{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0282,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002,
      "grad_norm": 12.427441596984863,
      "learning_rate": 1.8e-07,
      "loss": 0.4661,
      "step": 10
    },
    {
      "epoch": 0.004,
      "grad_norm": 10.644728660583496,
      "learning_rate": 3.8e-07,
      "loss": 0.43,
      "step": 20
    },
    {
      "epoch": 0.006,
      "grad_norm": 2.582233428955078,
      "learning_rate": 5.800000000000001e-07,
      "loss": 0.2937,
      "step": 30
    },
    {
      "epoch": 0.008,
      "grad_norm": 1.6104474067687988,
      "learning_rate": 7.8e-07,
      "loss": 0.2311,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3486788272857666,
      "learning_rate": 9.800000000000001e-07,
      "loss": 0.1944,
      "step": 50
    },
    {
      "epoch": 0.012,
      "grad_norm": 2.3484883308410645,
      "learning_rate": 1.1800000000000001e-06,
      "loss": 0.178,
      "step": 60
    },
    {
      "epoch": 0.014,
      "grad_norm": 2.1255300045013428,
      "learning_rate": 1.3800000000000001e-06,
      "loss": 0.1663,
      "step": 70
    },
    {
      "epoch": 0.016,
      "grad_norm": 1.2671189308166504,
      "learning_rate": 1.5800000000000001e-06,
      "loss": 0.1521,
      "step": 80
    },
    {
      "epoch": 0.018,
      "grad_norm": 2.0237274169921875,
      "learning_rate": 1.7800000000000001e-06,
      "loss": 0.1515,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4099950790405273,
      "learning_rate": 1.98e-06,
      "loss": 0.1481,
      "step": 100
    },
    {
      "epoch": 0.022,
      "grad_norm": 2.23960280418396,
      "learning_rate": 2.1800000000000003e-06,
      "loss": 0.1487,
      "step": 110
    },
    {
      "epoch": 0.024,
      "grad_norm": 1.7865636348724365,
      "learning_rate": 2.38e-06,
      "loss": 0.1359,
      "step": 120
    },
    {
      "epoch": 0.026,
      "grad_norm": 1.590512990951538,
      "learning_rate": 2.5800000000000003e-06,
      "loss": 0.1487,
      "step": 130
    },
    {
      "epoch": 0.028,
      "grad_norm": 3.781755208969116,
      "learning_rate": 2.7800000000000005e-06,
      "loss": 0.1381,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4385772943496704,
      "learning_rate": 2.9800000000000003e-06,
      "loss": 0.1384,
      "step": 150
    },
    {
      "epoch": 0.032,
      "grad_norm": 1.2209824323654175,
      "learning_rate": 3.1800000000000005e-06,
      "loss": 0.1407,
      "step": 160
    },
    {
      "epoch": 0.034,
      "grad_norm": 1.463581919670105,
      "learning_rate": 3.3800000000000007e-06,
      "loss": 0.1392,
      "step": 170
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.9795161485671997,
      "learning_rate": 3.58e-06,
      "loss": 0.1335,
      "step": 180
    },
    {
      "epoch": 0.038,
      "grad_norm": 1.5629451274871826,
      "learning_rate": 3.7800000000000002e-06,
      "loss": 0.1355,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.2003155946731567,
      "learning_rate": 3.980000000000001e-06,
      "loss": 0.136,
      "step": 200
    },
    {
      "epoch": 0.042,
      "grad_norm": 1.3819934129714966,
      "learning_rate": 4.18e-06,
      "loss": 0.1359,
      "step": 210
    },
    {
      "epoch": 0.044,
      "grad_norm": 1.797484278678894,
      "learning_rate": 4.38e-06,
      "loss": 0.1318,
      "step": 220
    },
    {
      "epoch": 0.046,
      "grad_norm": 1.15238618850708,
      "learning_rate": 4.58e-06,
      "loss": 0.1303,
      "step": 230
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.8387578129768372,
      "learning_rate": 4.78e-06,
      "loss": 0.1359,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8264948129653931,
      "learning_rate": 4.980000000000001e-06,
      "loss": 0.1316,
      "step": 250
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.9605500102043152,
      "learning_rate": 5.18e-06,
      "loss": 0.1287,
      "step": 260
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.9171805381774902,
      "learning_rate": 5.380000000000001e-06,
      "loss": 0.135,
      "step": 270
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.7537838220596313,
      "learning_rate": 5.580000000000001e-06,
      "loss": 0.1317,
      "step": 280
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.9007892608642578,
      "learning_rate": 5.78e-06,
      "loss": 0.1295,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8143233060836792,
      "learning_rate": 5.98e-06,
      "loss": 0.1337,
      "step": 300
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.7211500406265259,
      "learning_rate": 6.18e-06,
      "loss": 0.1228,
      "step": 310
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.8526885509490967,
      "learning_rate": 6.380000000000001e-06,
      "loss": 0.1331,
      "step": 320
    },
    {
      "epoch": 0.066,
      "grad_norm": 1.7408725023269653,
      "learning_rate": 6.5800000000000005e-06,
      "loss": 0.127,
      "step": 330
    },
    {
      "epoch": 0.068,
      "grad_norm": 1.5572606325149536,
      "learning_rate": 6.780000000000001e-06,
      "loss": 0.1177,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.8059284687042236,
      "learning_rate": 6.98e-06,
      "loss": 0.1267,
      "step": 350
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.755484938621521,
      "learning_rate": 7.180000000000001e-06,
      "loss": 0.1278,
      "step": 360
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.9175857901573181,
      "learning_rate": 7.3800000000000005e-06,
      "loss": 0.121,
      "step": 370
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.8510663509368896,
      "learning_rate": 7.58e-06,
      "loss": 0.115,
      "step": 380
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.9119841456413269,
      "learning_rate": 7.78e-06,
      "loss": 0.1218,
      "step": 390
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6241216063499451,
      "learning_rate": 7.980000000000002e-06,
      "loss": 0.1238,
      "step": 400
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.8066564202308655,
      "learning_rate": 8.18e-06,
      "loss": 0.1205,
      "step": 410
    },
    {
      "epoch": 0.084,
      "grad_norm": 1.104693055152893,
      "learning_rate": 8.380000000000001e-06,
      "loss": 0.1225,
      "step": 420
    },
    {
      "epoch": 0.086,
      "grad_norm": 2.131037712097168,
      "learning_rate": 8.580000000000001e-06,
      "loss": 0.1321,
      "step": 430
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.8443548679351807,
      "learning_rate": 8.78e-06,
      "loss": 0.1278,
      "step": 440
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.8610478043556213,
      "learning_rate": 8.98e-06,
      "loss": 0.1169,
      "step": 450
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.9445344805717468,
      "learning_rate": 9.180000000000002e-06,
      "loss": 0.1293,
      "step": 460
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.8306660056114197,
      "learning_rate": 9.38e-06,
      "loss": 0.1213,
      "step": 470
    },
    {
      "epoch": 0.096,
      "grad_norm": 1.2315057516098022,
      "learning_rate": 9.58e-06,
      "loss": 0.1173,
      "step": 480
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.7147797346115112,
      "learning_rate": 9.780000000000001e-06,
      "loss": 0.1138,
      "step": 490
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8065736293792725,
      "learning_rate": 9.980000000000001e-06,
      "loss": 0.1225,
      "step": 500
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.6982239484786987,
      "learning_rate": 9.999901304280686e-06,
      "loss": 0.1225,
      "step": 510
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.906414806842804,
      "learning_rate": 9.999560138895238e-06,
      "loss": 0.1225,
      "step": 520
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.6281483769416809,
      "learning_rate": 9.99897530200195e-06,
      "loss": 0.1183,
      "step": 530
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.626783549785614,
      "learning_rate": 9.998146822104943e-06,
      "loss": 0.124,
      "step": 540
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.9497509598731995,
      "learning_rate": 9.997074739583162e-06,
      "loss": 0.1193,
      "step": 550
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.7951166033744812,
      "learning_rate": 9.995759106688394e-06,
      "loss": 0.1151,
      "step": 560
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.722017228603363,
      "learning_rate": 9.99419998754273e-06,
      "loss": 0.1205,
      "step": 570
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.7518423795700073,
      "learning_rate": 9.992397458135438e-06,
      "loss": 0.1231,
      "step": 580
    },
    {
      "epoch": 0.118,
      "grad_norm": 3.67732834815979,
      "learning_rate": 9.990351606319261e-06,
      "loss": 0.116,
      "step": 590
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7335011959075928,
      "learning_rate": 9.988062531806127e-06,
      "loss": 0.1138,
      "step": 600
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.6158178448677063,
      "learning_rate": 9.9855303461623e-06,
      "loss": 0.1116,
      "step": 610
    },
    {
      "epoch": 0.124,
      "grad_norm": 1.0562201738357544,
      "learning_rate": 9.982755172802933e-06,
      "loss": 0.1122,
      "step": 620
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.7680226564407349,
      "learning_rate": 9.979737146986064e-06,
      "loss": 0.1108,
      "step": 630
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.935513973236084,
      "learning_rate": 9.976476415806013e-06,
      "loss": 0.1189,
      "step": 640
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7690663933753967,
      "learning_rate": 9.972973138186217e-06,
      "loss": 0.1201,
      "step": 650
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.8854688405990601,
      "learning_rate": 9.969227484871485e-06,
      "loss": 0.1144,
      "step": 660
    },
    {
      "epoch": 0.134,
      "grad_norm": 1.5783987045288086,
      "learning_rate": 9.965239638419673e-06,
      "loss": 0.1173,
      "step": 670
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.7775712013244629,
      "learning_rate": 9.961009793192793e-06,
      "loss": 0.1199,
      "step": 680
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.8870111107826233,
      "learning_rate": 9.956538155347534e-06,
      "loss": 0.1221,
      "step": 690
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.9003025889396667,
      "learning_rate": 9.951824942825215e-06,
      "loss": 0.1144,
      "step": 700
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.8907985687255859,
      "learning_rate": 9.946870385341167e-06,
      "loss": 0.111,
      "step": 710
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.7512484192848206,
      "learning_rate": 9.94167472437353e-06,
      "loss": 0.1233,
      "step": 720
    },
    {
      "epoch": 0.146,
      "grad_norm": 1.176719069480896,
      "learning_rate": 9.936238213151491e-06,
      "loss": 0.1077,
      "step": 730
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.5437051057815552,
      "learning_rate": 9.930561116642936e-06,
      "loss": 0.1159,
      "step": 740
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.4601088762283325,
      "learning_rate": 9.92464371154154e-06,
      "loss": 0.1121,
      "step": 750
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.6581608057022095,
      "learning_rate": 9.918486286253279e-06,
      "loss": 0.1155,
      "step": 760
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.7549941539764404,
      "learning_rate": 9.912089140882377e-06,
      "loss": 0.1138,
      "step": 770
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.8127408623695374,
      "learning_rate": 9.90545258721667e-06,
      "loss": 0.113,
      "step": 780
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.6606084108352661,
      "learning_rate": 9.898576948712427e-06,
      "loss": 0.1171,
      "step": 790
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.90671306848526,
      "learning_rate": 9.891462560478562e-06,
      "loss": 0.1096,
      "step": 800
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.6799761652946472,
      "learning_rate": 9.884109769260326e-06,
      "loss": 0.1175,
      "step": 810
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.6381652355194092,
      "learning_rate": 9.876518933422385e-06,
      "loss": 0.1148,
      "step": 820
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.5381753444671631,
      "learning_rate": 9.868690422931372e-06,
      "loss": 0.121,
      "step": 830
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.7047041654586792,
      "learning_rate": 9.860624619337844e-06,
      "loss": 0.1128,
      "step": 840
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.9245678186416626,
      "learning_rate": 9.852321915757688e-06,
      "loss": 0.1172,
      "step": 850
    },
    {
      "epoch": 0.172,
      "grad_norm": 1.1964024305343628,
      "learning_rate": 9.843782716852963e-06,
      "loss": 0.1156,
      "step": 860
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.6748765110969543,
      "learning_rate": 9.835007438812177e-06,
      "loss": 0.116,
      "step": 870
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.9408971071243286,
      "learning_rate": 9.825996509330001e-06,
      "loss": 0.1134,
      "step": 880
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.5978791117668152,
      "learning_rate": 9.816750367586424e-06,
      "loss": 0.1148,
      "step": 890
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.8566645383834839,
      "learning_rate": 9.807269464225355e-06,
      "loss": 0.1117,
      "step": 900
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.7490561604499817,
      "learning_rate": 9.797554261332644e-06,
      "loss": 0.1116,
      "step": 910
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.5367133617401123,
      "learning_rate": 9.787605232413575e-06,
      "loss": 0.1132,
      "step": 920
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.7432158589363098,
      "learning_rate": 9.777422862369782e-06,
      "loss": 0.1251,
      "step": 930
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.7126122117042542,
      "learning_rate": 9.767007647475618e-06,
      "loss": 0.1066,
      "step": 940
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2850806713104248,
      "learning_rate": 9.756360095353957e-06,
      "loss": 0.1122,
      "step": 950
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.4591011106967926,
      "learning_rate": 9.745480724951473e-06,
      "loss": 0.1101,
      "step": 960
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.6034103035926819,
      "learning_rate": 9.73437006651333e-06,
      "loss": 0.1145,
      "step": 970
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.6527671813964844,
      "learning_rate": 9.723028661557345e-06,
      "loss": 0.1074,
      "step": 980
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.49524030089378357,
      "learning_rate": 9.711457062847596e-06,
      "loss": 0.1101,
      "step": 990
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7419334053993225,
      "learning_rate": 9.699655834367479e-06,
      "loss": 0.1133,
      "step": 1000
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.508783221244812,
      "learning_rate": 9.687625551292219e-06,
      "loss": 0.1156,
      "step": 1010
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.6890400648117065,
      "learning_rate": 9.675366799960842e-06,
      "loss": 0.1095,
      "step": 1020
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.574763834476471,
      "learning_rate": 9.662880177847595e-06,
      "loss": 0.114,
      "step": 1030
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.5926764607429504,
      "learning_rate": 9.650166293532822e-06,
      "loss": 0.1087,
      "step": 1040
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6142486333847046,
      "learning_rate": 9.637225766673309e-06,
      "loss": 0.1061,
      "step": 1050
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.5107919573783875,
      "learning_rate": 9.624059227972077e-06,
      "loss": 0.1154,
      "step": 1060
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.7692158818244934,
      "learning_rate": 9.610667319147648e-06,
      "loss": 0.1128,
      "step": 1070
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.5525968074798584,
      "learning_rate": 9.597050692902765e-06,
      "loss": 0.1096,
      "step": 1080
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.6440755724906921,
      "learning_rate": 9.583210012892582e-06,
      "loss": 0.1157,
      "step": 1090
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5787026882171631,
      "learning_rate": 9.569145953692316e-06,
      "loss": 0.1042,
      "step": 1100
    },
    {
      "epoch": 0.222,
      "grad_norm": 0.8566828966140747,
      "learning_rate": 9.554859200764371e-06,
      "loss": 0.109,
      "step": 1110
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.5807632207870483,
      "learning_rate": 9.540350450424927e-06,
      "loss": 0.1094,
      "step": 1120
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.5819889307022095,
      "learning_rate": 9.525620409810009e-06,
      "loss": 0.1065,
      "step": 1130
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.46392130851745605,
      "learning_rate": 9.510669796841014e-06,
      "loss": 0.1058,
      "step": 1140
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.063761830329895,
      "learning_rate": 9.495499340189729e-06,
      "loss": 0.1096,
      "step": 1150
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.7276637554168701,
      "learning_rate": 9.480109779242805e-06,
      "loss": 0.118,
      "step": 1160
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.5671316385269165,
      "learning_rate": 9.464501864065735e-06,
      "loss": 0.1077,
      "step": 1170
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.7350747585296631,
      "learning_rate": 9.448676355366282e-06,
      "loss": 0.1105,
      "step": 1180
    },
    {
      "epoch": 0.238,
      "grad_norm": 0.5442182421684265,
      "learning_rate": 9.432634024457414e-06,
      "loss": 0.1058,
      "step": 1190
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9185284376144409,
      "learning_rate": 9.41637565321971e-06,
      "loss": 0.1026,
      "step": 1200
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.7028173208236694,
      "learning_rate": 9.399902034063244e-06,
      "loss": 0.108,
      "step": 1210
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.5107137560844421,
      "learning_rate": 9.383213969888972e-06,
      "loss": 0.1148,
      "step": 1220
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.5467950701713562,
      "learning_rate": 9.366312274049602e-06,
      "loss": 0.1007,
      "step": 1230
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.8054739832878113,
      "learning_rate": 9.349197770309942e-06,
      "loss": 0.1057,
      "step": 1240
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5686827301979065,
      "learning_rate": 9.33187129280676e-06,
      "loss": 0.1071,
      "step": 1250
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.6124984622001648,
      "learning_rate": 9.314333686008125e-06,
      "loss": 0.1095,
      "step": 1260
    },
    {
      "epoch": 0.254,
      "grad_norm": 1.458778977394104,
      "learning_rate": 9.296585804672253e-06,
      "loss": 0.1072,
      "step": 1270
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.6592323780059814,
      "learning_rate": 9.278628513805838e-06,
      "loss": 0.1009,
      "step": 1280
    },
    {
      "epoch": 0.258,
      "grad_norm": 0.5499700903892517,
      "learning_rate": 9.260462688621906e-06,
      "loss": 0.109,
      "step": 1290
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5235973596572876,
      "learning_rate": 9.242089214497146e-06,
      "loss": 0.1044,
      "step": 1300
    },
    {
      "epoch": 0.262,
      "grad_norm": 0.5331101417541504,
      "learning_rate": 9.223508986928766e-06,
      "loss": 0.103,
      "step": 1310
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.4890216290950775,
      "learning_rate": 9.204722911490847e-06,
      "loss": 0.1059,
      "step": 1320
    },
    {
      "epoch": 0.266,
      "grad_norm": 0.7136434316635132,
      "learning_rate": 9.1857319037902e-06,
      "loss": 0.1063,
      "step": 1330
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.5715893507003784,
      "learning_rate": 9.16653688942175e-06,
      "loss": 0.1147,
      "step": 1340
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.6043694615364075,
      "learning_rate": 9.147138803923417e-06,
      "loss": 0.11,
      "step": 1350
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.720506489276886,
      "learning_rate": 9.12753859273052e-06,
      "loss": 0.1042,
      "step": 1360
    },
    {
      "epoch": 0.274,
      "grad_norm": 0.8681615591049194,
      "learning_rate": 9.107737211129702e-06,
      "loss": 0.1041,
      "step": 1370
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.49558717012405396,
      "learning_rate": 9.087735624212365e-06,
      "loss": 0.1109,
      "step": 1380
    },
    {
      "epoch": 0.278,
      "grad_norm": 0.5701916217803955,
      "learning_rate": 9.06753480682764e-06,
      "loss": 0.1118,
      "step": 1390
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.503829836845398,
      "learning_rate": 9.047135743534866e-06,
      "loss": 0.1086,
      "step": 1400
    },
    {
      "epoch": 0.282,
      "grad_norm": 0.3906455636024475,
      "learning_rate": 9.026539428555609e-06,
      "loss": 0.1082,
      "step": 1410
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.6386545300483704,
      "learning_rate": 9.005746865725206e-06,
      "loss": 0.1061,
      "step": 1420
    },
    {
      "epoch": 0.286,
      "grad_norm": 0.5736138224601746,
      "learning_rate": 8.984759068443832e-06,
      "loss": 0.1117,
      "step": 1430
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.8654852509498596,
      "learning_rate": 8.963577059627117e-06,
      "loss": 0.1075,
      "step": 1440
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6457758545875549,
      "learning_rate": 8.942201871656292e-06,
      "loss": 0.1098,
      "step": 1450
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.4607081115245819,
      "learning_rate": 8.920634546327857e-06,
      "loss": 0.1033,
      "step": 1460
    },
    {
      "epoch": 0.294,
      "grad_norm": 0.4616289734840393,
      "learning_rate": 8.898876134802827e-06,
      "loss": 0.1073,
      "step": 1470
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.5038948655128479,
      "learning_rate": 8.87692769755548e-06,
      "loss": 0.1086,
      "step": 1480
    },
    {
      "epoch": 0.298,
      "grad_norm": 0.5729373097419739,
      "learning_rate": 8.854790304321682e-06,
      "loss": 0.1061,
      "step": 1490
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.48975491523742676,
      "learning_rate": 8.83246503404675e-06,
      "loss": 0.1099,
      "step": 1500
    },
    {
      "epoch": 0.302,
      "grad_norm": 0.5203743577003479,
      "learning_rate": 8.80995297483286e-06,
      "loss": 0.1075,
      "step": 1510
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.4938233196735382,
      "learning_rate": 8.78725522388602e-06,
      "loss": 0.105,
      "step": 1520
    },
    {
      "epoch": 0.306,
      "grad_norm": 0.5147393941879272,
      "learning_rate": 8.764372887462587e-06,
      "loss": 0.109,
      "step": 1530
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.6048959493637085,
      "learning_rate": 8.741307080815357e-06,
      "loss": 0.1051,
      "step": 1540
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4202023446559906,
      "learning_rate": 8.718058928139205e-06,
      "loss": 0.1071,
      "step": 1550
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.7164034247398376,
      "learning_rate": 8.694629562516295e-06,
      "loss": 0.1111,
      "step": 1560
    },
    {
      "epoch": 0.314,
      "grad_norm": 0.5305653214454651,
      "learning_rate": 8.671020125860851e-06,
      "loss": 0.0985,
      "step": 1570
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.48124703764915466,
      "learning_rate": 8.647231768863513e-06,
      "loss": 0.1051,
      "step": 1580
    },
    {
      "epoch": 0.318,
      "grad_norm": 0.586521327495575,
      "learning_rate": 8.623265650935233e-06,
      "loss": 0.1056,
      "step": 1590
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.7912370562553406,
      "learning_rate": 8.599122940150795e-06,
      "loss": 0.1042,
      "step": 1600
    },
    {
      "epoch": 0.322,
      "grad_norm": 0.5484232306480408,
      "learning_rate": 8.574804813191859e-06,
      "loss": 0.1008,
      "step": 1610
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.4200885593891144,
      "learning_rate": 8.550312455289624e-06,
      "loss": 0.1058,
      "step": 1620
    },
    {
      "epoch": 0.326,
      "grad_norm": 0.5569013357162476,
      "learning_rate": 8.525647060167063e-06,
      "loss": 0.1087,
      "step": 1630
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.42352986335754395,
      "learning_rate": 8.500809829980734e-06,
      "loss": 0.1003,
      "step": 1640
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.5038197636604309,
      "learning_rate": 8.4758019752622e-06,
      "loss": 0.098,
      "step": 1650
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.5340690612792969,
      "learning_rate": 8.450624714859016e-06,
      "loss": 0.107,
      "step": 1660
    },
    {
      "epoch": 0.334,
      "grad_norm": 0.5682323575019836,
      "learning_rate": 8.425279275875336e-06,
      "loss": 0.1034,
      "step": 1670
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.5063655376434326,
      "learning_rate": 8.399766893612096e-06,
      "loss": 0.1005,
      "step": 1680
    },
    {
      "epoch": 0.338,
      "grad_norm": 0.6555968523025513,
      "learning_rate": 8.374088811506819e-06,
      "loss": 0.1074,
      "step": 1690
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4122048318386078,
      "learning_rate": 8.348246281072998e-06,
      "loss": 0.1035,
      "step": 1700
    },
    {
      "epoch": 0.342,
      "grad_norm": 0.6033665537834167,
      "learning_rate": 8.32224056183911e-06,
      "loss": 0.1004,
      "step": 1710
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.6186772584915161,
      "learning_rate": 8.296072921287217e-06,
      "loss": 0.1059,
      "step": 1720
    },
    {
      "epoch": 0.346,
      "grad_norm": 0.5782063603401184,
      "learning_rate": 8.269744634791207e-06,
      "loss": 0.1026,
      "step": 1730
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.5529699921607971,
      "learning_rate": 8.243256985554622e-06,
      "loss": 0.1052,
      "step": 1740
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.43343502283096313,
      "learning_rate": 8.21661126454811e-06,
      "loss": 0.105,
      "step": 1750
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.5857189297676086,
      "learning_rate": 8.189808770446528e-06,
      "loss": 0.1049,
      "step": 1760
    },
    {
      "epoch": 0.354,
      "grad_norm": 0.6525639891624451,
      "learning_rate": 8.162850809565623e-06,
      "loss": 0.0974,
      "step": 1770
    },
    {
      "epoch": 0.356,
      "grad_norm": 2.085735321044922,
      "learning_rate": 8.135738695798377e-06,
      "loss": 0.1067,
      "step": 1780
    },
    {
      "epoch": 0.358,
      "grad_norm": 0.5422115325927734,
      "learning_rate": 8.108473750550965e-06,
      "loss": 0.1086,
      "step": 1790
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4821435213088989,
      "learning_rate": 8.081057302678352e-06,
      "loss": 0.0956,
      "step": 1800
    },
    {
      "epoch": 0.362,
      "grad_norm": 0.6426023244857788,
      "learning_rate": 8.053490688419532e-06,
      "loss": 0.1072,
      "step": 1810
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.5745382905006409,
      "learning_rate": 8.02577525133239e-06,
      "loss": 0.1077,
      "step": 1820
    },
    {
      "epoch": 0.366,
      "grad_norm": 0.5834154486656189,
      "learning_rate": 7.997912342228232e-06,
      "loss": 0.0991,
      "step": 1830
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.4103539288043976,
      "learning_rate": 7.969903319105935e-06,
      "loss": 0.1022,
      "step": 1840
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7636476755142212,
      "learning_rate": 7.941749547085778e-06,
      "loss": 0.1028,
      "step": 1850
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.4373137354850769,
      "learning_rate": 7.913452398342882e-06,
      "loss": 0.099,
      "step": 1860
    },
    {
      "epoch": 0.374,
      "grad_norm": 0.5366427302360535,
      "learning_rate": 7.88501325204036e-06,
      "loss": 0.1059,
      "step": 1870
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.49244028329849243,
      "learning_rate": 7.856433494262078e-06,
      "loss": 0.1055,
      "step": 1880
    },
    {
      "epoch": 0.378,
      "grad_norm": 0.7082386016845703,
      "learning_rate": 7.827714517945116e-06,
      "loss": 0.1042,
      "step": 1890
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6199413537979126,
      "learning_rate": 7.798857722811857e-06,
      "loss": 0.1072,
      "step": 1900
    },
    {
      "epoch": 0.382,
      "grad_norm": 0.5586550831794739,
      "learning_rate": 7.769864515301787e-06,
      "loss": 0.1119,
      "step": 1910
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.5374526977539062,
      "learning_rate": 7.740736308502939e-06,
      "loss": 0.1074,
      "step": 1920
    },
    {
      "epoch": 0.386,
      "grad_norm": 0.6747333407402039,
      "learning_rate": 7.711474522083015e-06,
      "loss": 0.1052,
      "step": 1930
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.8605504631996155,
      "learning_rate": 7.682080582220206e-06,
      "loss": 0.1048,
      "step": 1940
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8798643350601196,
      "learning_rate": 7.652555921533671e-06,
      "loss": 0.0994,
      "step": 1950
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.39183422923088074,
      "learning_rate": 7.622901979013717e-06,
      "loss": 0.1023,
      "step": 1960
    },
    {
      "epoch": 0.394,
      "grad_norm": 0.5172967314720154,
      "learning_rate": 7.5931201999516715e-06,
      "loss": 0.1029,
      "step": 1970
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.4358147084712982,
      "learning_rate": 7.563212035869426e-06,
      "loss": 0.1112,
      "step": 1980
    },
    {
      "epoch": 0.398,
      "grad_norm": 0.9168479442596436,
      "learning_rate": 7.533178944448705e-06,
      "loss": 0.1026,
      "step": 1990
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4735708236694336,
      "learning_rate": 7.503022389460014e-06,
      "loss": 0.1075,
      "step": 2000
    },
    {
      "epoch": 0.402,
      "grad_norm": 2.389822244644165,
      "learning_rate": 7.4727438406912986e-06,
      "loss": 0.1052,
      "step": 2010
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.5340117812156677,
      "learning_rate": 7.44234477387631e-06,
      "loss": 0.1036,
      "step": 2020
    },
    {
      "epoch": 0.406,
      "grad_norm": 0.4848586916923523,
      "learning_rate": 7.411826670622676e-06,
      "loss": 0.0993,
      "step": 2030
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.5334826111793518,
      "learning_rate": 7.381191018339697e-06,
      "loss": 0.1037,
      "step": 2040
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6843870282173157,
      "learning_rate": 7.350439310165842e-06,
      "loss": 0.0961,
      "step": 2050
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.4742767810821533,
      "learning_rate": 7.319573044895986e-06,
      "loss": 0.103,
      "step": 2060
    },
    {
      "epoch": 0.414,
      "grad_norm": 0.7538304924964905,
      "learning_rate": 7.288593726908351e-06,
      "loss": 0.1068,
      "step": 2070
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.4505075216293335,
      "learning_rate": 7.257502866091192e-06,
      "loss": 0.0987,
      "step": 2080
    },
    {
      "epoch": 0.418,
      "grad_norm": 0.4845210015773773,
      "learning_rate": 7.226301977769199e-06,
      "loss": 0.1019,
      "step": 2090
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.7197276949882507,
      "learning_rate": 7.194992582629654e-06,
      "loss": 0.1009,
      "step": 2100
    },
    {
      "epoch": 0.422,
      "grad_norm": 0.8190245032310486,
      "learning_rate": 7.1635762066483035e-06,
      "loss": 0.1066,
      "step": 2110
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.4387541711330414,
      "learning_rate": 7.1320543810149945e-06,
      "loss": 0.0943,
      "step": 2120
    },
    {
      "epoch": 0.426,
      "grad_norm": 0.5509855151176453,
      "learning_rate": 7.100428642059033e-06,
      "loss": 0.104,
      "step": 2130
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.5081747174263,
      "learning_rate": 7.0687005311743195e-06,
      "loss": 0.1016,
      "step": 2140
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.5941420793533325,
      "learning_rate": 7.036871594744218e-06,
      "loss": 0.1094,
      "step": 2150
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.47250431776046753,
      "learning_rate": 7.0049433840661875e-06,
      "loss": 0.1022,
      "step": 2160
    },
    {
      "epoch": 0.434,
      "grad_norm": 0.6311111450195312,
      "learning_rate": 6.97291745527617e-06,
      "loss": 0.1014,
      "step": 2170
    },
    {
      "epoch": 0.436,
      "grad_norm": 1.6460589170455933,
      "learning_rate": 6.940795369272754e-06,
      "loss": 0.1093,
      "step": 2180
    },
    {
      "epoch": 0.438,
      "grad_norm": 0.5381661653518677,
      "learning_rate": 6.908578691641092e-06,
      "loss": 0.0997,
      "step": 2190
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.5386653542518616,
      "learning_rate": 6.876268992576605e-06,
      "loss": 0.0996,
      "step": 2200
    },
    {
      "epoch": 0.442,
      "grad_norm": 0.5878283381462097,
      "learning_rate": 6.843867846808438e-06,
      "loss": 0.1011,
      "step": 2210
    },
    {
      "epoch": 0.444,
      "grad_norm": 1.202819585800171,
      "learning_rate": 6.811376833522729e-06,
      "loss": 0.1015,
      "step": 2220
    },
    {
      "epoch": 0.446,
      "grad_norm": 0.6832873225212097,
      "learning_rate": 6.778797536285625e-06,
      "loss": 0.1007,
      "step": 2230
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.7227466106414795,
      "learning_rate": 6.746131542966112e-06,
      "loss": 0.1036,
      "step": 2240
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.462429016828537,
      "learning_rate": 6.713380445658618e-06,
      "loss": 0.1012,
      "step": 2250
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.5341704487800598,
      "learning_rate": 6.680545840605423e-06,
      "loss": 0.103,
      "step": 2260
    },
    {
      "epoch": 0.454,
      "grad_norm": 0.5896704196929932,
      "learning_rate": 6.647629328118852e-06,
      "loss": 0.1045,
      "step": 2270
    },
    {
      "epoch": 0.456,
      "grad_norm": 4.748597145080566,
      "learning_rate": 6.614632512503289e-06,
      "loss": 0.0989,
      "step": 2280
    },
    {
      "epoch": 0.458,
      "grad_norm": 0.5261045098304749,
      "learning_rate": 6.58155700197697e-06,
      "loss": 0.1017,
      "step": 2290
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5400019884109497,
      "learning_rate": 6.548404408593622e-06,
      "loss": 0.1028,
      "step": 2300
    },
    {
      "epoch": 0.462,
      "grad_norm": 0.5167948603630066,
      "learning_rate": 6.5151763481638705e-06,
      "loss": 0.101,
      "step": 2310
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.5450933575630188,
      "learning_rate": 6.481874440176506e-06,
      "loss": 0.1069,
      "step": 2320
    },
    {
      "epoch": 0.466,
      "grad_norm": 0.8595630526542664,
      "learning_rate": 6.448500307719537e-06,
      "loss": 0.1004,
      "step": 2330
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.47659191489219666,
      "learning_rate": 6.415055577401101e-06,
      "loss": 0.0994,
      "step": 2340
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.49259841442108154,
      "learning_rate": 6.3815418792701686e-06,
      "loss": 0.0985,
      "step": 2350
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.5162966251373291,
      "learning_rate": 6.3479608467371055e-06,
      "loss": 0.1004,
      "step": 2360
    },
    {
      "epoch": 0.474,
      "grad_norm": 0.43801945447921753,
      "learning_rate": 6.314314116494061e-06,
      "loss": 0.1009,
      "step": 2370
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.5107311010360718,
      "learning_rate": 6.280603328435199e-06,
      "loss": 0.0944,
      "step": 2380
    },
    {
      "epoch": 0.478,
      "grad_norm": 0.5112555027008057,
      "learning_rate": 6.24683012557677e-06,
      "loss": 0.1034,
      "step": 2390
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.40618887543678284,
      "learning_rate": 6.212996153977038e-06,
      "loss": 0.1062,
      "step": 2400
    },
    {
      "epoch": 0.482,
      "grad_norm": 0.6284693479537964,
      "learning_rate": 6.179103062656042e-06,
      "loss": 0.0955,
      "step": 2410
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.42741313576698303,
      "learning_rate": 6.145152503515239e-06,
      "loss": 0.1031,
      "step": 2420
    },
    {
      "epoch": 0.486,
      "grad_norm": 0.5457072854042053,
      "learning_rate": 6.111146131256983e-06,
      "loss": 0.1109,
      "step": 2430
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.4269944727420807,
      "learning_rate": 6.077085603303883e-06,
      "loss": 0.098,
      "step": 2440
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5703105926513672,
      "learning_rate": 6.04297257971802e-06,
      "loss": 0.0988,
      "step": 2450
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.3747485280036926,
      "learning_rate": 6.008808723120035e-06,
      "loss": 0.1019,
      "step": 2460
    },
    {
      "epoch": 0.494,
      "grad_norm": 0.4911485016345978,
      "learning_rate": 5.974595698608103e-06,
      "loss": 0.1007,
      "step": 2470
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.5864652991294861,
      "learning_rate": 5.94033517367677e-06,
      "loss": 0.0988,
      "step": 2480
    },
    {
      "epoch": 0.498,
      "grad_norm": 0.5941429138183594,
      "learning_rate": 5.906028818135687e-06,
      "loss": 0.1036,
      "step": 2490
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6516585946083069,
      "learning_rate": 5.871678304028224e-06,
      "loss": 0.0895,
      "step": 2500
    },
    {
      "epoch": 0.502,
      "grad_norm": 0.4426342844963074,
      "learning_rate": 5.837285305549978e-06,
      "loss": 0.0968,
      "step": 2510
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.4874410033226013,
      "learning_rate": 5.802851498967173e-06,
      "loss": 0.1018,
      "step": 2520
    },
    {
      "epoch": 0.506,
      "grad_norm": 0.5964179039001465,
      "learning_rate": 5.768378562534962e-06,
      "loss": 0.1037,
      "step": 2530
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.43735530972480774,
      "learning_rate": 5.733868176415633e-06,
      "loss": 0.0993,
      "step": 2540
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.6398305296897888,
      "learning_rate": 5.6993220225967214e-06,
      "loss": 0.0943,
      "step": 2550
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.5659822225570679,
      "learning_rate": 5.6647417848090225e-06,
      "loss": 0.0991,
      "step": 2560
    },
    {
      "epoch": 0.514,
      "grad_norm": 0.6617985963821411,
      "learning_rate": 5.630129148444543e-06,
      "loss": 0.0957,
      "step": 2570
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.3739423453807831,
      "learning_rate": 5.59548580047435e-06,
      "loss": 0.0951,
      "step": 2580
    },
    {
      "epoch": 0.518,
      "grad_norm": 0.9471829533576965,
      "learning_rate": 5.560813429366345e-06,
      "loss": 0.0965,
      "step": 2590
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7911657094955444,
      "learning_rate": 5.526113725002984e-06,
      "loss": 0.1031,
      "step": 2600
    },
    {
      "epoch": 0.522,
      "grad_norm": 0.6069086194038391,
      "learning_rate": 5.491388378598899e-06,
      "loss": 0.0976,
      "step": 2610
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.47589465975761414,
      "learning_rate": 5.456639082618489e-06,
      "loss": 0.0918,
      "step": 2620
    },
    {
      "epoch": 0.526,
      "grad_norm": 0.4524126946926117,
      "learning_rate": 5.4218675306934145e-06,
      "loss": 0.0979,
      "step": 2630
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.6384430527687073,
      "learning_rate": 5.3870754175400595e-06,
      "loss": 0.0987,
      "step": 2640
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.40055084228515625,
      "learning_rate": 5.352264438876935e-06,
      "loss": 0.097,
      "step": 2650
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.5200124382972717,
      "learning_rate": 5.317436291342031e-06,
      "loss": 0.0973,
      "step": 2660
    },
    {
      "epoch": 0.534,
      "grad_norm": 0.6301565766334534,
      "learning_rate": 5.282592672410124e-06,
      "loss": 0.105,
      "step": 2670
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.48175889253616333,
      "learning_rate": 5.247735280310041e-06,
      "loss": 0.0941,
      "step": 2680
    },
    {
      "epoch": 0.538,
      "grad_norm": 0.3721906542778015,
      "learning_rate": 5.212865813941899e-06,
      "loss": 0.1017,
      "step": 2690
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.38474878668785095,
      "learning_rate": 5.177985972794293e-06,
      "loss": 0.0914,
      "step": 2700
    },
    {
      "epoch": 0.542,
      "grad_norm": 0.5831676721572876,
      "learning_rate": 5.143097456861474e-06,
      "loss": 0.0959,
      "step": 2710
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.5077164173126221,
      "learning_rate": 5.1082019665604895e-06,
      "loss": 0.0977,
      "step": 2720
    },
    {
      "epoch": 0.546,
      "grad_norm": 0.5812519788742065,
      "learning_rate": 5.073301202648304e-06,
      "loss": 0.0874,
      "step": 2730
    },
    {
      "epoch": 0.548,
      "grad_norm": 2.192858934402466,
      "learning_rate": 5.038396866138915e-06,
      "loss": 0.1039,
      "step": 2740
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8018079996109009,
      "learning_rate": 5.003490658220438e-06,
      "loss": 0.0999,
      "step": 2750
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.4818994104862213,
      "learning_rate": 4.968584280172206e-06,
      "loss": 0.0895,
      "step": 2760
    },
    {
      "epoch": 0.554,
      "grad_norm": 0.7325239181518555,
      "learning_rate": 4.933679433281837e-06,
      "loss": 0.1034,
      "step": 2770
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.6191554665565491,
      "learning_rate": 4.898777818762325e-06,
      "loss": 0.0995,
      "step": 2780
    },
    {
      "epoch": 0.558,
      "grad_norm": 0.775681734085083,
      "learning_rate": 4.863881137669123e-06,
      "loss": 0.0952,
      "step": 2790
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4037740230560303,
      "learning_rate": 4.828991090817238e-06,
      "loss": 0.0952,
      "step": 2800
    },
    {
      "epoch": 0.562,
      "grad_norm": 0.43431025743484497,
      "learning_rate": 4.794109378698327e-06,
      "loss": 0.1076,
      "step": 2810
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.4915035665035248,
      "learning_rate": 4.759237701397831e-06,
      "loss": 0.1032,
      "step": 2820
    },
    {
      "epoch": 0.566,
      "grad_norm": 0.4426998794078827,
      "learning_rate": 4.7243777585121034e-06,
      "loss": 0.0979,
      "step": 2830
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.44436115026474,
      "learning_rate": 4.689531249065581e-06,
      "loss": 0.0937,
      "step": 2840
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9476667642593384,
      "learning_rate": 4.654699871427972e-06,
      "loss": 0.1045,
      "step": 2850
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.6479189395904541,
      "learning_rate": 4.619885323231484e-06,
      "loss": 0.1003,
      "step": 2860
    },
    {
      "epoch": 0.574,
      "grad_norm": 4.131319522857666,
      "learning_rate": 4.5850893012880806e-06,
      "loss": 0.1047,
      "step": 2870
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.4022497832775116,
      "learning_rate": 4.5503135015067815e-06,
      "loss": 0.1002,
      "step": 2880
    },
    {
      "epoch": 0.578,
      "grad_norm": 0.6141453981399536,
      "learning_rate": 4.5155596188110055e-06,
      "loss": 0.0973,
      "step": 2890
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.520523190498352,
      "learning_rate": 4.4808293470559645e-06,
      "loss": 0.1045,
      "step": 2900
    },
    {
      "epoch": 0.582,
      "grad_norm": 0.41964802145957947,
      "learning_rate": 4.446124378946108e-06,
      "loss": 0.1089,
      "step": 2910
    },
    {
      "epoch": 0.584,
      "grad_norm": 1.8227766752243042,
      "learning_rate": 4.4114464059526185e-06,
      "loss": 0.0872,
      "step": 2920
    },
    {
      "epoch": 0.586,
      "grad_norm": 0.427450031042099,
      "learning_rate": 4.376797118230978e-06,
      "loss": 0.0966,
      "step": 2930
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.5866195559501648,
      "learning_rate": 4.342178204538588e-06,
      "loss": 0.1014,
      "step": 2940
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.7429317235946655,
      "learning_rate": 4.307591352152459e-06,
      "loss": 0.1011,
      "step": 2950
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.578780472278595,
      "learning_rate": 4.273038246786986e-06,
      "loss": 0.1034,
      "step": 2960
    },
    {
      "epoch": 0.594,
      "grad_norm": 0.45780378580093384,
      "learning_rate": 4.238520572511773e-06,
      "loss": 0.0988,
      "step": 2970
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.4824381470680237,
      "learning_rate": 4.204040011669567e-06,
      "loss": 0.1011,
      "step": 2980
    },
    {
      "epoch": 0.598,
      "grad_norm": 0.6699690222740173,
      "learning_rate": 4.169598244794261e-06,
      "loss": 0.0958,
      "step": 2990
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6796635389328003,
      "learning_rate": 4.135196950528982e-06,
      "loss": 0.1013,
      "step": 3000
    },
    {
      "epoch": 0.602,
      "grad_norm": 0.4701448380947113,
      "learning_rate": 4.100837805544279e-06,
      "loss": 0.0899,
      "step": 3010
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.4026479125022888,
      "learning_rate": 4.066522484456406e-06,
      "loss": 0.0996,
      "step": 3020
    },
    {
      "epoch": 0.606,
      "grad_norm": 0.6788782477378845,
      "learning_rate": 4.032252659745699e-06,
      "loss": 0.0988,
      "step": 3030
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.45941832661628723,
      "learning_rate": 3.9980300016750696e-06,
      "loss": 0.1006,
      "step": 3040
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7753358483314514,
      "learning_rate": 3.963856178208588e-06,
      "loss": 0.0943,
      "step": 3050
    },
    {
      "epoch": 0.612,
      "grad_norm": 0.5039435625076294,
      "learning_rate": 3.9297328549302e-06,
      "loss": 0.0987,
      "step": 3060
    },
    {
      "epoch": 0.614,
      "grad_norm": 0.6418735384941101,
      "learning_rate": 3.895661694962542e-06,
      "loss": 0.0983,
      "step": 3070
    },
    {
      "epoch": 0.616,
      "grad_norm": 1.4340955018997192,
      "learning_rate": 3.86164435888588e-06,
      "loss": 0.1017,
      "step": 3080
    },
    {
      "epoch": 0.618,
      "grad_norm": 0.6281638145446777,
      "learning_rate": 3.827682504657187e-06,
      "loss": 0.1016,
      "step": 3090
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8099610805511475,
      "learning_rate": 3.793777787529325e-06,
      "loss": 0.1028,
      "step": 3100
    },
    {
      "epoch": 0.622,
      "grad_norm": 0.46766799688339233,
      "learning_rate": 3.759931859970374e-06,
      "loss": 0.0992,
      "step": 3110
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.6429394483566284,
      "learning_rate": 3.7261463715830902e-06,
      "loss": 0.0993,
      "step": 3120
    },
    {
      "epoch": 0.626,
      "grad_norm": 0.5106363892555237,
      "learning_rate": 3.6924229690245163e-06,
      "loss": 0.0984,
      "step": 3130
    },
    {
      "epoch": 0.628,
      "grad_norm": 0.3694756031036377,
      "learning_rate": 3.6587632959257168e-06,
      "loss": 0.0919,
      "step": 3140
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.38708898425102234,
      "learning_rate": 3.625168992811671e-06,
      "loss": 0.0992,
      "step": 3150
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.4585551619529724,
      "learning_rate": 3.5916416970213173e-06,
      "loss": 0.0947,
      "step": 3160
    },
    {
      "epoch": 0.634,
      "grad_norm": 0.4839175045490265,
      "learning_rate": 3.5581830426277554e-06,
      "loss": 0.1006,
      "step": 3170
    },
    {
      "epoch": 0.636,
      "grad_norm": 0.6512119770050049,
      "learning_rate": 3.524794660358593e-06,
      "loss": 0.0975,
      "step": 3180
    },
    {
      "epoch": 0.638,
      "grad_norm": 0.6606001853942871,
      "learning_rate": 3.491478177516484e-06,
      "loss": 0.0965,
      "step": 3190
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.5618588924407959,
      "learning_rate": 3.4582352178997937e-06,
      "loss": 0.0972,
      "step": 3200
    },
    {
      "epoch": 0.642,
      "grad_norm": 0.9269917607307434,
      "learning_rate": 3.4250674017234774e-06,
      "loss": 0.1047,
      "step": 3210
    },
    {
      "epoch": 0.644,
      "grad_norm": 0.4144897162914276,
      "learning_rate": 3.3919763455401016e-06,
      "loss": 0.1007,
      "step": 3220
    },
    {
      "epoch": 0.646,
      "grad_norm": 0.541193425655365,
      "learning_rate": 3.358963662161062e-06,
      "loss": 0.0927,
      "step": 3230
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.7467935085296631,
      "learning_rate": 3.3260309605779717e-06,
      "loss": 0.0959,
      "step": 3240
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.5478144288063049,
      "learning_rate": 3.293179845884245e-06,
      "loss": 0.102,
      "step": 3250
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.4977323114871979,
      "learning_rate": 3.260411919196866e-06,
      "loss": 0.0961,
      "step": 3260
    },
    {
      "epoch": 0.654,
      "grad_norm": 0.5541924834251404,
      "learning_rate": 3.227728777578353e-06,
      "loss": 0.1031,
      "step": 3270
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.570885419845581,
      "learning_rate": 3.195132013958918e-06,
      "loss": 0.0984,
      "step": 3280
    },
    {
      "epoch": 0.658,
      "grad_norm": 1.277539849281311,
      "learning_rate": 3.1626232170588343e-06,
      "loss": 0.0952,
      "step": 3290
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4125811457633972,
      "learning_rate": 3.130203971310999e-06,
      "loss": 0.0996,
      "step": 3300
    },
    {
      "epoch": 0.662,
      "grad_norm": 0.5309910774230957,
      "learning_rate": 3.097875856783713e-06,
      "loss": 0.0954,
      "step": 3310
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.553604006767273,
      "learning_rate": 3.0656404491036696e-06,
      "loss": 0.1029,
      "step": 3320
    },
    {
      "epoch": 0.666,
      "grad_norm": 0.740545928478241,
      "learning_rate": 3.033499319379163e-06,
      "loss": 0.0964,
      "step": 3330
    },
    {
      "epoch": 0.668,
      "grad_norm": 0.37706542015075684,
      "learning_rate": 3.001454034123512e-06,
      "loss": 0.104,
      "step": 3340
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.4644536077976227,
      "learning_rate": 2.969506155178711e-06,
      "loss": 0.0963,
      "step": 3350
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.4521162509918213,
      "learning_rate": 2.9376572396393047e-06,
      "loss": 0.0977,
      "step": 3360
    },
    {
      "epoch": 0.674,
      "grad_norm": 0.5465760827064514,
      "learning_rate": 2.905908839776509e-06,
      "loss": 0.097,
      "step": 3370
    },
    {
      "epoch": 0.676,
      "grad_norm": 0.6061195135116577,
      "learning_rate": 2.874262502962537e-06,
      "loss": 0.0937,
      "step": 3380
    },
    {
      "epoch": 0.678,
      "grad_norm": 1.0016474723815918,
      "learning_rate": 2.8427197715952047e-06,
      "loss": 0.1016,
      "step": 3390
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5500060319900513,
      "learning_rate": 2.811282183022736e-06,
      "loss": 0.0996,
      "step": 3400
    },
    {
      "epoch": 0.682,
      "grad_norm": 0.5355184078216553,
      "learning_rate": 2.779951269468847e-06,
      "loss": 0.1028,
      "step": 3410
    },
    {
      "epoch": 0.684,
      "grad_norm": 0.5095385909080505,
      "learning_rate": 2.7487285579580635e-06,
      "loss": 0.095,
      "step": 3420
    },
    {
      "epoch": 0.686,
      "grad_norm": 0.6504735350608826,
      "learning_rate": 2.717615570241294e-06,
      "loss": 0.0971,
      "step": 3430
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.6753683686256409,
      "learning_rate": 2.686613822721666e-06,
      "loss": 0.0995,
      "step": 3440
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.6328535676002502,
      "learning_rate": 2.6557248263806175e-06,
      "loss": 0.1017,
      "step": 3450
    },
    {
      "epoch": 0.692,
      "grad_norm": 0.6124202013015747,
      "learning_rate": 2.6249500867042523e-06,
      "loss": 0.1084,
      "step": 3460
    },
    {
      "epoch": 0.694,
      "grad_norm": 0.510665774345398,
      "learning_rate": 2.5942911036099657e-06,
      "loss": 0.1038,
      "step": 3470
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.5020245909690857,
      "learning_rate": 2.5637493713733376e-06,
      "loss": 0.096,
      "step": 3480
    },
    {
      "epoch": 0.698,
      "grad_norm": 0.6111953854560852,
      "learning_rate": 2.533326378555314e-06,
      "loss": 0.0977,
      "step": 3490
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.49325230717658997,
      "learning_rate": 2.5030236079296443e-06,
      "loss": 0.0959,
      "step": 3500
    },
    {
      "epoch": 0.702,
      "grad_norm": 0.4303823709487915,
      "learning_rate": 2.4728425364106136e-06,
      "loss": 0.0924,
      "step": 3510
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.5887752771377563,
      "learning_rate": 2.442784634981071e-06,
      "loss": 0.0964,
      "step": 3520
    },
    {
      "epoch": 0.706,
      "grad_norm": 0.7012388706207275,
      "learning_rate": 2.412851368620726e-06,
      "loss": 0.0958,
      "step": 3530
    },
    {
      "epoch": 0.708,
      "grad_norm": 0.4575226902961731,
      "learning_rate": 2.3830441962347528e-06,
      "loss": 0.1011,
      "step": 3540
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.7122695446014404,
      "learning_rate": 2.353364570582681e-06,
      "loss": 0.0958,
      "step": 3550
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.5158506631851196,
      "learning_rate": 2.323813938207593e-06,
      "loss": 0.094,
      "step": 3560
    },
    {
      "epoch": 0.714,
      "grad_norm": 1.2165895700454712,
      "learning_rate": 2.294393739365621e-06,
      "loss": 0.1015,
      "step": 3570
    },
    {
      "epoch": 0.716,
      "grad_norm": 0.4191378355026245,
      "learning_rate": 2.265105407955752e-06,
      "loss": 0.1023,
      "step": 3580
    },
    {
      "epoch": 0.718,
      "grad_norm": 0.5682792663574219,
      "learning_rate": 2.235950371449938e-06,
      "loss": 0.0987,
      "step": 3590
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.4805779755115509,
      "learning_rate": 2.2069300508235273e-06,
      "loss": 0.0991,
      "step": 3600
    },
    {
      "epoch": 0.722,
      "grad_norm": 0.5979336500167847,
      "learning_rate": 2.1780458604860056e-06,
      "loss": 0.095,
      "step": 3610
    },
    {
      "epoch": 0.724,
      "grad_norm": 0.48464110493659973,
      "learning_rate": 2.14929920821206e-06,
      "loss": 0.1018,
      "step": 3620
    },
    {
      "epoch": 0.726,
      "grad_norm": 0.8631208539009094,
      "learning_rate": 2.1206914950729673e-06,
      "loss": 0.0965,
      "step": 3630
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.5129687786102295,
      "learning_rate": 2.0922241153683064e-06,
      "loss": 0.1012,
      "step": 3640
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4799841642379761,
      "learning_rate": 2.063898456558002e-06,
      "loss": 0.0954,
      "step": 3650
    },
    {
      "epoch": 0.732,
      "grad_norm": 0.698362410068512,
      "learning_rate": 2.035715899194704e-06,
      "loss": 0.0922,
      "step": 3660
    },
    {
      "epoch": 0.734,
      "grad_norm": 0.5779376029968262,
      "learning_rate": 2.007677816856498e-06,
      "loss": 0.0908,
      "step": 3670
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.37104320526123047,
      "learning_rate": 1.979785576079961e-06,
      "loss": 0.0907,
      "step": 3680
    },
    {
      "epoch": 0.738,
      "grad_norm": 0.9988909363746643,
      "learning_rate": 1.95204053629356e-06,
      "loss": 0.0991,
      "step": 3690
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.6738276481628418,
      "learning_rate": 1.9244440497513895e-06,
      "loss": 0.0883,
      "step": 3700
    },
    {
      "epoch": 0.742,
      "grad_norm": 1.0084031820297241,
      "learning_rate": 1.896997461467272e-06,
      "loss": 0.0963,
      "step": 3710
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.47921252250671387,
      "learning_rate": 1.8697021091491991e-06,
      "loss": 0.0951,
      "step": 3720
    },
    {
      "epoch": 0.746,
      "grad_norm": 0.697775661945343,
      "learning_rate": 1.842559323134136e-06,
      "loss": 0.0946,
      "step": 3730
    },
    {
      "epoch": 0.748,
      "grad_norm": 0.74412602186203,
      "learning_rate": 1.8155704263231777e-06,
      "loss": 0.0889,
      "step": 3740
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5915809273719788,
      "learning_rate": 1.7887367341170781e-06,
      "loss": 0.097,
      "step": 3750
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.8403988480567932,
      "learning_rate": 1.762059554352143e-06,
      "loss": 0.1004,
      "step": 3760
    },
    {
      "epoch": 0.754,
      "grad_norm": 0.763522207736969,
      "learning_rate": 1.7355401872364759e-06,
      "loss": 0.0987,
      "step": 3770
    },
    {
      "epoch": 0.756,
      "grad_norm": 0.5236890316009521,
      "learning_rate": 1.709179925286617e-06,
      "loss": 0.0993,
      "step": 3780
    },
    {
      "epoch": 0.758,
      "grad_norm": 1.012764811515808,
      "learning_rate": 1.6829800532645447e-06,
      "loss": 0.1035,
      "step": 3790
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.4698300361633301,
      "learning_rate": 1.6569418481150596e-06,
      "loss": 0.0944,
      "step": 3800
    },
    {
      "epoch": 0.762,
      "grad_norm": 0.545480489730835,
      "learning_rate": 1.6310665789035468e-06,
      "loss": 0.0961,
      "step": 3810
    },
    {
      "epoch": 0.764,
      "grad_norm": 1.2716432809829712,
      "learning_rate": 1.605355506754121e-06,
      "loss": 0.0964,
      "step": 3820
    },
    {
      "epoch": 0.766,
      "grad_norm": 0.6809777021408081,
      "learning_rate": 1.5798098847881664e-06,
      "loss": 0.0901,
      "step": 3830
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.45785894989967346,
      "learning_rate": 1.554430958063259e-06,
      "loss": 0.0931,
      "step": 3840
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5503713488578796,
      "learning_rate": 1.529219963512481e-06,
      "loss": 0.0963,
      "step": 3850
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.4882219731807709,
      "learning_rate": 1.5041781298841424e-06,
      "loss": 0.105,
      "step": 3860
    },
    {
      "epoch": 0.774,
      "grad_norm": 1.1814230680465698,
      "learning_rate": 1.4793066776818843e-06,
      "loss": 0.098,
      "step": 3870
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.51359623670578,
      "learning_rate": 1.4546068191051988e-06,
      "loss": 0.0985,
      "step": 3880
    },
    {
      "epoch": 0.778,
      "grad_norm": 0.4637075960636139,
      "learning_rate": 1.4300797579903476e-06,
      "loss": 0.0955,
      "step": 3890
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.7598605155944824,
      "learning_rate": 1.4057266897516842e-06,
      "loss": 0.103,
      "step": 3900
    },
    {
      "epoch": 0.782,
      "grad_norm": 0.4668324291706085,
      "learning_rate": 1.3815488013233986e-06,
      "loss": 0.0928,
      "step": 3910
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.46290475130081177,
      "learning_rate": 1.3575472711016634e-06,
      "loss": 0.0921,
      "step": 3920
    },
    {
      "epoch": 0.786,
      "grad_norm": 0.5419790148735046,
      "learning_rate": 1.333723268887201e-06,
      "loss": 0.0909,
      "step": 3930
    },
    {
      "epoch": 0.788,
      "grad_norm": 0.6386012434959412,
      "learning_rate": 1.3100779558282673e-06,
      "loss": 0.0903,
      "step": 3940
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.1641706228256226,
      "learning_rate": 1.2866124843640614e-06,
      "loss": 0.0965,
      "step": 3950
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.8818916082382202,
      "learning_rate": 1.2633279981685608e-06,
      "loss": 0.1001,
      "step": 3960
    },
    {
      "epoch": 0.794,
      "grad_norm": 0.4657428562641144,
      "learning_rate": 1.240225632094773e-06,
      "loss": 0.0949,
      "step": 3970
    },
    {
      "epoch": 0.796,
      "grad_norm": 0.6301387548446655,
      "learning_rate": 1.217306512119425e-06,
      "loss": 0.1017,
      "step": 3980
    },
    {
      "epoch": 0.798,
      "grad_norm": 0.5192793011665344,
      "learning_rate": 1.1945717552880919e-06,
      "loss": 0.0949,
      "step": 3990
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.445238322019577,
      "learning_rate": 1.1720224696607474e-06,
      "loss": 0.0933,
      "step": 4000
    },
    {
      "epoch": 0.802,
      "grad_norm": 0.7800849676132202,
      "learning_rate": 1.1496597542577603e-06,
      "loss": 0.0937,
      "step": 4010
    },
    {
      "epoch": 0.804,
      "grad_norm": 0.5425428748130798,
      "learning_rate": 1.1274846990063314e-06,
      "loss": 0.0927,
      "step": 4020
    },
    {
      "epoch": 0.806,
      "grad_norm": 0.5033382177352905,
      "learning_rate": 1.1054983846873684e-06,
      "loss": 0.0909,
      "step": 4030
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.6877299547195435,
      "learning_rate": 1.0837018828828133e-06,
      "loss": 0.0905,
      "step": 4040
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5039352774620056,
      "learning_rate": 1.0620962559234144e-06,
      "loss": 0.0938,
      "step": 4050
    },
    {
      "epoch": 0.812,
      "grad_norm": 0.48495736718177795,
      "learning_rate": 1.0406825568369478e-06,
      "loss": 0.0947,
      "step": 4060
    },
    {
      "epoch": 0.814,
      "grad_norm": 0.3968163728713989,
      "learning_rate": 1.0194618292968972e-06,
      "loss": 0.0898,
      "step": 4070
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.542452871799469,
      "learning_rate": 9.984351075715848e-07,
      "loss": 0.0916,
      "step": 4080
    },
    {
      "epoch": 0.818,
      "grad_norm": 0.6849305629730225,
      "learning_rate": 9.77603416473763e-07,
      "loss": 0.0968,
      "step": 4090
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.279828667640686,
      "learning_rate": 9.569677713106673e-07,
      "loss": 0.0954,
      "step": 4100
    },
    {
      "epoch": 0.822,
      "grad_norm": 0.5056818723678589,
      "learning_rate": 9.365291778345303e-07,
      "loss": 0.0852,
      "step": 4110
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.4796373248100281,
      "learning_rate": 9.162886321935632e-07,
      "loss": 0.0898,
      "step": 4120
    },
    {
      "epoch": 0.826,
      "grad_norm": 0.7538776397705078,
      "learning_rate": 8.962471208834056e-07,
      "loss": 0.0916,
      "step": 4130
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.499845951795578,
      "learning_rate": 8.764056206990446e-07,
      "loss": 0.0985,
      "step": 4140
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.0565054416656494,
      "learning_rate": 8.567650986872061e-07,
      "loss": 0.0913,
      "step": 4150
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.47425204515457153,
      "learning_rate": 8.373265120992252e-07,
      "loss": 0.1015,
      "step": 4160
    },
    {
      "epoch": 0.834,
      "grad_norm": 0.5785176753997803,
      "learning_rate": 8.180908083443884e-07,
      "loss": 0.0904,
      "step": 4170
    },
    {
      "epoch": 0.836,
      "grad_norm": 0.5388156175613403,
      "learning_rate": 7.990589249437591e-07,
      "loss": 0.0947,
      "step": 4180
    },
    {
      "epoch": 0.838,
      "grad_norm": 0.4170565903186798,
      "learning_rate": 7.802317894844835e-07,
      "loss": 0.0899,
      "step": 4190
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.5308743715286255,
      "learning_rate": 7.61610319574585e-07,
      "loss": 0.0929,
      "step": 4200
    },
    {
      "epoch": 0.842,
      "grad_norm": 0.583870530128479,
      "learning_rate": 7.43195422798233e-07,
      "loss": 0.0865,
      "step": 4210
    },
    {
      "epoch": 0.844,
      "grad_norm": 0.5489416122436523,
      "learning_rate": 7.249879966715174e-07,
      "loss": 0.1026,
      "step": 4220
    },
    {
      "epoch": 0.846,
      "grad_norm": 0.5027689933776855,
      "learning_rate": 7.069889285987025e-07,
      "loss": 0.1007,
      "step": 4230
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.4889448583126068,
      "learning_rate": 6.891990958289724e-07,
      "loss": 0.0875,
      "step": 4240
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.2699036598205566,
      "learning_rate": 6.716193654136788e-07,
      "loss": 0.0973,
      "step": 4250
    },
    {
      "epoch": 0.852,
      "grad_norm": 0.5371798276901245,
      "learning_rate": 6.542505941640803e-07,
      "loss": 0.0943,
      "step": 4260
    },
    {
      "epoch": 0.854,
      "grad_norm": 0.6962149739265442,
      "learning_rate": 6.370936286095842e-07,
      "loss": 0.0909,
      "step": 4270
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.5354880094528198,
      "learning_rate": 6.201493049564883e-07,
      "loss": 0.093,
      "step": 4280
    },
    {
      "epoch": 0.858,
      "grad_norm": 0.627852737903595,
      "learning_rate": 6.034184490472195e-07,
      "loss": 0.1,
      "step": 4290
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5209593176841736,
      "learning_rate": 5.869018763200929e-07,
      "loss": 0.0944,
      "step": 4300
    },
    {
      "epoch": 0.862,
      "grad_norm": 0.5602114200592041,
      "learning_rate": 5.706003917695619e-07,
      "loss": 0.0944,
      "step": 4310
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.4925883114337921,
      "learning_rate": 5.545147899069836e-07,
      "loss": 0.1006,
      "step": 4320
    },
    {
      "epoch": 0.866,
      "grad_norm": 0.5700224041938782,
      "learning_rate": 5.386458547219026e-07,
      "loss": 0.0908,
      "step": 4330
    },
    {
      "epoch": 0.868,
      "grad_norm": 1.1226917505264282,
      "learning_rate": 5.229943596438297e-07,
      "loss": 0.098,
      "step": 4340
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.6325000524520874,
      "learning_rate": 5.075610675045567e-07,
      "loss": 0.0918,
      "step": 4350
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.6880750060081482,
      "learning_rate": 4.92346730500966e-07,
      "loss": 0.0891,
      "step": 4360
    },
    {
      "epoch": 0.874,
      "grad_norm": 0.9880423545837402,
      "learning_rate": 4.773520901583801e-07,
      "loss": 0.0932,
      "step": 4370
    },
    {
      "epoch": 0.876,
      "grad_norm": 0.43653684854507446,
      "learning_rate": 4.625778772944156e-07,
      "loss": 0.0919,
      "step": 4380
    },
    {
      "epoch": 0.878,
      "grad_norm": 0.5817267298698425,
      "learning_rate": 4.480248119833641e-07,
      "loss": 0.0976,
      "step": 4390
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5401520729064941,
      "learning_rate": 4.33693603521097e-07,
      "loss": 0.0984,
      "step": 4400
    },
    {
      "epoch": 0.882,
      "grad_norm": 0.5201219320297241,
      "learning_rate": 4.195849503904975e-07,
      "loss": 0.0909,
      "step": 4410
    },
    {
      "epoch": 0.884,
      "grad_norm": 0.46396180987358093,
      "learning_rate": 4.056995402274122e-07,
      "loss": 0.092,
      "step": 4420
    },
    {
      "epoch": 0.886,
      "grad_norm": 0.8054051995277405,
      "learning_rate": 3.920380497871473e-07,
      "loss": 0.0924,
      "step": 4430
    },
    {
      "epoch": 0.888,
      "grad_norm": 1.6495290994644165,
      "learning_rate": 3.7860114491147017e-07,
      "loss": 0.0856,
      "step": 4440
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5194240212440491,
      "learning_rate": 3.6538948049616886e-07,
      "loss": 0.1016,
      "step": 4450
    },
    {
      "epoch": 0.892,
      "grad_norm": 0.9510396718978882,
      "learning_rate": 3.524037004591274e-07,
      "loss": 0.092,
      "step": 4460
    },
    {
      "epoch": 0.894,
      "grad_norm": 0.5282657146453857,
      "learning_rate": 3.396444377089453e-07,
      "loss": 0.1007,
      "step": 4470
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.5371118187904358,
      "learning_rate": 3.271123141140886e-07,
      "loss": 0.0928,
      "step": 4480
    },
    {
      "epoch": 0.898,
      "grad_norm": 0.4065883457660675,
      "learning_rate": 3.148079404725801e-07,
      "loss": 0.0895,
      "step": 4490
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5543450117111206,
      "learning_rate": 3.027319164822329e-07,
      "loss": 0.0912,
      "step": 4500
    },
    {
      "epoch": 0.902,
      "grad_norm": 0.5739998817443848,
      "learning_rate": 2.908848307114198e-07,
      "loss": 0.0938,
      "step": 4510
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.48197928071022034,
      "learning_rate": 2.792672605703867e-07,
      "loss": 0.0912,
      "step": 4520
    },
    {
      "epoch": 0.906,
      "grad_norm": 0.9697645902633667,
      "learning_rate": 2.6787977228311336e-07,
      "loss": 0.0929,
      "step": 4530
    },
    {
      "epoch": 0.908,
      "grad_norm": 0.7662408351898193,
      "learning_rate": 2.5672292085971276e-07,
      "loss": 0.0976,
      "step": 4540
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5820423364639282,
      "learning_rate": 2.457972500693834e-07,
      "loss": 0.0952,
      "step": 4550
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.6280630826950073,
      "learning_rate": 2.351032924139063e-07,
      "loss": 0.0895,
      "step": 4560
    },
    {
      "epoch": 0.914,
      "grad_norm": 0.8524745106697083,
      "learning_rate": 2.2464156910168954e-07,
      "loss": 0.0931,
      "step": 4570
    },
    {
      "epoch": 0.916,
      "grad_norm": 0.8388736844062805,
      "learning_rate": 2.1441259002236924e-07,
      "loss": 0.0984,
      "step": 4580
    },
    {
      "epoch": 0.918,
      "grad_norm": 0.7623071074485779,
      "learning_rate": 2.0441685372195487e-07,
      "loss": 0.0914,
      "step": 4590
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7167821526527405,
      "learning_rate": 1.9465484737853092e-07,
      "loss": 0.092,
      "step": 4600
    },
    {
      "epoch": 0.922,
      "grad_norm": 0.9387440085411072,
      "learning_rate": 1.8512704677851489e-07,
      "loss": 0.0918,
      "step": 4610
    },
    {
      "epoch": 0.924,
      "grad_norm": 1.0186631679534912,
      "learning_rate": 1.758339162934658e-07,
      "loss": 0.0914,
      "step": 4620
    },
    {
      "epoch": 0.926,
      "grad_norm": 0.7060619592666626,
      "learning_rate": 1.6677590885745388e-07,
      "loss": 0.095,
      "step": 4630
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.7647775411605835,
      "learning_rate": 1.5795346594498162e-07,
      "loss": 0.0942,
      "step": 4640
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5398228168487549,
      "learning_rate": 1.4936701754947104e-07,
      "loss": 0.0965,
      "step": 4650
    },
    {
      "epoch": 0.932,
      "grad_norm": 0.6493486762046814,
      "learning_rate": 1.4101698216230254e-07,
      "loss": 0.0975,
      "step": 4660
    },
    {
      "epoch": 0.934,
      "grad_norm": 0.5592340230941772,
      "learning_rate": 1.3290376675242022e-07,
      "loss": 0.091,
      "step": 4670
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.6127602458000183,
      "learning_rate": 1.2502776674649776e-07,
      "loss": 0.0955,
      "step": 4680
    },
    {
      "epoch": 0.938,
      "grad_norm": 0.6314841508865356,
      "learning_rate": 1.1738936600966366e-07,
      "loss": 0.0947,
      "step": 4690
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.6018589735031128,
      "learning_rate": 1.0998893682679479e-07,
      "loss": 0.1013,
      "step": 4700
    },
    {
      "epoch": 0.942,
      "grad_norm": 0.5869537591934204,
      "learning_rate": 1.0282683988436792e-07,
      "loss": 0.0952,
      "step": 4710
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.5971767902374268,
      "learning_rate": 9.590342425288446e-08,
      "loss": 0.0855,
      "step": 4720
    },
    {
      "epoch": 0.946,
      "grad_norm": 0.6858376860618591,
      "learning_rate": 8.921902736985399e-08,
      "loss": 0.0964,
      "step": 4730
    },
    {
      "epoch": 0.948,
      "grad_norm": 0.6425265669822693,
      "learning_rate": 8.277397502335194e-08,
      "loss": 0.0873,
      "step": 4740
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6217800378799438,
      "learning_rate": 7.656858133613498e-08,
      "loss": 0.0932,
      "step": 4750
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.7790467143058777,
      "learning_rate": 7.060314875033836e-08,
      "loss": 0.0884,
      "step": 4760
    },
    {
      "epoch": 0.954,
      "grad_norm": 0.7824919819831848,
      "learning_rate": 6.487796801272983e-08,
      "loss": 0.0994,
      "step": 4770
    },
    {
      "epoch": 0.956,
      "grad_norm": 1.0274417400360107,
      "learning_rate": 5.939331816054161e-08,
      "loss": 0.096,
      "step": 4780
    },
    {
      "epoch": 0.958,
      "grad_norm": 1.30277681350708,
      "learning_rate": 5.414946650786957e-08,
      "loss": 0.0875,
      "step": 4790
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.7545578479766846,
      "learning_rate": 4.914666863264528e-08,
      "loss": 0.09,
      "step": 4800
    },
    {
      "epoch": 0.962,
      "grad_norm": 0.5341658592224121,
      "learning_rate": 4.438516836417994e-08,
      "loss": 0.0904,
      "step": 4810
    },
    {
      "epoch": 0.964,
      "grad_norm": 0.954708993434906,
      "learning_rate": 3.986519777127884e-08,
      "loss": 0.0878,
      "step": 4820
    },
    {
      "epoch": 0.966,
      "grad_norm": 3.0081846714019775,
      "learning_rate": 3.558697715093207e-08,
      "loss": 0.1018,
      "step": 4830
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.6550202965736389,
      "learning_rate": 3.1550715017575895e-08,
      "loss": 0.0933,
      "step": 4840
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.0143718719482422,
      "learning_rate": 2.7756608092933678e-08,
      "loss": 0.1012,
      "step": 4850
    },
    {
      "epoch": 1.0002,
      "grad_norm": 0.7272701859474182,
      "learning_rate": 2.4204841296424086e-08,
      "loss": 0.1023,
      "step": 4860
    },
    {
      "epoch": 1.0022,
      "grad_norm": 0.5501855611801147,
      "learning_rate": 2.0895587736149414e-08,
      "loss": 0.0984,
      "step": 4870
    },
    {
      "epoch": 1.0042,
      "grad_norm": 0.729179322719574,
      "learning_rate": 1.7829008700460116e-08,
      "loss": 0.0895,
      "step": 4880
    },
    {
      "epoch": 1.0062,
      "grad_norm": 0.6516703963279724,
      "learning_rate": 1.500525365009109e-08,
      "loss": 0.0921,
      "step": 4890
    },
    {
      "epoch": 1.0082,
      "grad_norm": 0.6412553191184998,
      "learning_rate": 1.2424460210881394e-08,
      "loss": 0.1001,
      "step": 4900
    },
    {
      "epoch": 1.0102,
      "grad_norm": 0.8296138644218445,
      "learning_rate": 1.008675416706073e-08,
      "loss": 0.0834,
      "step": 4910
    },
    {
      "epoch": 1.0122,
      "grad_norm": 0.6063240766525269,
      "learning_rate": 7.992249455124889e-09,
      "loss": 0.0831,
      "step": 4920
    },
    {
      "epoch": 1.0142,
      "grad_norm": 0.6920802593231201,
      "learning_rate": 6.141048158277429e-09,
      "loss": 0.0919,
      "step": 4930
    },
    {
      "epoch": 1.0162,
      "grad_norm": 0.6563531160354614,
      "learning_rate": 4.533240501459202e-09,
      "loss": 0.0952,
      "step": 4940
    },
    {
      "epoch": 1.0182,
      "grad_norm": 0.6383795142173767,
      "learning_rate": 3.1689048469457638e-09,
      "loss": 0.091,
      "step": 4950
    },
    {
      "epoch": 1.0202,
      "grad_norm": 0.6429073214530945,
      "learning_rate": 2.0481076905332074e-09,
      "loss": 0.0974,
      "step": 4960
    },
    {
      "epoch": 1.0222,
      "grad_norm": 0.4217958450317383,
      "learning_rate": 1.170903658293532e-09,
      "loss": 0.091,
      "step": 4970
    },
    {
      "epoch": 1.0242,
      "grad_norm": 0.648520827293396,
      "learning_rate": 5.373355039128836e-10,
      "loss": 0.0889,
      "step": 4980
    },
    {
      "epoch": 1.0262,
      "grad_norm": 0.6270639896392822,
      "learning_rate": 1.4743410661044454e-10,
      "loss": 0.0999,
      "step": 4990
    },
    {
      "epoch": 1.0282,
      "grad_norm": 0.7285539507865906,
      "learning_rate": 1.2184696296380083e-12,
      "loss": 0.1012,
      "step": 5000
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4735039884017795e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}