      "loss": 0.6236,
+      "step": 1239
+    },
+    {
+      "epoch": 1.94258872651357,
+      "grad_norm": 0.9107227921485901,
+      "learning_rate": 3.93240668655403e-06,
+      "loss": 0.7069,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9441544885177453,
+      "grad_norm": 1.0516945123672485,
+      "learning_rate": 3.930679521026517e-06,
+      "loss": 0.773,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9457202505219207,
+      "grad_norm": 0.9941951632499695,
+      "learning_rate": 3.92895133950704e-06,
+      "loss": 0.6002,
+      "step": 1242
+    },
+    {
+      "epoch": 1.947286012526096,
+      "grad_norm": 1.031784176826477,
+      "learning_rate": 3.927222143222859e-06,
+      "loss": 0.7737,
+      "step": 1243
+    },
+    {
+      "epoch": 1.9488517745302714,
+      "grad_norm": 0.986473798751831,
+      "learning_rate": 3.925491933401961e-06,
+      "loss": 0.7455,
+      "step": 1244
+    },
+    {
+      "epoch": 1.9504175365344467,
+      "grad_norm": 1.109964370727539,
+      "learning_rate": 3.923760711273044e-06,
+      "loss": 0.7048,
+      "step": 1245
+    },
+    {
+      "epoch": 1.951983298538622,
+      "grad_norm": 1.1303372383117676,
+      "learning_rate": 3.922028478065532e-06,
+      "loss": 0.749,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9535490605427976,
+      "grad_norm": 0.999824583530426,
+      "learning_rate": 3.920295235009562e-06,
+      "loss": 0.7596,
+      "step": 1247
+    },
+    {
+      "epoch": 1.9551148225469728,
+      "grad_norm": 1.0440820455551147,
+      "learning_rate": 3.918560983335993e-06,
+      "loss": 0.7841,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9566805845511483,
+      "grad_norm": 1.0420323610305786,
+      "learning_rate": 3.916825724276395e-06,
+      "loss": 0.7837,
+      "step": 1249
+    },
+    {
+      "epoch": 1.9582463465553235,
+      "grad_norm": 1.001997947692871,
+      "learning_rate": 3.915089459063058e-06,
+      "loss": 0.6956,
+      "step": 1250
+    },
+    {
+      "epoch": 1.959812108559499,
+      "grad_norm": 1.030626893043518,
+      "learning_rate": 3.913352188928984e-06,
+      "loss": 0.74,
+      "step": 1251
+    },
+    {
+      "epoch": 1.9613778705636742,
+      "grad_norm": 1.0981760025024414,
+      "learning_rate": 3.911613915107888e-06,
+      "loss": 0.713,
+      "step": 1252
+    },
+    {
+      "epoch": 1.9629436325678498,
+      "grad_norm": 1.020772933959961,
+      "learning_rate": 3.9098746388342006e-06,
+      "loss": 0.7605,
+      "step": 1253
+    },
+    {
+      "epoch": 1.964509394572025,
+      "grad_norm": 1.0206503868103027,
+      "learning_rate": 3.908134361343062e-06,
+      "loss": 0.7176,
+      "step": 1254
+    },
+    {
+      "epoch": 1.9660751565762005,
+      "grad_norm": 1.0471689701080322,
+      "learning_rate": 3.906393083870325e-06,
+      "loss": 0.7698,
+      "step": 1255
+    },
+    {
+      "epoch": 1.9676409185803758,
+      "grad_norm": 1.0658202171325684,
+      "learning_rate": 3.904650807652549e-06,
+      "loss": 0.7179,
+      "step": 1256
+    },
+    {
+      "epoch": 1.9692066805845512,
+      "grad_norm": 1.196192741394043,
+      "learning_rate": 3.902907533927007e-06,
+      "loss": 0.7801,
+      "step": 1257
+    },
+    {
+      "epoch": 1.9707724425887265,
+      "grad_norm": 1.0874967575073242,
+      "learning_rate": 3.90116326393168e-06,
+      "loss": 0.7672,
+      "step": 1258
+    },
+    {
+      "epoch": 1.9723382045929019,
+      "grad_norm": 1.153617024421692,
+      "learning_rate": 3.899417998905253e-06,
+      "loss": 0.752,
+      "step": 1259
+    },
+    {
+      "epoch": 1.9739039665970772,
+      "grad_norm": 1.0750453472137451,
+      "learning_rate": 3.89767174008712e-06,
+      "loss": 0.7471,
+      "step": 1260
+    },
+    {
+      "epoch": 1.9754697286012526,
+      "grad_norm": 1.0693799257278442,
+      "learning_rate": 3.895924488717381e-06,
+      "loss": 0.6704,
+      "step": 1261
+    },
+    {
+      "epoch": 1.977035490605428,
+      "grad_norm": 0.9552210569381714,
+      "learning_rate": 3.89417624603684e-06,
+      "loss": 0.7793,
+      "step": 1262
+    },
+    {
+      "epoch": 1.9786012526096033,
+      "grad_norm": 0.96712726354599,
+      "learning_rate": 3.892427013287007e-06,
+      "loss": 0.6843,
+      "step": 1263
+    },
+    {
+      "epoch": 1.9801670146137789,
+      "grad_norm": 1.1453518867492676,
+      "learning_rate": 3.890676791710091e-06,
+      "loss": 0.7532,
+      "step": 1264
+    },
+    {
+      "epoch": 1.981732776617954,
+      "grad_norm": 1.0294159650802612,
+      "learning_rate": 3.888925582549006e-06,
+      "loss": 0.703,
+      "step": 1265
+    },
+    {
+      "epoch": 1.9832985386221296,
+      "grad_norm": 1.0790715217590332,
+      "learning_rate": 3.8871733870473686e-06,
+      "loss": 0.6731,
+      "step": 1266
+    },
+    {
+      "epoch": 1.9848643006263047,
+      "grad_norm": 1.052460789680481,
+      "learning_rate": 3.8854202064494915e-06,
+      "loss": 0.748,
+      "step": 1267
+    },
+    {
+      "epoch": 1.9864300626304803,
+      "grad_norm": 1.112462043762207,
+      "learning_rate": 3.883666042000392e-06,
+      "loss": 0.7408,
+      "step": 1268
+    },
+    {
+      "epoch": 1.9879958246346554,
+      "grad_norm": 1.0738476514816284,
+      "learning_rate": 3.881910894945783e-06,
+      "loss": 0.759,
+      "step": 1269
+    },
+    {
+      "epoch": 1.989561586638831,
+      "grad_norm": 1.136472463607788,
+      "learning_rate": 3.8801547665320775e-06,
+      "loss": 0.754,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9911273486430061,
+      "grad_norm": 1.0640734434127808,
+      "learning_rate": 3.87839765800638e-06,
+      "loss": 0.719,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9926931106471817,
+      "grad_norm": 0.9155470132827759,
+      "learning_rate": 3.876639570616498e-06,
+      "loss": 0.762,
+      "step": 1272
+    },
+    {
+      "epoch": 1.994258872651357,
+      "grad_norm": 1.015743374824524,
+      "learning_rate": 3.87488050561093e-06,
+      "loss": 0.6926,
+      "step": 1273
+    },
+    {
+      "epoch": 1.9958246346555324,
+      "grad_norm": 1.0829862356185913,
+      "learning_rate": 3.87312046423887e-06,
+      "loss": 0.7644,
+      "step": 1274
+    },
+    {
+      "epoch": 1.9973903966597077,
+      "grad_norm": 1.0962239503860474,
+      "learning_rate": 3.871359447750205e-06,
+      "loss": 0.7038,
+      "step": 1275
+    },
+    {
+      "epoch": 1.998956158663883,
+      "grad_norm": 1.0470675230026245,
+      "learning_rate": 3.869597457395514e-06,
+      "loss": 0.7629,
+      "step": 1276
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3828,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 638,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.224064837298422e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}