{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 200,
  "global_step": 2666,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0007501875468867217,
      "grad_norm": 0.13885293900966644,
      "learning_rate": 3.7453183520599253e-07,
      "loss": 0.1628,
      "step": 1
    },
    {
      "epoch": 0.0015003750937734434,
      "grad_norm": 0.18549223244190216,
      "learning_rate": 7.490636704119851e-07,
      "loss": 0.2412,
      "step": 2
    },
    {
      "epoch": 0.002250562640660165,
      "grad_norm": 0.1155385822057724,
      "learning_rate": 1.1235955056179775e-06,
      "loss": 0.1319,
      "step": 3
    },
    {
      "epoch": 0.003000750187546887,
      "grad_norm": 0.15342453122138977,
      "learning_rate": 1.4981273408239701e-06,
      "loss": 0.2292,
      "step": 4
    },
    {
      "epoch": 0.0037509377344336083,
      "grad_norm": 0.12306933850049973,
      "learning_rate": 1.8726591760299627e-06,
      "loss": 0.1508,
      "step": 5
    },
    {
      "epoch": 0.00450112528132033,
      "grad_norm": 0.191707581281662,
      "learning_rate": 2.247191011235955e-06,
      "loss": 0.2169,
      "step": 6
    },
    {
      "epoch": 0.005251312828207052,
      "grad_norm": 0.13260212540626526,
      "learning_rate": 2.621722846441948e-06,
      "loss": 0.1464,
      "step": 7
    },
    {
      "epoch": 0.006001500375093774,
      "grad_norm": 0.10345903784036636,
      "learning_rate": 2.9962546816479402e-06,
      "loss": 0.1349,
      "step": 8
    },
    {
      "epoch": 0.006751687921980495,
      "grad_norm": 0.10904919356107712,
      "learning_rate": 3.3707865168539327e-06,
      "loss": 0.1499,
      "step": 9
    },
    {
      "epoch": 0.007501875468867217,
      "grad_norm": 0.12517283856868744,
      "learning_rate": 3.7453183520599255e-06,
      "loss": 0.18,
      "step": 10
    },
    {
      "epoch": 0.008252063015753939,
      "grad_norm": 0.12791502475738525,
      "learning_rate": 4.1198501872659175e-06,
      "loss": 0.101,
      "step": 11
    },
    {
      "epoch": 0.00900225056264066,
      "grad_norm": 0.17475782334804535,
      "learning_rate": 4.49438202247191e-06,
      "loss": 0.2105,
      "step": 12
    },
    {
      "epoch": 0.009752438109527382,
      "grad_norm": 0.17885951697826385,
      "learning_rate": 4.868913857677903e-06,
      "loss": 0.2074,
      "step": 13
    },
    {
      "epoch": 0.010502625656414103,
      "grad_norm": 0.3866336941719055,
      "learning_rate": 5.243445692883896e-06,
      "loss": 0.3252,
      "step": 14
    },
    {
      "epoch": 0.011252813203300824,
      "grad_norm": 0.13714569807052612,
      "learning_rate": 5.617977528089888e-06,
      "loss": 0.1515,
      "step": 15
    },
    {
      "epoch": 0.012003000750187547,
      "grad_norm": 0.2914994955062866,
      "learning_rate": 5.9925093632958805e-06,
      "loss": 0.1959,
      "step": 16
    },
    {
      "epoch": 0.012753188297074268,
      "grad_norm": 0.09515587240457535,
      "learning_rate": 6.367041198501873e-06,
      "loss": 0.0907,
      "step": 17
    },
    {
      "epoch": 0.01350337584396099,
      "grad_norm": 0.11435846239328384,
      "learning_rate": 6.741573033707865e-06,
      "loss": 0.1067,
      "step": 18
    },
    {
      "epoch": 0.014253563390847712,
      "grad_norm": 0.11161951720714569,
      "learning_rate": 7.116104868913858e-06,
      "loss": 0.1085,
      "step": 19
    },
    {
      "epoch": 0.015003750937734433,
      "grad_norm": 0.2902528941631317,
      "learning_rate": 7.490636704119851e-06,
      "loss": 0.3206,
      "step": 20
    },
    {
      "epoch": 0.015753938484621154,
      "grad_norm": 0.1824679672718048,
      "learning_rate": 7.865168539325843e-06,
      "loss": 0.1762,
      "step": 21
    },
    {
      "epoch": 0.016504126031507877,
      "grad_norm": 0.19523698091506958,
      "learning_rate": 8.239700374531835e-06,
      "loss": 0.2151,
      "step": 22
    },
    {
      "epoch": 0.0172543135783946,
      "grad_norm": 0.18187110126018524,
      "learning_rate": 8.614232209737828e-06,
      "loss": 0.1724,
      "step": 23
    },
    {
      "epoch": 0.01800450112528132,
      "grad_norm": 0.2814335823059082,
      "learning_rate": 8.98876404494382e-06,
      "loss": 0.1468,
      "step": 24
    },
    {
      "epoch": 0.018754688672168042,
      "grad_norm": 0.19284458458423615,
      "learning_rate": 9.363295880149813e-06,
      "loss": 0.1513,
      "step": 25
    },
    {
      "epoch": 0.019504876219054765,
      "grad_norm": 0.1800457239151001,
      "learning_rate": 9.737827715355806e-06,
      "loss": 0.1771,
      "step": 26
    },
    {
      "epoch": 0.020255063765941484,
      "grad_norm": 0.14044098556041718,
      "learning_rate": 1.0112359550561798e-05,
      "loss": 0.1252,
      "step": 27
    },
    {
      "epoch": 0.021005251312828207,
      "grad_norm": 0.30116593837738037,
      "learning_rate": 1.0486891385767791e-05,
      "loss": 0.2511,
      "step": 28
    },
    {
      "epoch": 0.02175543885971493,
      "grad_norm": 0.144786074757576,
      "learning_rate": 1.0861423220973783e-05,
      "loss": 0.1473,
      "step": 29
    },
    {
      "epoch": 0.02250562640660165,
      "grad_norm": 0.1813868284225464,
      "learning_rate": 1.1235955056179776e-05,
      "loss": 0.1257,
      "step": 30
    },
    {
      "epoch": 0.023255813953488372,
      "grad_norm": 0.14789240062236786,
      "learning_rate": 1.161048689138577e-05,
      "loss": 0.1564,
      "step": 31
    },
    {
      "epoch": 0.024006001500375095,
      "grad_norm": 0.11978735774755478,
      "learning_rate": 1.1985018726591761e-05,
      "loss": 0.0879,
      "step": 32
    },
    {
      "epoch": 0.024756189047261814,
      "grad_norm": 0.24735777080059052,
      "learning_rate": 1.2359550561797752e-05,
      "loss": 0.1878,
      "step": 33
    },
    {
      "epoch": 0.025506376594148537,
      "grad_norm": 0.16269366443157196,
      "learning_rate": 1.2734082397003746e-05,
      "loss": 0.1552,
      "step": 34
    },
    {
      "epoch": 0.02625656414103526,
      "grad_norm": 0.25208044052124023,
      "learning_rate": 1.3108614232209737e-05,
      "loss": 0.227,
      "step": 35
    },
    {
      "epoch": 0.02700675168792198,
      "grad_norm": 0.16651014983654022,
      "learning_rate": 1.348314606741573e-05,
      "loss": 0.1223,
      "step": 36
    },
    {
      "epoch": 0.0277569392348087,
      "grad_norm": 0.21705418825149536,
      "learning_rate": 1.3857677902621724e-05,
      "loss": 0.1634,
      "step": 37
    },
    {
      "epoch": 0.028507126781695424,
      "grad_norm": 0.1602638065814972,
      "learning_rate": 1.4232209737827715e-05,
      "loss": 0.1174,
      "step": 38
    },
    {
      "epoch": 0.029257314328582147,
      "grad_norm": 0.37907129526138306,
      "learning_rate": 1.4606741573033709e-05,
      "loss": 0.2483,
      "step": 39
    },
    {
      "epoch": 0.030007501875468866,
      "grad_norm": 0.13058216869831085,
      "learning_rate": 1.4981273408239702e-05,
      "loss": 0.117,
      "step": 40
    },
    {
      "epoch": 0.03075768942235559,
      "grad_norm": 0.11338390409946442,
      "learning_rate": 1.5355805243445692e-05,
      "loss": 0.1113,
      "step": 41
    },
    {
      "epoch": 0.03150787696924231,
      "grad_norm": 0.2932577133178711,
      "learning_rate": 1.5730337078651687e-05,
      "loss": 0.1604,
      "step": 42
    },
    {
      "epoch": 0.03225806451612903,
      "grad_norm": 0.17441685497760773,
      "learning_rate": 1.610486891385768e-05,
      "loss": 0.1427,
      "step": 43
    },
    {
      "epoch": 0.033008252063015754,
      "grad_norm": 0.2918965518474579,
      "learning_rate": 1.647940074906367e-05,
      "loss": 0.14,
      "step": 44
    },
    {
      "epoch": 0.03375843960990248,
      "grad_norm": 0.2671569883823395,
      "learning_rate": 1.6853932584269665e-05,
      "loss": 0.1558,
      "step": 45
    },
    {
      "epoch": 0.0345086271567892,
      "grad_norm": 0.2006506770849228,
      "learning_rate": 1.7228464419475657e-05,
      "loss": 0.1355,
      "step": 46
    },
    {
      "epoch": 0.035258814703675916,
      "grad_norm": 0.15463334321975708,
      "learning_rate": 1.760299625468165e-05,
      "loss": 0.1249,
      "step": 47
    },
    {
      "epoch": 0.03600900225056264,
      "grad_norm": 0.394148588180542,
      "learning_rate": 1.797752808988764e-05,
      "loss": 0.2662,
      "step": 48
    },
    {
      "epoch": 0.03675918979744936,
      "grad_norm": 0.32093098759651184,
      "learning_rate": 1.8352059925093635e-05,
      "loss": 0.2,
      "step": 49
    },
    {
      "epoch": 0.037509377344336084,
      "grad_norm": 0.127495676279068,
      "learning_rate": 1.8726591760299626e-05,
      "loss": 0.1226,
      "step": 50
    },
    {
      "epoch": 0.03825956489122281,
      "grad_norm": 0.1350037157535553,
      "learning_rate": 1.9101123595505618e-05,
      "loss": 0.0691,
      "step": 51
    },
    {
      "epoch": 0.03900975243810953,
      "grad_norm": 0.24726983904838562,
      "learning_rate": 1.9475655430711613e-05,
      "loss": 0.1451,
      "step": 52
    },
    {
      "epoch": 0.03975993998499625,
      "grad_norm": 0.30878591537475586,
      "learning_rate": 1.9850187265917604e-05,
      "loss": 0.2132,
      "step": 53
    },
    {
      "epoch": 0.04051012753188297,
      "grad_norm": 0.2271026372909546,
      "learning_rate": 2.0224719101123596e-05,
      "loss": 0.132,
      "step": 54
    },
    {
      "epoch": 0.04126031507876969,
      "grad_norm": 0.2296169400215149,
      "learning_rate": 2.059925093632959e-05,
      "loss": 0.1274,
      "step": 55
    },
    {
      "epoch": 0.042010502625656414,
      "grad_norm": 0.1105537861585617,
      "learning_rate": 2.0973782771535582e-05,
      "loss": 0.0744,
      "step": 56
    },
    {
      "epoch": 0.04276069017254314,
      "grad_norm": 0.2400723546743393,
      "learning_rate": 2.1348314606741574e-05,
      "loss": 0.1027,
      "step": 57
    },
    {
      "epoch": 0.04351087771942986,
      "grad_norm": 0.5178837180137634,
      "learning_rate": 2.1722846441947566e-05,
      "loss": 0.1871,
      "step": 58
    },
    {
      "epoch": 0.04426106526631658,
      "grad_norm": 0.16341057419776917,
      "learning_rate": 2.209737827715356e-05,
      "loss": 0.0965,
      "step": 59
    },
    {
      "epoch": 0.0450112528132033,
      "grad_norm": 0.1322585940361023,
      "learning_rate": 2.2471910112359552e-05,
      "loss": 0.1001,
      "step": 60
    },
    {
      "epoch": 0.04576144036009002,
      "grad_norm": 0.20453320443630219,
      "learning_rate": 2.2846441947565544e-05,
      "loss": 0.1281,
      "step": 61
    },
    {
      "epoch": 0.046511627906976744,
      "grad_norm": 0.16821111738681793,
      "learning_rate": 2.322097378277154e-05,
      "loss": 0.0779,
      "step": 62
    },
    {
      "epoch": 0.047261815453863466,
      "grad_norm": 0.23807474970817566,
      "learning_rate": 2.359550561797753e-05,
      "loss": 0.16,
      "step": 63
    },
    {
      "epoch": 0.04801200300075019,
      "grad_norm": 0.22742918133735657,
      "learning_rate": 2.3970037453183522e-05,
      "loss": 0.12,
      "step": 64
    },
    {
      "epoch": 0.04876219054763691,
      "grad_norm": 0.15059584379196167,
      "learning_rate": 2.4344569288389517e-05,
      "loss": 0.085,
      "step": 65
    },
    {
      "epoch": 0.04951237809452363,
      "grad_norm": 0.4061318635940552,
      "learning_rate": 2.4719101123595505e-05,
      "loss": 0.107,
      "step": 66
    },
    {
      "epoch": 0.05026256564141035,
      "grad_norm": 0.14806576073169708,
      "learning_rate": 2.50936329588015e-05,
      "loss": 0.0733,
      "step": 67
    },
    {
      "epoch": 0.05101275318829707,
      "grad_norm": 0.22779157757759094,
      "learning_rate": 2.546816479400749e-05,
      "loss": 0.0834,
      "step": 68
    },
    {
      "epoch": 0.051762940735183796,
      "grad_norm": 0.14479000866413116,
      "learning_rate": 2.5842696629213486e-05,
      "loss": 0.1031,
      "step": 69
    },
    {
      "epoch": 0.05251312828207052,
      "grad_norm": 0.134308323264122,
      "learning_rate": 2.6217228464419475e-05,
      "loss": 0.0768,
      "step": 70
    },
    {
      "epoch": 0.05326331582895724,
      "grad_norm": 0.2619442641735077,
      "learning_rate": 2.6591760299625466e-05,
      "loss": 0.1374,
      "step": 71
    },
    {
      "epoch": 0.05401350337584396,
      "grad_norm": 0.15653356909751892,
      "learning_rate": 2.696629213483146e-05,
      "loss": 0.1015,
      "step": 72
    },
    {
      "epoch": 0.05476369092273068,
      "grad_norm": 0.16244176030158997,
      "learning_rate": 2.7340823970037456e-05,
      "loss": 0.0686,
      "step": 73
    },
    {
      "epoch": 0.0555138784696174,
      "grad_norm": 0.2526662051677704,
      "learning_rate": 2.7715355805243448e-05,
      "loss": 0.1234,
      "step": 74
    },
    {
      "epoch": 0.056264066016504126,
      "grad_norm": 0.22778008878231049,
      "learning_rate": 2.8089887640449443e-05,
      "loss": 0.1534,
      "step": 75
    },
    {
      "epoch": 0.05701425356339085,
      "grad_norm": 0.27490678429603577,
      "learning_rate": 2.846441947565543e-05,
      "loss": 0.1303,
      "step": 76
    },
    {
      "epoch": 0.05776444111027757,
      "grad_norm": 0.13964544236660004,
      "learning_rate": 2.8838951310861422e-05,
      "loss": 0.0939,
      "step": 77
    },
    {
      "epoch": 0.058514628657164294,
      "grad_norm": 0.20751768350601196,
      "learning_rate": 2.9213483146067417e-05,
      "loss": 0.1046,
      "step": 78
    },
    {
      "epoch": 0.05926481620405101,
      "grad_norm": 0.2151343822479248,
      "learning_rate": 2.958801498127341e-05,
      "loss": 0.0932,
      "step": 79
    },
    {
      "epoch": 0.06001500375093773,
      "grad_norm": 0.13716846704483032,
      "learning_rate": 2.9962546816479404e-05,
      "loss": 0.0689,
      "step": 80
    },
    {
      "epoch": 0.060765191297824456,
      "grad_norm": 0.204776793718338,
      "learning_rate": 3.0337078651685396e-05,
      "loss": 0.0959,
      "step": 81
    },
    {
      "epoch": 0.06151537884471118,
      "grad_norm": 0.37804269790649414,
      "learning_rate": 3.0711610486891384e-05,
      "loss": 0.1014,
      "step": 82
    },
    {
      "epoch": 0.0622655663915979,
      "grad_norm": 0.2825986444950104,
      "learning_rate": 3.108614232209738e-05,
      "loss": 0.1371,
      "step": 83
    },
    {
      "epoch": 0.06301575393848462,
      "grad_norm": 0.1622900366783142,
      "learning_rate": 3.1460674157303374e-05,
      "loss": 0.0767,
      "step": 84
    },
    {
      "epoch": 0.06376594148537135,
      "grad_norm": 0.21700279414653778,
      "learning_rate": 3.183520599250936e-05,
      "loss": 0.0826,
      "step": 85
    },
    {
      "epoch": 0.06451612903225806,
      "grad_norm": 0.14375083148479462,
      "learning_rate": 3.220973782771536e-05,
      "loss": 0.0766,
      "step": 86
    },
    {
      "epoch": 0.06526631657914479,
      "grad_norm": 0.20387856662273407,
      "learning_rate": 3.258426966292135e-05,
      "loss": 0.091,
      "step": 87
    },
    {
      "epoch": 0.06601650412603151,
      "grad_norm": 0.20541413128376007,
      "learning_rate": 3.295880149812734e-05,
      "loss": 0.1184,
      "step": 88
    },
    {
      "epoch": 0.06676669167291822,
      "grad_norm": 0.1971144676208496,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.1029,
      "step": 89
    },
    {
      "epoch": 0.06751687921980495,
      "grad_norm": 0.18307439982891083,
      "learning_rate": 3.370786516853933e-05,
      "loss": 0.0993,
      "step": 90
    },
    {
      "epoch": 0.06826706676669167,
      "grad_norm": 0.1350952833890915,
      "learning_rate": 3.408239700374532e-05,
      "loss": 0.0529,
      "step": 91
    },
    {
      "epoch": 0.0690172543135784,
      "grad_norm": 0.15601865947246552,
      "learning_rate": 3.445692883895131e-05,
      "loss": 0.0772,
      "step": 92
    },
    {
      "epoch": 0.06976744186046512,
      "grad_norm": 0.09120382368564606,
      "learning_rate": 3.483146067415731e-05,
      "loss": 0.055,
      "step": 93
    },
    {
      "epoch": 0.07051762940735183,
      "grad_norm": 0.16563227772712708,
      "learning_rate": 3.52059925093633e-05,
      "loss": 0.0651,
      "step": 94
    },
    {
      "epoch": 0.07126781695423856,
      "grad_norm": 0.19384774565696716,
      "learning_rate": 3.558052434456929e-05,
      "loss": 0.0749,
      "step": 95
    },
    {
      "epoch": 0.07201800450112528,
      "grad_norm": 0.1468099057674408,
      "learning_rate": 3.595505617977528e-05,
      "loss": 0.0673,
      "step": 96
    },
    {
      "epoch": 0.072768192048012,
      "grad_norm": 0.30378851294517517,
      "learning_rate": 3.6329588014981274e-05,
      "loss": 0.0638,
      "step": 97
    },
    {
      "epoch": 0.07351837959489872,
      "grad_norm": 0.2536369562149048,
      "learning_rate": 3.670411985018727e-05,
      "loss": 0.0932,
      "step": 98
    },
    {
      "epoch": 0.07426856714178545,
      "grad_norm": 0.1923932582139969,
      "learning_rate": 3.7078651685393264e-05,
      "loss": 0.0732,
      "step": 99
    },
    {
      "epoch": 0.07501875468867217,
      "grad_norm": 0.20941613614559174,
      "learning_rate": 3.745318352059925e-05,
      "loss": 0.0971,
      "step": 100
    },
    {
      "epoch": 0.07576894223555888,
      "grad_norm": 0.17410755157470703,
      "learning_rate": 3.782771535580524e-05,
      "loss": 0.081,
      "step": 101
    },
    {
      "epoch": 0.07651912978244561,
      "grad_norm": 0.1977808177471161,
      "learning_rate": 3.8202247191011236e-05,
      "loss": 0.0933,
      "step": 102
    },
    {
      "epoch": 0.07726931732933233,
      "grad_norm": 0.25273486971855164,
      "learning_rate": 3.857677902621723e-05,
      "loss": 0.0739,
      "step": 103
    },
    {
      "epoch": 0.07801950487621906,
      "grad_norm": 0.29576393961906433,
      "learning_rate": 3.8951310861423226e-05,
      "loss": 0.0938,
      "step": 104
    },
    {
      "epoch": 0.07876969242310577,
      "grad_norm": 0.1943545937538147,
      "learning_rate": 3.9325842696629214e-05,
      "loss": 0.0993,
      "step": 105
    },
    {
      "epoch": 0.0795198799699925,
      "grad_norm": 0.2052256166934967,
      "learning_rate": 3.970037453183521e-05,
      "loss": 0.0803,
      "step": 106
    },
    {
      "epoch": 0.08027006751687922,
      "grad_norm": 0.2309311181306839,
      "learning_rate": 4.00749063670412e-05,
      "loss": 0.0663,
      "step": 107
    },
    {
      "epoch": 0.08102025506376594,
      "grad_norm": 0.22197790443897247,
      "learning_rate": 4.044943820224719e-05,
      "loss": 0.1032,
      "step": 108
    },
    {
      "epoch": 0.08177044261065267,
      "grad_norm": 0.13494251668453217,
      "learning_rate": 4.082397003745319e-05,
      "loss": 0.0557,
      "step": 109
    },
    {
      "epoch": 0.08252063015753938,
      "grad_norm": 0.2530130445957184,
      "learning_rate": 4.119850187265918e-05,
      "loss": 0.0802,
      "step": 110
    },
    {
      "epoch": 0.08327081770442611,
      "grad_norm": 0.20438532531261444,
      "learning_rate": 4.157303370786517e-05,
      "loss": 0.1289,
      "step": 111
    },
    {
      "epoch": 0.08402100525131283,
      "grad_norm": 0.1981775015592575,
      "learning_rate": 4.1947565543071165e-05,
      "loss": 0.0578,
      "step": 112
    },
    {
      "epoch": 0.08477119279819954,
      "grad_norm": 0.30348509550094604,
      "learning_rate": 4.232209737827715e-05,
      "loss": 0.0826,
      "step": 113
    },
    {
      "epoch": 0.08552138034508627,
      "grad_norm": 0.21688346564769745,
      "learning_rate": 4.269662921348315e-05,
      "loss": 0.0693,
      "step": 114
    },
    {
      "epoch": 0.08627156789197299,
      "grad_norm": 0.21740299463272095,
      "learning_rate": 4.307116104868914e-05,
      "loss": 0.0788,
      "step": 115
    },
    {
      "epoch": 0.08702175543885972,
      "grad_norm": 0.29491832852363586,
      "learning_rate": 4.344569288389513e-05,
      "loss": 0.084,
      "step": 116
    },
    {
      "epoch": 0.08777194298574643,
      "grad_norm": 0.1748553067445755,
      "learning_rate": 4.3820224719101126e-05,
      "loss": 0.0671,
      "step": 117
    },
    {
      "epoch": 0.08852213053263316,
      "grad_norm": 0.19076122343540192,
      "learning_rate": 4.419475655430712e-05,
      "loss": 0.0715,
      "step": 118
    },
    {
      "epoch": 0.08927231807951988,
      "grad_norm": 0.2707376778125763,
      "learning_rate": 4.456928838951311e-05,
      "loss": 0.1027,
      "step": 119
    },
    {
      "epoch": 0.0900225056264066,
      "grad_norm": 0.1251923143863678,
      "learning_rate": 4.4943820224719104e-05,
      "loss": 0.0594,
      "step": 120
    },
    {
      "epoch": 0.09077269317329333,
      "grad_norm": 0.18228654563426971,
      "learning_rate": 4.531835205992509e-05,
      "loss": 0.0581,
      "step": 121
    },
    {
      "epoch": 0.09152288072018004,
      "grad_norm": 0.20524358749389648,
      "learning_rate": 4.569288389513109e-05,
      "loss": 0.0627,
      "step": 122
    },
    {
      "epoch": 0.09227306826706677,
      "grad_norm": 0.10450001060962677,
      "learning_rate": 4.606741573033708e-05,
      "loss": 0.0731,
      "step": 123
    },
    {
      "epoch": 0.09302325581395349,
      "grad_norm": 0.18328773975372314,
      "learning_rate": 4.644194756554308e-05,
      "loss": 0.092,
      "step": 124
    },
    {
      "epoch": 0.09377344336084022,
      "grad_norm": 0.208187073469162,
      "learning_rate": 4.6816479400749066e-05,
      "loss": 0.0639,
      "step": 125
    },
    {
      "epoch": 0.09452363090772693,
      "grad_norm": 0.2710714638233185,
      "learning_rate": 4.719101123595506e-05,
      "loss": 0.0841,
      "step": 126
    },
    {
      "epoch": 0.09527381845461365,
      "grad_norm": 0.2559587359428406,
      "learning_rate": 4.756554307116105e-05,
      "loss": 0.0984,
      "step": 127
    },
    {
      "epoch": 0.09602400600150038,
      "grad_norm": 0.16492657363414764,
      "learning_rate": 4.7940074906367044e-05,
      "loss": 0.0491,
      "step": 128
    },
    {
      "epoch": 0.0967741935483871,
      "grad_norm": 0.15822099149227142,
      "learning_rate": 4.831460674157304e-05,
      "loss": 0.0395,
      "step": 129
    },
    {
      "epoch": 0.09752438109527382,
      "grad_norm": 0.1362650990486145,
      "learning_rate": 4.8689138576779034e-05,
      "loss": 0.0516,
      "step": 130
    },
    {
      "epoch": 0.09827456864216054,
      "grad_norm": 0.28588271141052246,
      "learning_rate": 4.906367041198502e-05,
      "loss": 0.1132,
      "step": 131
    },
    {
      "epoch": 0.09902475618904726,
      "grad_norm": 0.3099716305732727,
      "learning_rate": 4.943820224719101e-05,
      "loss": 0.0726,
      "step": 132
    },
    {
      "epoch": 0.09977494373593399,
      "grad_norm": 0.10133063048124313,
      "learning_rate": 4.9812734082397005e-05,
      "loss": 0.0515,
      "step": 133
    },
    {
      "epoch": 0.1005251312828207,
      "grad_norm": 0.18876691162586212,
      "learning_rate": 5.0187265917603e-05,
      "loss": 0.0918,
      "step": 134
    },
    {
      "epoch": 0.10127531882970743,
      "grad_norm": 0.3820139765739441,
      "learning_rate": 5.0561797752808995e-05,
      "loss": 0.1141,
      "step": 135
    },
    {
      "epoch": 0.10202550637659415,
      "grad_norm": 0.26029330492019653,
      "learning_rate": 5.093632958801498e-05,
      "loss": 0.0843,
      "step": 136
    },
    {
      "epoch": 0.10277569392348088,
      "grad_norm": 0.17507338523864746,
      "learning_rate": 5.131086142322098e-05,
      "loss": 0.0433,
      "step": 137
    },
    {
      "epoch": 0.10352588147036759,
      "grad_norm": 0.2125600278377533,
      "learning_rate": 5.168539325842697e-05,
      "loss": 0.0931,
      "step": 138
    },
    {
      "epoch": 0.10427606901725431,
      "grad_norm": 0.2325267344713211,
      "learning_rate": 5.205992509363297e-05,
      "loss": 0.048,
      "step": 139
    },
    {
      "epoch": 0.10502625656414104,
      "grad_norm": 0.1814715564250946,
      "learning_rate": 5.243445692883895e-05,
      "loss": 0.0708,
      "step": 140
    },
    {
      "epoch": 0.10577644411102775,
      "grad_norm": 0.13349854946136475,
      "learning_rate": 5.2808988764044944e-05,
      "loss": 0.0653,
      "step": 141
    },
    {
      "epoch": 0.10652663165791448,
      "grad_norm": 0.13655948638916016,
      "learning_rate": 5.318352059925093e-05,
      "loss": 0.0407,
      "step": 142
    },
    {
      "epoch": 0.1072768192048012,
      "grad_norm": 0.24276535212993622,
      "learning_rate": 5.355805243445693e-05,
      "loss": 0.0837,
      "step": 143
    },
    {
      "epoch": 0.10802700675168792,
      "grad_norm": 0.20750541985034943,
      "learning_rate": 5.393258426966292e-05,
      "loss": 0.0766,
      "step": 144
    },
    {
      "epoch": 0.10877719429857464,
      "grad_norm": 0.37364712357521057,
      "learning_rate": 5.430711610486892e-05,
      "loss": 0.0974,
      "step": 145
    },
    {
      "epoch": 0.10952738184546136,
      "grad_norm": 0.44260385632514954,
      "learning_rate": 5.468164794007491e-05,
      "loss": 0.0679,
      "step": 146
    },
    {
      "epoch": 0.11027756939234809,
      "grad_norm": 0.21232500672340393,
      "learning_rate": 5.50561797752809e-05,
      "loss": 0.0753,
      "step": 147
    },
    {
      "epoch": 0.1110277569392348,
      "grad_norm": 0.4461332857608795,
      "learning_rate": 5.5430711610486895e-05,
      "loss": 0.1543,
      "step": 148
    },
    {
      "epoch": 0.11177794448612154,
      "grad_norm": 0.26154395937919617,
      "learning_rate": 5.580524344569289e-05,
      "loss": 0.1068,
      "step": 149
    },
    {
      "epoch": 0.11252813203300825,
      "grad_norm": 0.22164535522460938,
      "learning_rate": 5.6179775280898885e-05,
      "loss": 0.0687,
      "step": 150
    },
    {
      "epoch": 0.11327831957989497,
      "grad_norm": 0.23080644011497498,
      "learning_rate": 5.6554307116104874e-05,
      "loss": 0.0623,
      "step": 151
    },
    {
      "epoch": 0.1140285071267817,
      "grad_norm": 0.20958930253982544,
      "learning_rate": 5.692883895131086e-05,
      "loss": 0.058,
      "step": 152
    },
    {
      "epoch": 0.11477869467366841,
      "grad_norm": 0.14259661734104156,
      "learning_rate": 5.730337078651685e-05,
      "loss": 0.0608,
      "step": 153
    },
    {
      "epoch": 0.11552888222055514,
      "grad_norm": 0.14598913490772247,
      "learning_rate": 5.7677902621722845e-05,
      "loss": 0.0652,
      "step": 154
    },
    {
      "epoch": 0.11627906976744186,
      "grad_norm": 0.11540807038545609,
      "learning_rate": 5.805243445692884e-05,
      "loss": 0.0338,
      "step": 155
    },
    {
      "epoch": 0.11702925731432859,
      "grad_norm": 0.1490037888288498,
      "learning_rate": 5.8426966292134835e-05,
      "loss": 0.0489,
      "step": 156
    },
    {
      "epoch": 0.1177794448612153,
      "grad_norm": 0.16232560575008392,
      "learning_rate": 5.880149812734082e-05,
      "loss": 0.0514,
      "step": 157
    },
    {
      "epoch": 0.11852963240810202,
      "grad_norm": 0.22876153886318207,
      "learning_rate": 5.917602996254682e-05,
      "loss": 0.0899,
      "step": 158
    },
    {
      "epoch": 0.11927981995498875,
      "grad_norm": 0.2434694468975067,
      "learning_rate": 5.955056179775281e-05,
      "loss": 0.0626,
      "step": 159
    },
    {
      "epoch": 0.12003000750187547,
      "grad_norm": 0.29315224289894104,
      "learning_rate": 5.992509363295881e-05,
      "loss": 0.0916,
      "step": 160
    },
    {
      "epoch": 0.1207801950487622,
      "grad_norm": 0.16403856873512268,
      "learning_rate": 6.02996254681648e-05,
      "loss": 0.0608,
      "step": 161
    },
    {
      "epoch": 0.12153038259564891,
      "grad_norm": 0.16882139444351196,
      "learning_rate": 6.067415730337079e-05,
      "loss": 0.0986,
      "step": 162
    },
    {
      "epoch": 0.12228057014253563,
      "grad_norm": 0.1665971726179123,
      "learning_rate": 6.104868913857679e-05,
      "loss": 0.0773,
      "step": 163
    },
    {
      "epoch": 0.12303075768942236,
      "grad_norm": 0.16212251782417297,
      "learning_rate": 6.142322097378277e-05,
      "loss": 0.0775,
      "step": 164
    },
    {
      "epoch": 0.12378094523630907,
      "grad_norm": 0.12361068278551102,
      "learning_rate": 6.179775280898876e-05,
      "loss": 0.0535,
      "step": 165
    },
    {
      "epoch": 0.1245311327831958,
      "grad_norm": 0.12108980119228363,
      "learning_rate": 6.217228464419476e-05,
      "loss": 0.0474,
      "step": 166
    },
    {
      "epoch": 0.12528132033008252,
      "grad_norm": 0.12282898277044296,
      "learning_rate": 6.254681647940075e-05,
      "loss": 0.0419,
      "step": 167
    },
    {
      "epoch": 0.12603150787696923,
      "grad_norm": 0.12157752364873886,
      "learning_rate": 6.292134831460675e-05,
      "loss": 0.0481,
      "step": 168
    },
    {
      "epoch": 0.12678169542385595,
      "grad_norm": 0.17761266231536865,
      "learning_rate": 6.329588014981274e-05,
      "loss": 0.0621,
      "step": 169
    },
    {
      "epoch": 0.1275318829707427,
      "grad_norm": 0.19057996571063995,
      "learning_rate": 6.367041198501872e-05,
      "loss": 0.0649,
      "step": 170
    },
    {
      "epoch": 0.1282820705176294,
      "grad_norm": 0.2076871246099472,
      "learning_rate": 6.404494382022472e-05,
      "loss": 0.077,
      "step": 171
    },
    {
      "epoch": 0.12903225806451613,
      "grad_norm": 0.19672785699367523,
      "learning_rate": 6.441947565543071e-05,
      "loss": 0.067,
      "step": 172
    },
    {
      "epoch": 0.12978244561140284,
      "grad_norm": 0.19830764830112457,
      "learning_rate": 6.479400749063671e-05,
      "loss": 0.068,
      "step": 173
    },
    {
      "epoch": 0.13053263315828958,
      "grad_norm": 0.160996675491333,
      "learning_rate": 6.51685393258427e-05,
      "loss": 0.0613,
      "step": 174
    },
    {
      "epoch": 0.1312828207051763,
      "grad_norm": 0.14649710059165955,
      "learning_rate": 6.55430711610487e-05,
      "loss": 0.0432,
      "step": 175
    },
    {
      "epoch": 0.13203300825206302,
      "grad_norm": 0.22120599448680878,
      "learning_rate": 6.591760299625468e-05,
      "loss": 0.0517,
      "step": 176
    },
    {
      "epoch": 0.13278319579894973,
      "grad_norm": 0.17631229758262634,
      "learning_rate": 6.629213483146067e-05,
      "loss": 0.0816,
      "step": 177
    },
    {
      "epoch": 0.13353338334583645,
      "grad_norm": 0.2152678519487381,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0522,
      "step": 178
    },
    {
      "epoch": 0.1342835708927232,
      "grad_norm": 0.22123490273952484,
      "learning_rate": 6.704119850187266e-05,
      "loss": 0.0771,
      "step": 179
    },
    {
      "epoch": 0.1350337584396099,
      "grad_norm": 0.37060683965682983,
      "learning_rate": 6.741573033707866e-05,
      "loss": 0.0641,
      "step": 180
    },
    {
      "epoch": 0.13578394598649662,
      "grad_norm": 0.26469045877456665,
      "learning_rate": 6.779026217228464e-05,
      "loss": 0.0889,
      "step": 181
    },
    {
      "epoch": 0.13653413353338334,
      "grad_norm": 0.19371746480464935,
      "learning_rate": 6.816479400749064e-05,
      "loss": 0.0506,
      "step": 182
    },
    {
      "epoch": 0.13728432108027006,
      "grad_norm": 0.1641053408384323,
      "learning_rate": 6.853932584269663e-05,
      "loss": 0.0672,
      "step": 183
    },
    {
      "epoch": 0.1380345086271568,
      "grad_norm": 0.2839615046977997,
      "learning_rate": 6.891385767790263e-05,
      "loss": 0.0929,
      "step": 184
    },
    {
      "epoch": 0.13878469617404351,
      "grad_norm": 0.20816008746623993,
      "learning_rate": 6.928838951310862e-05,
      "loss": 0.0597,
      "step": 185
    },
    {
      "epoch": 0.13953488372093023,
      "grad_norm": 0.1270139068365097,
      "learning_rate": 6.966292134831462e-05,
      "loss": 0.0484,
      "step": 186
    },
    {
      "epoch": 0.14028507126781695,
      "grad_norm": 0.11964840441942215,
      "learning_rate": 7.003745318352061e-05,
      "loss": 0.0459,
      "step": 187
    },
    {
      "epoch": 0.14103525881470366,
      "grad_norm": 0.2520076334476471,
      "learning_rate": 7.04119850187266e-05,
      "loss": 0.1065,
      "step": 188
    },
    {
      "epoch": 0.1417854463615904,
      "grad_norm": 0.23095941543579102,
      "learning_rate": 7.078651685393259e-05,
      "loss": 0.0969,
      "step": 189
    },
    {
      "epoch": 0.14253563390847712,
      "grad_norm": 0.2514033913612366,
      "learning_rate": 7.116104868913858e-05,
      "loss": 0.0812,
      "step": 190
    },
    {
      "epoch": 0.14328582145536384,
      "grad_norm": 0.20976576209068298,
      "learning_rate": 7.153558052434456e-05,
      "loss": 0.0905,
      "step": 191
    },
    {
      "epoch": 0.14403600900225055,
      "grad_norm": 0.18989625573158264,
      "learning_rate": 7.191011235955056e-05,
      "loss": 0.062,
      "step": 192
    },
    {
      "epoch": 0.1447861965491373,
      "grad_norm": 0.1842532753944397,
      "learning_rate": 7.228464419475655e-05,
      "loss": 0.0588,
      "step": 193
    },
    {
      "epoch": 0.145536384096024,
      "grad_norm": 0.2708159387111664,
      "learning_rate": 7.265917602996255e-05,
      "loss": 0.1267,
      "step": 194
    },
    {
      "epoch": 0.14628657164291073,
      "grad_norm": 0.15724222362041473,
      "learning_rate": 7.303370786516854e-05,
      "loss": 0.066,
      "step": 195
    },
    {
      "epoch": 0.14703675918979744,
      "grad_norm": 0.13891983032226562,
      "learning_rate": 7.340823970037454e-05,
      "loss": 0.0687,
      "step": 196
    },
    {
      "epoch": 0.14778694673668416,
      "grad_norm": 0.10517305880784988,
      "learning_rate": 7.378277153558053e-05,
      "loss": 0.0509,
      "step": 197
    },
    {
      "epoch": 0.1485371342835709,
      "grad_norm": 0.26864030957221985,
      "learning_rate": 7.415730337078653e-05,
      "loss": 0.0655,
      "step": 198
    },
    {
      "epoch": 0.14928732183045762,
      "grad_norm": 0.20357665419578552,
      "learning_rate": 7.453183520599252e-05,
      "loss": 0.0651,
      "step": 199
    },
    {
      "epoch": 0.15003750937734434,
      "grad_norm": 0.20443439483642578,
      "learning_rate": 7.49063670411985e-05,
      "loss": 0.093,
      "step": 200
    },
    {
      "epoch": 0.15003750937734434,
      "eval_loss": 0.07821805030107498,
      "eval_runtime": 2.6636,
      "eval_samples_per_second": 20.273,
      "eval_steps_per_second": 5.256,
      "step": 200
    },
    {
      "epoch": 0.15078769692423105,
      "grad_norm": 0.2003447413444519,
      "learning_rate": 7.52808988764045e-05,
      "loss": 0.0846,
      "step": 201
    },
    {
      "epoch": 0.15153788447111777,
      "grad_norm": 0.148345485329628,
      "learning_rate": 7.565543071161048e-05,
      "loss": 0.0649,
      "step": 202
    },
    {
      "epoch": 0.1522880720180045,
      "grad_norm": 0.11558017134666443,
      "learning_rate": 7.602996254681648e-05,
      "loss": 0.0565,
      "step": 203
    },
    {
      "epoch": 0.15303825956489123,
      "grad_norm": 0.197371244430542,
      "learning_rate": 7.640449438202247e-05,
      "loss": 0.0704,
      "step": 204
    },
    {
      "epoch": 0.15378844711177794,
      "grad_norm": 0.1327037215232849,
      "learning_rate": 7.677902621722847e-05,
      "loss": 0.049,
      "step": 205
    },
    {
      "epoch": 0.15453863465866466,
      "grad_norm": 0.1790093332529068,
      "learning_rate": 7.715355805243446e-05,
      "loss": 0.0799,
      "step": 206
    },
    {
      "epoch": 0.15528882220555137,
      "grad_norm": 0.13199618458747864,
      "learning_rate": 7.752808988764046e-05,
      "loss": 0.0499,
      "step": 207
    },
    {
      "epoch": 0.15603900975243812,
      "grad_norm": 0.2246326208114624,
      "learning_rate": 7.790262172284645e-05,
      "loss": 0.0972,
      "step": 208
    },
    {
      "epoch": 0.15678919729932483,
      "grad_norm": 0.13354603946208954,
      "learning_rate": 7.827715355805245e-05,
      "loss": 0.0718,
      "step": 209
    },
    {
      "epoch": 0.15753938484621155,
      "grad_norm": 0.12375898659229279,
      "learning_rate": 7.865168539325843e-05,
      "loss": 0.0659,
      "step": 210
    },
    {
      "epoch": 0.15828957239309827,
      "grad_norm": 0.2637913227081299,
      "learning_rate": 7.902621722846442e-05,
      "loss": 0.0918,
      "step": 211
    },
    {
      "epoch": 0.159039759939985,
      "grad_norm": 0.23306505382061005,
      "learning_rate": 7.940074906367042e-05,
      "loss": 0.0789,
      "step": 212
    },
    {
      "epoch": 0.15978994748687173,
      "grad_norm": 0.1299886852502823,
      "learning_rate": 7.97752808988764e-05,
      "loss": 0.0605,
      "step": 213
    },
    {
      "epoch": 0.16054013503375844,
      "grad_norm": 0.21600346267223358,
      "learning_rate": 8.01498127340824e-05,
      "loss": 0.0624,
      "step": 214
    },
    {
      "epoch": 0.16129032258064516,
      "grad_norm": 0.16830679774284363,
      "learning_rate": 8.052434456928839e-05,
      "loss": 0.0756,
      "step": 215
    },
    {
      "epoch": 0.16204051012753187,
      "grad_norm": 0.4400419294834137,
      "learning_rate": 8.089887640449438e-05,
      "loss": 0.1069,
      "step": 216
    },
    {
      "epoch": 0.16279069767441862,
      "grad_norm": 0.15384125709533691,
      "learning_rate": 8.127340823970038e-05,
      "loss": 0.0756,
      "step": 217
    },
    {
      "epoch": 0.16354088522130533,
      "grad_norm": 0.1943519562482834,
      "learning_rate": 8.164794007490637e-05,
      "loss": 0.0847,
      "step": 218
    },
    {
      "epoch": 0.16429107276819205,
      "grad_norm": 0.17785924673080444,
      "learning_rate": 8.202247191011237e-05,
      "loss": 0.0641,
      "step": 219
    },
    {
      "epoch": 0.16504126031507876,
      "grad_norm": 0.342724472284317,
      "learning_rate": 8.239700374531836e-05,
      "loss": 0.1031,
      "step": 220
    },
    {
      "epoch": 0.16579144786196548,
      "grad_norm": 0.18801873922348022,
      "learning_rate": 8.277153558052434e-05,
      "loss": 0.0822,
      "step": 221
    },
    {
      "epoch": 0.16654163540885222,
      "grad_norm": 0.14902432262897491,
      "learning_rate": 8.314606741573034e-05,
      "loss": 0.0596,
      "step": 222
    },
    {
      "epoch": 0.16729182295573894,
      "grad_norm": 0.12913598120212555,
      "learning_rate": 8.352059925093633e-05,
      "loss": 0.069,
      "step": 223
    },
    {
      "epoch": 0.16804201050262565,
      "grad_norm": 0.13187310099601746,
      "learning_rate": 8.389513108614233e-05,
      "loss": 0.0647,
      "step": 224
    },
    {
      "epoch": 0.16879219804951237,
      "grad_norm": 0.15238381922245026,
      "learning_rate": 8.426966292134831e-05,
      "loss": 0.0566,
      "step": 225
    },
    {
      "epoch": 0.1695423855963991,
      "grad_norm": 0.32936590909957886,
      "learning_rate": 8.46441947565543e-05,
      "loss": 0.0842,
      "step": 226
    },
    {
      "epoch": 0.17029257314328583,
      "grad_norm": 0.16082966327667236,
      "learning_rate": 8.50187265917603e-05,
      "loss": 0.0703,
      "step": 227
    },
    {
      "epoch": 0.17104276069017255,
      "grad_norm": 0.15576517581939697,
      "learning_rate": 8.53932584269663e-05,
      "loss": 0.0776,
      "step": 228
    },
    {
      "epoch": 0.17179294823705926,
      "grad_norm": 0.14493928849697113,
      "learning_rate": 8.576779026217229e-05,
      "loss": 0.0522,
      "step": 229
    },
    {
      "epoch": 0.17254313578394598,
      "grad_norm": 0.14679628610610962,
      "learning_rate": 8.614232209737829e-05,
      "loss": 0.0743,
      "step": 230
    },
    {
      "epoch": 0.17329332333083272,
      "grad_norm": 0.19002974033355713,
      "learning_rate": 8.651685393258427e-05,
      "loss": 0.0923,
      "step": 231
    },
    {
      "epoch": 0.17404351087771944,
      "grad_norm": 0.284746915102005,
      "learning_rate": 8.689138576779026e-05,
      "loss": 0.1601,
      "step": 232
    },
    {
      "epoch": 0.17479369842460615,
      "grad_norm": 0.299917995929718,
      "learning_rate": 8.726591760299626e-05,
      "loss": 0.0805,
      "step": 233
    },
    {
      "epoch": 0.17554388597149287,
      "grad_norm": 0.18388903141021729,
      "learning_rate": 8.764044943820225e-05,
      "loss": 0.0675,
      "step": 234
    },
    {
      "epoch": 0.17629407351837958,
      "grad_norm": 0.18255142867565155,
      "learning_rate": 8.801498127340825e-05,
      "loss": 0.0948,
      "step": 235
    },
    {
      "epoch": 0.17704426106526633,
      "grad_norm": 0.14342963695526123,
      "learning_rate": 8.838951310861424e-05,
      "loss": 0.0717,
      "step": 236
    },
    {
      "epoch": 0.17779444861215304,
      "grad_norm": 0.22409477829933167,
      "learning_rate": 8.876404494382022e-05,
      "loss": 0.1043,
      "step": 237
    },
    {
      "epoch": 0.17854463615903976,
      "grad_norm": 0.13333116471767426,
      "learning_rate": 8.913857677902622e-05,
      "loss": 0.0661,
      "step": 238
    },
    {
      "epoch": 0.17929482370592648,
      "grad_norm": 0.18588890135288239,
      "learning_rate": 8.951310861423221e-05,
      "loss": 0.0722,
      "step": 239
    },
    {
      "epoch": 0.1800450112528132,
      "grad_norm": 0.13051137328147888,
      "learning_rate": 8.988764044943821e-05,
      "loss": 0.0554,
      "step": 240
    },
    {
      "epoch": 0.18079519879969994,
      "grad_norm": 0.18558816611766815,
      "learning_rate": 9.02621722846442e-05,
      "loss": 0.0782,
      "step": 241
    },
    {
      "epoch": 0.18154538634658665,
      "grad_norm": 0.16871553659439087,
      "learning_rate": 9.063670411985018e-05,
      "loss": 0.0699,
      "step": 242
    },
    {
      "epoch": 0.18229557389347337,
      "grad_norm": 0.19408437609672546,
      "learning_rate": 9.101123595505618e-05,
      "loss": 0.0937,
      "step": 243
    },
    {
      "epoch": 0.18304576144036008,
      "grad_norm": 0.19955557584762573,
      "learning_rate": 9.138576779026217e-05,
      "loss": 0.0844,
      "step": 244
    },
    {
      "epoch": 0.1837959489872468,
      "grad_norm": 0.1998845785856247,
      "learning_rate": 9.176029962546817e-05,
      "loss": 0.0749,
      "step": 245
    },
    {
      "epoch": 0.18454613653413354,
      "grad_norm": 0.1685137152671814,
      "learning_rate": 9.213483146067416e-05,
      "loss": 0.0875,
      "step": 246
    },
    {
      "epoch": 0.18529632408102026,
      "grad_norm": 0.13906221091747284,
      "learning_rate": 9.250936329588016e-05,
      "loss": 0.0604,
      "step": 247
    },
    {
      "epoch": 0.18604651162790697,
      "grad_norm": 0.11521290242671967,
      "learning_rate": 9.288389513108615e-05,
      "loss": 0.0438,
      "step": 248
    },
    {
      "epoch": 0.1867966991747937,
      "grad_norm": 0.15000270307064056,
      "learning_rate": 9.325842696629214e-05,
      "loss": 0.046,
      "step": 249
    },
    {
      "epoch": 0.18754688672168043,
      "grad_norm": 0.10028818994760513,
      "learning_rate": 9.363295880149813e-05,
      "loss": 0.0526,
      "step": 250
    },
    {
      "epoch": 0.18829707426856715,
      "grad_norm": 0.15884524583816528,
      "learning_rate": 9.400749063670413e-05,
      "loss": 0.0537,
      "step": 251
    },
    {
      "epoch": 0.18904726181545387,
      "grad_norm": 0.1237359270453453,
      "learning_rate": 9.438202247191012e-05,
      "loss": 0.0507,
      "step": 252
    },
    {
      "epoch": 0.18979744936234058,
      "grad_norm": 0.1722472906112671,
      "learning_rate": 9.47565543071161e-05,
      "loss": 0.0709,
      "step": 253
    },
    {
      "epoch": 0.1905476369092273,
      "grad_norm": 0.1201101616024971,
      "learning_rate": 9.51310861423221e-05,
      "loss": 0.0475,
      "step": 254
    },
    {
      "epoch": 0.19129782445611404,
      "grad_norm": 0.1595461666584015,
      "learning_rate": 9.550561797752809e-05,
      "loss": 0.0815,
      "step": 255
    },
    {
      "epoch": 0.19204801200300076,
      "grad_norm": 0.16221360862255096,
      "learning_rate": 9.588014981273409e-05,
      "loss": 0.0715,
      "step": 256
    },
    {
      "epoch": 0.19279819954988747,
      "grad_norm": 0.1589856594800949,
      "learning_rate": 9.625468164794008e-05,
      "loss": 0.0585,
      "step": 257
    },
    {
      "epoch": 0.1935483870967742,
      "grad_norm": 0.2617347538471222,
      "learning_rate": 9.662921348314608e-05,
      "loss": 0.0698,
      "step": 258
    },
    {
      "epoch": 0.1942985746436609,
      "grad_norm": 0.15338610112667084,
      "learning_rate": 9.700374531835207e-05,
      "loss": 0.0502,
      "step": 259
    },
    {
      "epoch": 0.19504876219054765,
      "grad_norm": 0.18816447257995605,
      "learning_rate": 9.737827715355807e-05,
      "loss": 0.0843,
      "step": 260
    },
    {
      "epoch": 0.19579894973743436,
      "grad_norm": 0.1613040417432785,
      "learning_rate": 9.775280898876405e-05,
      "loss": 0.0639,
      "step": 261
    },
    {
      "epoch": 0.19654913728432108,
      "grad_norm": 0.18618257343769073,
      "learning_rate": 9.812734082397004e-05,
      "loss": 0.0842,
      "step": 262
    },
    {
      "epoch": 0.1972993248312078,
      "grad_norm": 0.17338407039642334,
      "learning_rate": 9.850187265917602e-05,
      "loss": 0.0565,
      "step": 263
    },
    {
      "epoch": 0.1980495123780945,
      "grad_norm": 0.4943162798881531,
      "learning_rate": 9.887640449438202e-05,
      "loss": 0.1323,
      "step": 264
    },
    {
      "epoch": 0.19879969992498125,
      "grad_norm": 0.18759749829769135,
      "learning_rate": 9.925093632958801e-05,
      "loss": 0.1042,
      "step": 265
    },
    {
      "epoch": 0.19954988747186797,
      "grad_norm": 0.12457593530416489,
      "learning_rate": 9.962546816479401e-05,
      "loss": 0.0541,
      "step": 266
    },
    {
      "epoch": 0.2003000750187547,
      "grad_norm": 0.17804086208343506,
      "learning_rate": 0.0001,
      "loss": 0.071,
      "step": 267
    },
    {
      "epoch": 0.2010502625656414,
      "grad_norm": 0.10970258712768555,
      "learning_rate": 9.99999571274618e-05,
      "loss": 0.0418,
      "step": 268
    },
    {
      "epoch": 0.20180045011252815,
      "grad_norm": 0.11806442588567734,
      "learning_rate": 9.999982850992069e-05,
      "loss": 0.052,
      "step": 269
    },
    {
      "epoch": 0.20255063765941486,
      "grad_norm": 0.1669778674840927,
      "learning_rate": 9.999961414759727e-05,
      "loss": 0.0855,
      "step": 270
    },
    {
      "epoch": 0.20330082520630158,
      "grad_norm": 0.10167306661605835,
      "learning_rate": 9.999931404085912e-05,
      "loss": 0.0363,
      "step": 271
    },
    {
      "epoch": 0.2040510127531883,
      "grad_norm": 0.15393097698688507,
      "learning_rate": 9.999892819022092e-05,
      "loss": 0.0794,
      "step": 272
    },
    {
      "epoch": 0.204801200300075,
      "grad_norm": 0.12911133468151093,
      "learning_rate": 9.999845659634435e-05,
      "loss": 0.0613,
      "step": 273
    },
    {
      "epoch": 0.20555138784696175,
      "grad_norm": 0.17678648233413696,
      "learning_rate": 9.999789926003814e-05,
      "loss": 0.075,
      "step": 274
    },
    {
      "epoch": 0.20630157539384847,
      "grad_norm": 0.09235060214996338,
      "learning_rate": 9.999725618225808e-05,
      "loss": 0.0682,
      "step": 275
    },
    {
      "epoch": 0.20705176294073518,
      "grad_norm": 0.14196725189685822,
      "learning_rate": 9.999652736410698e-05,
      "loss": 0.0545,
      "step": 276
    },
    {
      "epoch": 0.2078019504876219,
      "grad_norm": 0.15323829650878906,
      "learning_rate": 9.999571280683468e-05,
      "loss": 0.0685,
      "step": 277
    },
    {
      "epoch": 0.20855213803450862,
      "grad_norm": 0.15211105346679688,
      "learning_rate": 9.99948125118381e-05,
      "loss": 0.074,
      "step": 278
    },
    {
      "epoch": 0.20930232558139536,
      "grad_norm": 0.12618400156497955,
      "learning_rate": 9.999382648066113e-05,
      "loss": 0.0621,
      "step": 279
    },
    {
      "epoch": 0.21005251312828208,
      "grad_norm": 0.16087733209133148,
      "learning_rate": 9.999275471499472e-05,
      "loss": 0.0999,
      "step": 280
    },
    {
      "epoch": 0.2108027006751688,
      "grad_norm": 0.1869620382785797,
      "learning_rate": 9.999159721667685e-05,
      "loss": 0.0839,
      "step": 281
    },
    {
      "epoch": 0.2115528882220555,
      "grad_norm": 0.15936337411403656,
      "learning_rate": 9.999035398769252e-05,
      "loss": 0.07,
      "step": 282
    },
    {
      "epoch": 0.21230307576894222,
      "grad_norm": 0.19384759664535522,
      "learning_rate": 9.998902503017372e-05,
      "loss": 0.1308,
      "step": 283
    },
    {
      "epoch": 0.21305326331582897,
      "grad_norm": 0.29375916719436646,
      "learning_rate": 9.99876103463995e-05,
      "loss": 0.0701,
      "step": 284
    },
    {
      "epoch": 0.21380345086271568,
      "grad_norm": 0.21341568231582642,
      "learning_rate": 9.998610993879589e-05,
      "loss": 0.0474,
      "step": 285
    },
    {
      "epoch": 0.2145536384096024,
      "grad_norm": 0.25241464376449585,
      "learning_rate": 9.998452380993597e-05,
      "loss": 0.068,
      "step": 286
    },
    {
      "epoch": 0.21530382595648911,
      "grad_norm": 0.18199339509010315,
      "learning_rate": 9.998285196253977e-05,
      "loss": 0.0693,
      "step": 287
    },
    {
      "epoch": 0.21605401350337583,
      "grad_norm": 0.11006759852170944,
      "learning_rate": 9.998109439947434e-05,
      "loss": 0.0392,
      "step": 288
    },
    {
      "epoch": 0.21680420105026257,
      "grad_norm": 0.132887065410614,
      "learning_rate": 9.997925112375375e-05,
      "loss": 0.0552,
      "step": 289
    },
    {
      "epoch": 0.2175543885971493,
      "grad_norm": 0.21448518335819244,
      "learning_rate": 9.997732213853902e-05,
      "loss": 0.0984,
      "step": 290
    },
    {
      "epoch": 0.218304576144036,
      "grad_norm": 0.18582187592983246,
      "learning_rate": 9.997530744713817e-05,
      "loss": 0.0898,
      "step": 291
    },
    {
      "epoch": 0.21905476369092272,
      "grad_norm": 0.2004450112581253,
      "learning_rate": 9.997320705300621e-05,
      "loss": 0.1014,
      "step": 292
    },
    {
      "epoch": 0.21980495123780946,
      "grad_norm": 0.15326693654060364,
      "learning_rate": 9.997102095974508e-05,
      "loss": 0.048,
      "step": 293
    },
    {
      "epoch": 0.22055513878469618,
      "grad_norm": 0.13675133883953094,
      "learning_rate": 9.996874917110378e-05,
      "loss": 0.0534,
      "step": 294
    },
    {
      "epoch": 0.2213053263315829,
      "grad_norm": 0.1403188854455948,
      "learning_rate": 9.996639169097811e-05,
      "loss": 0.0563,
      "step": 295
    },
    {
      "epoch": 0.2220555138784696,
      "grad_norm": 0.263437956571579,
      "learning_rate": 9.996394852341098e-05,
      "loss": 0.0865,
      "step": 296
    },
    {
      "epoch": 0.22280570142535633,
      "grad_norm": 0.17431458830833435,
      "learning_rate": 9.996141967259218e-05,
      "loss": 0.0582,
      "step": 297
    },
    {
      "epoch": 0.22355588897224307,
      "grad_norm": 0.19193384051322937,
      "learning_rate": 9.995880514285841e-05,
      "loss": 0.076,
      "step": 298
    },
    {
      "epoch": 0.2243060765191298,
      "grad_norm": 0.12529249489307404,
      "learning_rate": 9.995610493869336e-05,
      "loss": 0.074,
      "step": 299
    },
    {
      "epoch": 0.2250562640660165,
      "grad_norm": 0.20957054197788239,
      "learning_rate": 9.99533190647276e-05,
      "loss": 0.0807,
      "step": 300
    },
    {
      "epoch": 0.22580645161290322,
      "grad_norm": 0.14158689975738525,
      "learning_rate": 9.995044752573864e-05,
      "loss": 0.0549,
      "step": 301
    },
    {
      "epoch": 0.22655663915978994,
      "grad_norm": 0.09992564469575882,
      "learning_rate": 9.994749032665085e-05,
      "loss": 0.0511,
      "step": 302
    },
    {
      "epoch": 0.22730682670667668,
      "grad_norm": 0.21247920393943787,
      "learning_rate": 9.994444747253559e-05,
      "loss": 0.0645,
      "step": 303
    },
    {
      "epoch": 0.2280570142535634,
      "grad_norm": 0.10664507746696472,
      "learning_rate": 9.9941318968611e-05,
      "loss": 0.0357,
      "step": 304
    },
    {
      "epoch": 0.2288072018004501,
      "grad_norm": 0.19751517474651337,
      "learning_rate": 9.993810482024221e-05,
      "loss": 0.0761,
      "step": 305
    },
    {
      "epoch": 0.22955738934733683,
      "grad_norm": 0.13949774205684662,
      "learning_rate": 9.993480503294114e-05,
      "loss": 0.059,
      "step": 306
    },
    {
      "epoch": 0.23030757689422354,
      "grad_norm": 0.1572161316871643,
      "learning_rate": 9.993141961236661e-05,
      "loss": 0.0523,
      "step": 307
    },
    {
      "epoch": 0.23105776444111029,
      "grad_norm": 0.1472707837820053,
      "learning_rate": 9.992794856432426e-05,
      "loss": 0.0559,
      "step": 308
    },
    {
      "epoch": 0.231807951987997,
      "grad_norm": 0.2457486093044281,
      "learning_rate": 9.992439189476661e-05,
      "loss": 0.0948,
      "step": 309
    },
    {
      "epoch": 0.23255813953488372,
      "grad_norm": 0.12344438582658768,
      "learning_rate": 9.992074960979301e-05,
      "loss": 0.0447,
      "step": 310
    },
    {
      "epoch": 0.23330832708177043,
      "grad_norm": 0.17082221806049347,
      "learning_rate": 9.991702171564961e-05,
      "loss": 0.086,
      "step": 311
    },
    {
      "epoch": 0.23405851462865718,
      "grad_norm": 0.1637532263994217,
      "learning_rate": 9.991320821872939e-05,
      "loss": 0.0732,
      "step": 312
    },
    {
      "epoch": 0.2348087021755439,
      "grad_norm": 0.18615975975990295,
      "learning_rate": 9.990930912557209e-05,
      "loss": 0.075,
      "step": 313
    },
    {
      "epoch": 0.2355588897224306,
      "grad_norm": 0.10900649428367615,
      "learning_rate": 9.990532444286431e-05,
      "loss": 0.045,
      "step": 314
    },
    {
      "epoch": 0.23630907726931732,
      "grad_norm": 0.1474543958902359,
      "learning_rate": 9.990125417743937e-05,
      "loss": 0.0548,
      "step": 315
    },
    {
      "epoch": 0.23705926481620404,
      "grad_norm": 0.15200990438461304,
      "learning_rate": 9.989709833627736e-05,
      "loss": 0.0674,
      "step": 316
    },
    {
      "epoch": 0.23780945236309078,
      "grad_norm": 0.2598418891429901,
      "learning_rate": 9.989285692650518e-05,
      "loss": 0.1139,
      "step": 317
    },
    {
      "epoch": 0.2385596399099775,
      "grad_norm": 0.21949024498462677,
      "learning_rate": 9.98885299553964e-05,
      "loss": 0.0969,
      "step": 318
    },
    {
      "epoch": 0.23930982745686422,
      "grad_norm": 0.11348603665828705,
      "learning_rate": 9.988411743037134e-05,
      "loss": 0.0454,
      "step": 319
    },
    {
      "epoch": 0.24006001500375093,
      "grad_norm": 0.18390147387981415,
      "learning_rate": 9.987961935899706e-05,
      "loss": 0.0623,
      "step": 320
    },
    {
      "epoch": 0.24081020255063765,
      "grad_norm": 0.25778982043266296,
      "learning_rate": 9.987503574898731e-05,
      "loss": 0.0796,
      "step": 321
    },
    {
      "epoch": 0.2415603900975244,
      "grad_norm": 0.1339096575975418,
      "learning_rate": 9.987036660820255e-05,
      "loss": 0.0488,
      "step": 322
    },
    {
      "epoch": 0.2423105776444111,
      "grad_norm": 0.18395960330963135,
      "learning_rate": 9.986561194464985e-05,
      "loss": 0.0802,
      "step": 323
    },
    {
      "epoch": 0.24306076519129782,
      "grad_norm": 0.13696521520614624,
      "learning_rate": 9.986077176648303e-05,
      "loss": 0.0595,
      "step": 324
    },
    {
      "epoch": 0.24381095273818454,
      "grad_norm": 0.1406623274087906,
      "learning_rate": 9.985584608200251e-05,
      "loss": 0.0619,
      "step": 325
    },
    {
      "epoch": 0.24456114028507125,
      "grad_norm": 0.09625441581010818,
      "learning_rate": 9.985083489965534e-05,
      "loss": 0.0424,
      "step": 326
    },
    {
      "epoch": 0.245311327831958,
      "grad_norm": 0.12339973449707031,
      "learning_rate": 9.984573822803521e-05,
      "loss": 0.045,
      "step": 327
    },
    {
      "epoch": 0.24606151537884471,
      "grad_norm": 0.15541982650756836,
      "learning_rate": 9.984055607588242e-05,
      "loss": 0.0655,
      "step": 328
    },
    {
      "epoch": 0.24681170292573143,
      "grad_norm": 0.13536489009857178,
      "learning_rate": 9.983528845208384e-05,
      "loss": 0.0514,
      "step": 329
    },
    {
      "epoch": 0.24756189047261815,
      "grad_norm": 0.18628405034542084,
      "learning_rate": 9.982993536567293e-05,
      "loss": 0.0793,
      "step": 330
    },
    {
      "epoch": 0.2483120780195049,
      "grad_norm": 0.10619158297777176,
      "learning_rate": 9.98244968258297e-05,
      "loss": 0.0379,
      "step": 331
    },
    {
      "epoch": 0.2490622655663916,
      "grad_norm": 0.21251732110977173,
      "learning_rate": 9.981897284188073e-05,
      "loss": 0.1043,
      "step": 332
    },
    {
      "epoch": 0.24981245311327832,
      "grad_norm": 0.18133696913719177,
      "learning_rate": 9.981336342329909e-05,
      "loss": 0.0441,
      "step": 333
    },
    {
      "epoch": 0.25056264066016504,
      "grad_norm": 0.15581296384334564,
      "learning_rate": 9.980766857970438e-05,
      "loss": 0.0647,
      "step": 334
    },
    {
      "epoch": 0.25131282820705175,
      "grad_norm": 0.24356499314308167,
      "learning_rate": 9.98018883208627e-05,
      "loss": 0.073,
      "step": 335
    },
    {
      "epoch": 0.25206301575393847,
      "grad_norm": 0.10127891600131989,
      "learning_rate": 9.979602265668664e-05,
      "loss": 0.0461,
      "step": 336
    },
    {
      "epoch": 0.2528132033008252,
      "grad_norm": 0.13798540830612183,
      "learning_rate": 9.979007159723521e-05,
      "loss": 0.0595,
      "step": 337
    },
    {
      "epoch": 0.2535633908477119,
      "grad_norm": 0.13023167848587036,
      "learning_rate": 9.97840351527139e-05,
      "loss": 0.0604,
      "step": 338
    },
    {
      "epoch": 0.25431357839459867,
      "grad_norm": 0.12885215878486633,
      "learning_rate": 9.977791333347462e-05,
      "loss": 0.0564,
      "step": 339
    },
    {
      "epoch": 0.2550637659414854,
      "grad_norm": 0.23486711084842682,
      "learning_rate": 9.97717061500157e-05,
      "loss": 0.0684,
      "step": 340
    },
    {
      "epoch": 0.2558139534883721,
      "grad_norm": 0.09653028845787048,
      "learning_rate": 9.976541361298184e-05,
      "loss": 0.0477,
      "step": 341
    },
    {
      "epoch": 0.2565641410352588,
      "grad_norm": 0.18098174035549164,
      "learning_rate": 9.97590357331641e-05,
      "loss": 0.0573,
      "step": 342
    },
    {
      "epoch": 0.25731432858214554,
      "grad_norm": 0.1684122532606125,
      "learning_rate": 9.975257252149994e-05,
      "loss": 0.0499,
      "step": 343
    },
    {
      "epoch": 0.25806451612903225,
      "grad_norm": 0.2163161337375641,
      "learning_rate": 9.974602398907313e-05,
      "loss": 0.0719,
      "step": 344
    },
    {
      "epoch": 0.25881470367591897,
      "grad_norm": 0.1903722733259201,
      "learning_rate": 9.973939014711375e-05,
      "loss": 0.0758,
      "step": 345
    },
    {
      "epoch": 0.2595648912228057,
      "grad_norm": 0.2286502569913864,
      "learning_rate": 9.973267100699819e-05,
      "loss": 0.0647,
      "step": 346
    },
    {
      "epoch": 0.2603150787696924,
      "grad_norm": 0.20068782567977905,
      "learning_rate": 9.972586658024911e-05,
      "loss": 0.0849,
      "step": 347
    },
    {
      "epoch": 0.26106526631657917,
      "grad_norm": 0.17922401428222656,
      "learning_rate": 9.971897687853544e-05,
      "loss": 0.0526,
      "step": 348
    },
    {
      "epoch": 0.2618154538634659,
      "grad_norm": 0.301491379737854,
      "learning_rate": 9.971200191367234e-05,
      "loss": 0.0765,
      "step": 349
    },
    {
      "epoch": 0.2625656414103526,
      "grad_norm": 0.21195389330387115,
      "learning_rate": 9.970494169762117e-05,
      "loss": 0.0759,
      "step": 350
    },
    {
      "epoch": 0.2633158289572393,
      "grad_norm": 0.28137511014938354,
      "learning_rate": 9.969779624248954e-05,
      "loss": 0.0747,
      "step": 351
    },
    {
      "epoch": 0.26406601650412603,
      "grad_norm": 0.22566178441047668,
      "learning_rate": 9.969056556053116e-05,
      "loss": 0.1097,
      "step": 352
    },
    {
      "epoch": 0.26481620405101275,
      "grad_norm": 0.16110064089298248,
      "learning_rate": 9.968324966414597e-05,
      "loss": 0.0914,
      "step": 353
    },
    {
      "epoch": 0.26556639159789946,
      "grad_norm": 0.2024664431810379,
      "learning_rate": 9.967584856588e-05,
      "loss": 0.1056,
      "step": 354
    },
    {
      "epoch": 0.2663165791447862,
      "grad_norm": 0.21437528729438782,
      "learning_rate": 9.966836227842538e-05,
      "loss": 0.0681,
      "step": 355
    },
    {
      "epoch": 0.2670667666916729,
      "grad_norm": 0.14517563581466675,
      "learning_rate": 9.96607908146204e-05,
      "loss": 0.0538,
      "step": 356
    },
    {
      "epoch": 0.2678169542385596,
      "grad_norm": 0.1344057023525238,
      "learning_rate": 9.965313418744935e-05,
      "loss": 0.0672,
      "step": 357
    },
    {
      "epoch": 0.2685671417854464,
      "grad_norm": 0.08291507512331009,
      "learning_rate": 9.964539241004261e-05,
      "loss": 0.0373,
      "step": 358
    },
    {
      "epoch": 0.2693173293323331,
      "grad_norm": 0.3290295898914337,
      "learning_rate": 9.963756549567654e-05,
      "loss": 0.0939,
      "step": 359
    },
    {
      "epoch": 0.2700675168792198,
      "grad_norm": 0.20721577107906342,
      "learning_rate": 9.962965345777353e-05,
      "loss": 0.083,
      "step": 360
    },
    {
      "epoch": 0.27081770442610653,
      "grad_norm": 0.15423627197742462,
      "learning_rate": 9.962165630990196e-05,
      "loss": 0.0844,
      "step": 361
    },
    {
      "epoch": 0.27156789197299325,
      "grad_norm": 0.1754678636789322,
      "learning_rate": 9.961357406577617e-05,
      "loss": 0.0747,
      "step": 362
    },
    {
      "epoch": 0.27231807951987996,
      "grad_norm": 0.10178562998771667,
      "learning_rate": 9.960540673925636e-05,
      "loss": 0.0446,
      "step": 363
    },
    {
      "epoch": 0.2730682670667667,
      "grad_norm": 0.17141780257225037,
      "learning_rate": 9.959715434434873e-05,
      "loss": 0.0759,
      "step": 364
    },
    {
      "epoch": 0.2738184546136534,
      "grad_norm": 0.16186627745628357,
      "learning_rate": 9.958881689520531e-05,
      "loss": 0.0611,
      "step": 365
    },
    {
      "epoch": 0.2745686421605401,
      "grad_norm": 0.15801844000816345,
      "learning_rate": 9.958039440612402e-05,
      "loss": 0.0882,
      "step": 366
    },
    {
      "epoch": 0.2753188297074269,
      "grad_norm": 0.18964071571826935,
      "learning_rate": 9.957188689154859e-05,
      "loss": 0.063,
      "step": 367
    },
    {
      "epoch": 0.2760690172543136,
      "grad_norm": 0.13521847128868103,
      "learning_rate": 9.956329436606857e-05,
      "loss": 0.0784,
      "step": 368
    },
    {
      "epoch": 0.2768192048012003,
      "grad_norm": 0.1345323920249939,
      "learning_rate": 9.955461684441928e-05,
      "loss": 0.0655,
      "step": 369
    },
    {
      "epoch": 0.27756939234808703,
      "grad_norm": 0.2111639380455017,
      "learning_rate": 9.954585434148183e-05,
      "loss": 0.0848,
      "step": 370
    },
    {
      "epoch": 0.27831957989497375,
      "grad_norm": 0.14540591835975647,
      "learning_rate": 9.953700687228306e-05,
      "loss": 0.0664,
      "step": 371
    },
    {
      "epoch": 0.27906976744186046,
      "grad_norm": 0.13060738146305084,
      "learning_rate": 9.952807445199549e-05,
      "loss": 0.0599,
      "step": 372
    },
    {
      "epoch": 0.2798199549887472,
      "grad_norm": 0.12008094042539597,
      "learning_rate": 9.951905709593735e-05,
      "loss": 0.0573,
      "step": 373
    },
    {
      "epoch": 0.2805701425356339,
      "grad_norm": 0.1303444355726242,
      "learning_rate": 9.950995481957251e-05,
      "loss": 0.0783,
      "step": 374
    },
    {
      "epoch": 0.2813203300825206,
      "grad_norm": 0.12107301503419876,
      "learning_rate": 9.950076763851049e-05,
      "loss": 0.0585,
      "step": 375
    },
    {
      "epoch": 0.2820705176294073,
      "grad_norm": 0.11897249519824982,
      "learning_rate": 9.949149556850638e-05,
      "loss": 0.0649,
      "step": 376
    },
    {
      "epoch": 0.2828207051762941,
      "grad_norm": 0.16101312637329102,
      "learning_rate": 9.94821386254609e-05,
      "loss": 0.0831,
      "step": 377
    },
    {
      "epoch": 0.2835708927231808,
      "grad_norm": 0.16048072278499603,
      "learning_rate": 9.947269682542027e-05,
      "loss": 0.067,
      "step": 378
    },
    {
      "epoch": 0.2843210802700675,
      "grad_norm": 0.12512798607349396,
      "learning_rate": 9.946317018457622e-05,
      "loss": 0.0535,
      "step": 379
    },
    {
      "epoch": 0.28507126781695424,
      "grad_norm": 0.12598682940006256,
      "learning_rate": 9.945355871926605e-05,
      "loss": 0.0722,
      "step": 380
    },
    {
      "epoch": 0.28582145536384096,
      "grad_norm": 0.2111731618642807,
      "learning_rate": 9.944386244597244e-05,
      "loss": 0.0828,
      "step": 381
    },
    {
      "epoch": 0.2865716429107277,
      "grad_norm": 0.12089638411998749,
      "learning_rate": 9.943408138132357e-05,
      "loss": 0.0621,
      "step": 382
    },
    {
      "epoch": 0.2873218304576144,
      "grad_norm": 0.0753532275557518,
      "learning_rate": 9.942421554209297e-05,
      "loss": 0.0561,
      "step": 383
    },
    {
      "epoch": 0.2880720180045011,
      "grad_norm": 0.1740962713956833,
      "learning_rate": 9.94142649451996e-05,
      "loss": 0.0756,
      "step": 384
    },
    {
      "epoch": 0.2888222055513878,
      "grad_norm": 0.1652483493089676,
      "learning_rate": 9.940422960770776e-05,
      "loss": 0.0824,
      "step": 385
    },
    {
      "epoch": 0.2895723930982746,
      "grad_norm": 0.14878934621810913,
      "learning_rate": 9.939410954682706e-05,
      "loss": 0.0437,
      "step": 386
    },
    {
      "epoch": 0.2903225806451613,
      "grad_norm": 0.12823975086212158,
      "learning_rate": 9.938390477991242e-05,
      "loss": 0.0616,
      "step": 387
    },
    {
      "epoch": 0.291072768192048,
      "grad_norm": 0.11716058105230331,
      "learning_rate": 9.937361532446399e-05,
      "loss": 0.0538,
      "step": 388
    },
    {
      "epoch": 0.29182295573893474,
      "grad_norm": 0.1182708591222763,
      "learning_rate": 9.936324119812719e-05,
      "loss": 0.0578,
      "step": 389
    },
    {
      "epoch": 0.29257314328582146,
      "grad_norm": 0.16185607016086578,
      "learning_rate": 9.93527824186926e-05,
      "loss": 0.0629,
      "step": 390
    },
    {
      "epoch": 0.2933233308327082,
      "grad_norm": 0.16407489776611328,
      "learning_rate": 9.934223900409603e-05,
      "loss": 0.0475,
      "step": 391
    },
    {
      "epoch": 0.2940735183795949,
      "grad_norm": 0.1613144725561142,
      "learning_rate": 9.933161097241837e-05,
      "loss": 0.0459,
      "step": 392
    },
    {
      "epoch": 0.2948237059264816,
      "grad_norm": 0.23988468945026398,
      "learning_rate": 9.932089834188567e-05,
      "loss": 0.0712,
      "step": 393
    },
    {
      "epoch": 0.2955738934733683,
      "grad_norm": 0.15247158706188202,
      "learning_rate": 9.931010113086902e-05,
      "loss": 0.0633,
      "step": 394
    },
    {
      "epoch": 0.29632408102025504,
      "grad_norm": 0.19023269414901733,
      "learning_rate": 9.929921935788457e-05,
      "loss": 0.0715,
      "step": 395
    },
    {
      "epoch": 0.2970742685671418,
      "grad_norm": 0.15984871983528137,
      "learning_rate": 9.928825304159351e-05,
      "loss": 0.0639,
      "step": 396
    },
    {
      "epoch": 0.2978244561140285,
      "grad_norm": 0.1645509898662567,
      "learning_rate": 9.927720220080199e-05,
      "loss": 0.0586,
      "step": 397
    },
    {
      "epoch": 0.29857464366091524,
      "grad_norm": 0.1377629190683365,
      "learning_rate": 9.926606685446109e-05,
      "loss": 0.0764,
      "step": 398
    },
    {
      "epoch": 0.29932483120780196,
      "grad_norm": 0.11947868764400482,
      "learning_rate": 9.925484702166686e-05,
      "loss": 0.0512,
      "step": 399
    },
    {
      "epoch": 0.30007501875468867,
      "grad_norm": 0.10650676488876343,
      "learning_rate": 9.924354272166017e-05,
      "loss": 0.0361,
      "step": 400
    },
    {
      "epoch": 0.30007501875468867,
      "eval_loss": 0.07004823535680771,
      "eval_runtime": 2.655,
      "eval_samples_per_second": 20.339,
      "eval_steps_per_second": 5.273,
      "step": 400
    },
    {
      "epoch": 0.3008252063015754,
      "grad_norm": 0.13129869103431702,
      "learning_rate": 9.923215397382684e-05,
      "loss": 0.0503,
      "step": 401
    },
    {
      "epoch": 0.3015753938484621,
      "grad_norm": 0.11299598217010498,
      "learning_rate": 9.92206807976974e-05,
      "loss": 0.0361,
      "step": 402
    },
    {
      "epoch": 0.3023255813953488,
      "grad_norm": 0.21530146896839142,
      "learning_rate": 9.920912321294723e-05,
      "loss": 0.0583,
      "step": 403
    },
    {
      "epoch": 0.30307576894223553,
      "grad_norm": 0.10559721291065216,
      "learning_rate": 9.919748123939647e-05,
      "loss": 0.0494,
      "step": 404
    },
    {
      "epoch": 0.3038259564891223,
      "grad_norm": 0.22442474961280823,
      "learning_rate": 9.918575489700993e-05,
      "loss": 0.0808,
      "step": 405
    },
    {
      "epoch": 0.304576144036009,
      "grad_norm": 0.17437469959259033,
      "learning_rate": 9.917394420589716e-05,
      "loss": 0.0621,
      "step": 406
    },
    {
      "epoch": 0.30532633158289574,
      "grad_norm": 0.12458070367574692,
      "learning_rate": 9.916204918631231e-05,
      "loss": 0.0628,
      "step": 407
    },
    {
      "epoch": 0.30607651912978245,
      "grad_norm": 0.1025652214884758,
      "learning_rate": 9.915006985865416e-05,
      "loss": 0.0334,
      "step": 408
    },
    {
      "epoch": 0.30682670667666917,
      "grad_norm": 0.16310417652130127,
      "learning_rate": 9.913800624346612e-05,
      "loss": 0.0767,
      "step": 409
    },
    {
      "epoch": 0.3075768942235559,
      "grad_norm": 0.21571268141269684,
      "learning_rate": 9.912585836143606e-05,
      "loss": 0.0859,
      "step": 410
    },
    {
      "epoch": 0.3083270817704426,
      "grad_norm": 0.21639621257781982,
      "learning_rate": 9.911362623339642e-05,
      "loss": 0.0706,
      "step": 411
    },
    {
      "epoch": 0.3090772693173293,
      "grad_norm": 0.1594085842370987,
      "learning_rate": 9.91013098803241e-05,
      "loss": 0.0711,
      "step": 412
    },
    {
      "epoch": 0.30982745686421603,
      "grad_norm": 0.15109415352344513,
      "learning_rate": 9.908890932334042e-05,
      "loss": 0.0635,
      "step": 413
    },
    {
      "epoch": 0.31057764441110275,
      "grad_norm": 0.14152328670024872,
      "learning_rate": 9.907642458371111e-05,
      "loss": 0.0692,
      "step": 414
    },
    {
      "epoch": 0.3113278319579895,
      "grad_norm": 0.17827863991260529,
      "learning_rate": 9.906385568284629e-05,
      "loss": 0.0763,
      "step": 415
    },
    {
      "epoch": 0.31207801950487624,
      "grad_norm": 0.16184374690055847,
      "learning_rate": 9.905120264230036e-05,
      "loss": 0.0665,
      "step": 416
    },
    {
      "epoch": 0.31282820705176295,
      "grad_norm": 0.12294892221689224,
      "learning_rate": 9.903846548377206e-05,
      "loss": 0.0732,
      "step": 417
    },
    {
      "epoch": 0.31357839459864967,
      "grad_norm": 0.21045026183128357,
      "learning_rate": 9.902564422910436e-05,
      "loss": 0.0966,
      "step": 418
    },
    {
      "epoch": 0.3143285821455364,
      "grad_norm": 0.1927548348903656,
      "learning_rate": 9.901273890028444e-05,
      "loss": 0.0635,
      "step": 419
    },
    {
      "epoch": 0.3150787696924231,
      "grad_norm": 0.11646755039691925,
      "learning_rate": 9.899974951944367e-05,
      "loss": 0.0516,
      "step": 420
    },
    {
      "epoch": 0.3158289572393098,
      "grad_norm": 0.13698947429656982,
      "learning_rate": 9.898667610885757e-05,
      "loss": 0.0626,
      "step": 421
    },
    {
      "epoch": 0.31657914478619653,
      "grad_norm": 0.13424015045166016,
      "learning_rate": 9.897351869094573e-05,
      "loss": 0.0571,
      "step": 422
    },
    {
      "epoch": 0.31732933233308325,
      "grad_norm": 0.20071502029895782,
      "learning_rate": 9.896027728827185e-05,
      "loss": 0.0925,
      "step": 423
    },
    {
      "epoch": 0.31807951987997,
      "grad_norm": 0.14641085267066956,
      "learning_rate": 9.894695192354362e-05,
      "loss": 0.055,
      "step": 424
    },
    {
      "epoch": 0.31882970742685673,
      "grad_norm": 0.23519590497016907,
      "learning_rate": 9.893354261961274e-05,
      "loss": 0.0935,
      "step": 425
    },
    {
      "epoch": 0.31957989497374345,
      "grad_norm": 0.14721961319446564,
      "learning_rate": 9.892004939947482e-05,
      "loss": 0.0588,
      "step": 426
    },
    {
      "epoch": 0.32033008252063017,
      "grad_norm": 0.19134773313999176,
      "learning_rate": 9.890647228626944e-05,
      "loss": 0.084,
      "step": 427
    },
    {
      "epoch": 0.3210802700675169,
      "grad_norm": 0.18689143657684326,
      "learning_rate": 9.889281130327997e-05,
      "loss": 0.0695,
      "step": 428
    },
    {
      "epoch": 0.3218304576144036,
      "grad_norm": 0.16758248209953308,
      "learning_rate": 9.887906647393368e-05,
      "loss": 0.0738,
      "step": 429
    },
    {
      "epoch": 0.3225806451612903,
      "grad_norm": 0.13592371344566345,
      "learning_rate": 9.88652378218016e-05,
      "loss": 0.0846,
      "step": 430
    },
    {
      "epoch": 0.32333083270817703,
      "grad_norm": 0.16219943761825562,
      "learning_rate": 9.885132537059849e-05,
      "loss": 0.0805,
      "step": 431
    },
    {
      "epoch": 0.32408102025506375,
      "grad_norm": 0.10020235925912857,
      "learning_rate": 9.883732914418285e-05,
      "loss": 0.0498,
      "step": 432
    },
    {
      "epoch": 0.32483120780195046,
      "grad_norm": 0.14994826912879944,
      "learning_rate": 9.882324916655681e-05,
      "loss": 0.0618,
      "step": 433
    },
    {
      "epoch": 0.32558139534883723,
      "grad_norm": 0.23991218209266663,
      "learning_rate": 9.880908546186616e-05,
      "loss": 0.091,
      "step": 434
    },
    {
      "epoch": 0.32633158289572395,
      "grad_norm": 0.1507575809955597,
      "learning_rate": 9.879483805440027e-05,
      "loss": 0.0716,
      "step": 435
    },
    {
      "epoch": 0.32708177044261066,
      "grad_norm": 0.09811144322156906,
      "learning_rate": 9.8780506968592e-05,
      "loss": 0.0379,
      "step": 436
    },
    {
      "epoch": 0.3278319579894974,
      "grad_norm": 0.10230189561843872,
      "learning_rate": 9.876609222901781e-05,
      "loss": 0.0409,
      "step": 437
    },
    {
      "epoch": 0.3285821455363841,
      "grad_norm": 0.22452911734580994,
      "learning_rate": 9.875159386039749e-05,
      "loss": 0.0938,
      "step": 438
    },
    {
      "epoch": 0.3293323330832708,
      "grad_norm": 0.19232669472694397,
      "learning_rate": 9.873701188759438e-05,
      "loss": 0.1031,
      "step": 439
    },
    {
      "epoch": 0.3300825206301575,
      "grad_norm": 0.14251667261123657,
      "learning_rate": 9.872234633561509e-05,
      "loss": 0.0674,
      "step": 440
    },
    {
      "epoch": 0.33083270817704424,
      "grad_norm": 0.14763318002223969,
      "learning_rate": 9.87075972296096e-05,
      "loss": 0.0603,
      "step": 441
    },
    {
      "epoch": 0.33158289572393096,
      "grad_norm": 0.1664535254240036,
      "learning_rate": 9.86927645948712e-05,
      "loss": 0.0775,
      "step": 442
    },
    {
      "epoch": 0.33233308327081773,
      "grad_norm": 0.1140127182006836,
      "learning_rate": 9.867784845683637e-05,
      "loss": 0.0771,
      "step": 443
    },
    {
      "epoch": 0.33308327081770445,
      "grad_norm": 0.1934347301721573,
      "learning_rate": 9.866284884108481e-05,
      "loss": 0.0796,
      "step": 444
    },
    {
      "epoch": 0.33383345836459116,
      "grad_norm": 0.2020241767168045,
      "learning_rate": 9.864776577333941e-05,
      "loss": 0.0763,
      "step": 445
    },
    {
      "epoch": 0.3345836459114779,
      "grad_norm": 0.10327481478452682,
      "learning_rate": 9.863259927946613e-05,
      "loss": 0.0496,
      "step": 446
    },
    {
      "epoch": 0.3353338334583646,
      "grad_norm": 0.1181400716304779,
      "learning_rate": 9.861734938547405e-05,
      "loss": 0.0535,
      "step": 447
    },
    {
      "epoch": 0.3360840210052513,
      "grad_norm": 0.1940775215625763,
      "learning_rate": 9.860201611751518e-05,
      "loss": 0.0844,
      "step": 448
    },
    {
      "epoch": 0.336834208552138,
      "grad_norm": 0.12212885916233063,
      "learning_rate": 9.858659950188458e-05,
      "loss": 0.0664,
      "step": 449
    },
    {
      "epoch": 0.33758439609902474,
      "grad_norm": 0.14932742714881897,
      "learning_rate": 9.857109956502027e-05,
      "loss": 0.0786,
      "step": 450
    },
    {
      "epoch": 0.33833458364591146,
      "grad_norm": 0.15180030465126038,
      "learning_rate": 9.855551633350306e-05,
      "loss": 0.098,
      "step": 451
    },
    {
      "epoch": 0.3390847711927982,
      "grad_norm": 0.14569677412509918,
      "learning_rate": 9.853984983405668e-05,
      "loss": 0.0781,
      "step": 452
    },
    {
      "epoch": 0.33983495873968494,
      "grad_norm": 0.14552511274814606,
      "learning_rate": 9.852410009354766e-05,
      "loss": 0.0504,
      "step": 453
    },
    {
      "epoch": 0.34058514628657166,
      "grad_norm": 0.16309498250484467,
      "learning_rate": 9.850826713898521e-05,
      "loss": 0.0645,
      "step": 454
    },
    {
      "epoch": 0.3413353338334584,
      "grad_norm": 0.11566208302974701,
      "learning_rate": 9.849235099752132e-05,
      "loss": 0.0705,
      "step": 455
    },
    {
      "epoch": 0.3420855213803451,
      "grad_norm": 0.14972847700119019,
      "learning_rate": 9.847635169645058e-05,
      "loss": 0.0682,
      "step": 456
    },
    {
      "epoch": 0.3428357089272318,
      "grad_norm": 0.21967746317386627,
      "learning_rate": 9.846026926321024e-05,
      "loss": 0.0742,
      "step": 457
    },
    {
      "epoch": 0.3435858964741185,
      "grad_norm": 0.13541746139526367,
      "learning_rate": 9.844410372538006e-05,
      "loss": 0.0622,
      "step": 458
    },
    {
      "epoch": 0.34433608402100524,
      "grad_norm": 0.14925634860992432,
      "learning_rate": 9.842785511068239e-05,
      "loss": 0.0722,
      "step": 459
    },
    {
      "epoch": 0.34508627156789196,
      "grad_norm": 0.10999463498592377,
      "learning_rate": 9.841152344698197e-05,
      "loss": 0.0425,
      "step": 460
    },
    {
      "epoch": 0.34583645911477867,
      "grad_norm": 0.1686522364616394,
      "learning_rate": 9.8395108762286e-05,
      "loss": 0.0602,
      "step": 461
    },
    {
      "epoch": 0.34658664666166544,
      "grad_norm": 0.1271689385175705,
      "learning_rate": 9.837861108474404e-05,
      "loss": 0.0473,
      "step": 462
    },
    {
      "epoch": 0.34733683420855216,
      "grad_norm": 0.12710584700107574,
      "learning_rate": 9.8362030442648e-05,
      "loss": 0.0352,
      "step": 463
    },
    {
      "epoch": 0.3480870217554389,
      "grad_norm": 0.18865905702114105,
      "learning_rate": 9.834536686443204e-05,
      "loss": 0.07,
      "step": 464
    },
    {
      "epoch": 0.3488372093023256,
      "grad_norm": 0.12701815366744995,
      "learning_rate": 9.832862037867257e-05,
      "loss": 0.0504,
      "step": 465
    },
    {
      "epoch": 0.3495873968492123,
      "grad_norm": 0.26698964834213257,
      "learning_rate": 9.831179101408813e-05,
      "loss": 0.0944,
      "step": 466
    },
    {
      "epoch": 0.350337584396099,
      "grad_norm": 0.1741272509098053,
      "learning_rate": 9.829487879953946e-05,
      "loss": 0.0799,
      "step": 467
    },
    {
      "epoch": 0.35108777194298574,
      "grad_norm": 0.1799943894147873,
      "learning_rate": 9.827788376402932e-05,
      "loss": 0.0862,
      "step": 468
    },
    {
      "epoch": 0.35183795948987245,
      "grad_norm": 0.14110024273395538,
      "learning_rate": 9.826080593670253e-05,
      "loss": 0.062,
      "step": 469
    },
    {
      "epoch": 0.35258814703675917,
      "grad_norm": 0.30671846866607666,
      "learning_rate": 9.82436453468459e-05,
      "loss": 0.0911,
      "step": 470
    },
    {
      "epoch": 0.3533383345836459,
      "grad_norm": 0.2468806803226471,
      "learning_rate": 9.822640202388812e-05,
      "loss": 0.0777,
      "step": 471
    },
    {
      "epoch": 0.35408852213053266,
      "grad_norm": 0.12997083365917206,
      "learning_rate": 9.820907599739979e-05,
      "loss": 0.0566,
      "step": 472
    },
    {
      "epoch": 0.3548387096774194,
      "grad_norm": 0.16618408262729645,
      "learning_rate": 9.819166729709336e-05,
      "loss": 0.068,
      "step": 473
    },
    {
      "epoch": 0.3555888972243061,
      "grad_norm": 0.18094521760940552,
      "learning_rate": 9.817417595282304e-05,
      "loss": 0.0681,
      "step": 474
    },
    {
      "epoch": 0.3563390847711928,
      "grad_norm": 0.13062360882759094,
      "learning_rate": 9.815660199458476e-05,
      "loss": 0.0533,
      "step": 475
    },
    {
      "epoch": 0.3570892723180795,
      "grad_norm": 0.17325717210769653,
      "learning_rate": 9.81389454525161e-05,
      "loss": 0.0761,
      "step": 476
    },
    {
      "epoch": 0.35783945986496624,
      "grad_norm": 0.14805567264556885,
      "learning_rate": 9.812120635689632e-05,
      "loss": 0.0792,
      "step": 477
    },
    {
      "epoch": 0.35858964741185295,
      "grad_norm": 0.22663648426532745,
      "learning_rate": 9.810338473814621e-05,
      "loss": 0.0912,
      "step": 478
    },
    {
      "epoch": 0.35933983495873967,
      "grad_norm": 0.18868513405323029,
      "learning_rate": 9.808548062682812e-05,
      "loss": 0.0645,
      "step": 479
    },
    {
      "epoch": 0.3600900225056264,
      "grad_norm": 0.16635006666183472,
      "learning_rate": 9.80674940536458e-05,
      "loss": 0.0642,
      "step": 480
    },
    {
      "epoch": 0.36084021005251316,
      "grad_norm": 0.13103297352790833,
      "learning_rate": 9.804942504944445e-05,
      "loss": 0.0624,
      "step": 481
    },
    {
      "epoch": 0.36159039759939987,
      "grad_norm": 0.1823355257511139,
      "learning_rate": 9.803127364521067e-05,
      "loss": 0.068,
      "step": 482
    },
    {
      "epoch": 0.3623405851462866,
      "grad_norm": 0.09847483783960342,
      "learning_rate": 9.801303987207229e-05,
      "loss": 0.0557,
      "step": 483
    },
    {
      "epoch": 0.3630907726931733,
      "grad_norm": 0.20062148571014404,
      "learning_rate": 9.799472376129846e-05,
      "loss": 0.0981,
      "step": 484
    },
    {
      "epoch": 0.36384096024006,
      "grad_norm": 0.1890101432800293,
      "learning_rate": 9.79763253442995e-05,
      "loss": 0.0817,
      "step": 485
    },
    {
      "epoch": 0.36459114778694673,
      "grad_norm": 0.2274111807346344,
      "learning_rate": 9.795784465262689e-05,
      "loss": 0.0758,
      "step": 486
    },
    {
      "epoch": 0.36534133533383345,
      "grad_norm": 0.30035200715065,
      "learning_rate": 9.79392817179732e-05,
      "loss": 0.0743,
      "step": 487
    },
    {
      "epoch": 0.36609152288072017,
      "grad_norm": 0.10150797665119171,
      "learning_rate": 9.792063657217201e-05,
      "loss": 0.0486,
      "step": 488
    },
    {
      "epoch": 0.3668417104276069,
      "grad_norm": 0.10696882009506226,
      "learning_rate": 9.790190924719793e-05,
      "loss": 0.0415,
      "step": 489
    },
    {
      "epoch": 0.3675918979744936,
      "grad_norm": 0.16352054476737976,
      "learning_rate": 9.788309977516648e-05,
      "loss": 0.0831,
      "step": 490
    },
    {
      "epoch": 0.36834208552138037,
      "grad_norm": 0.15026482939720154,
      "learning_rate": 9.786420818833404e-05,
      "loss": 0.0474,
      "step": 491
    },
    {
      "epoch": 0.3690922730682671,
      "grad_norm": 0.13316163420677185,
      "learning_rate": 9.784523451909782e-05,
      "loss": 0.0525,
      "step": 492
    },
    {
      "epoch": 0.3698424606151538,
      "grad_norm": 0.19105587899684906,
      "learning_rate": 9.78261787999958e-05,
      "loss": 0.1024,
      "step": 493
    },
    {
      "epoch": 0.3705926481620405,
      "grad_norm": 0.20892846584320068,
      "learning_rate": 9.780704106370667e-05,
      "loss": 0.08,
      "step": 494
    },
    {
      "epoch": 0.37134283570892723,
      "grad_norm": 0.1441580057144165,
      "learning_rate": 9.778782134304976e-05,
      "loss": 0.057,
      "step": 495
    },
    {
      "epoch": 0.37209302325581395,
      "grad_norm": 0.10130858421325684,
      "learning_rate": 9.776851967098499e-05,
      "loss": 0.0416,
      "step": 496
    },
    {
      "epoch": 0.37284321080270066,
      "grad_norm": 0.17521536350250244,
      "learning_rate": 9.774913608061282e-05,
      "loss": 0.0666,
      "step": 497
    },
    {
      "epoch": 0.3735933983495874,
      "grad_norm": 0.18551689386367798,
      "learning_rate": 9.772967060517421e-05,
      "loss": 0.0669,
      "step": 498
    },
    {
      "epoch": 0.3743435858964741,
      "grad_norm": 0.15010853111743927,
      "learning_rate": 9.771012327805055e-05,
      "loss": 0.0507,
      "step": 499
    },
    {
      "epoch": 0.37509377344336087,
      "grad_norm": 0.18854506313800812,
      "learning_rate": 9.769049413276355e-05,
      "loss": 0.0627,
      "step": 500
    },
    {
      "epoch": 0.3758439609902476,
      "grad_norm": 0.18690386414527893,
      "learning_rate": 9.767078320297528e-05,
      "loss": 0.0773,
      "step": 501
    },
    {
      "epoch": 0.3765941485371343,
      "grad_norm": 0.19477099180221558,
      "learning_rate": 9.765099052248805e-05,
      "loss": 0.0585,
      "step": 502
    },
    {
      "epoch": 0.377344336084021,
      "grad_norm": 0.2110205888748169,
      "learning_rate": 9.763111612524434e-05,
      "loss": 0.065,
      "step": 503
    },
    {
      "epoch": 0.37809452363090773,
      "grad_norm": 0.11988997459411621,
      "learning_rate": 9.761116004532679e-05,
      "loss": 0.0523,
      "step": 504
    },
    {
      "epoch": 0.37884471117779445,
      "grad_norm": 0.1722867339849472,
      "learning_rate": 9.759112231695811e-05,
      "loss": 0.0819,
      "step": 505
    },
    {
      "epoch": 0.37959489872468116,
      "grad_norm": 0.19149507582187653,
      "learning_rate": 9.757100297450103e-05,
      "loss": 0.1036,
      "step": 506
    },
    {
      "epoch": 0.3803450862715679,
      "grad_norm": 0.14034023880958557,
      "learning_rate": 9.755080205245826e-05,
      "loss": 0.0473,
      "step": 507
    },
    {
      "epoch": 0.3810952738184546,
      "grad_norm": 0.18162596225738525,
      "learning_rate": 9.753051958547238e-05,
      "loss": 0.0597,
      "step": 508
    },
    {
      "epoch": 0.3818454613653413,
      "grad_norm": 0.2020900994539261,
      "learning_rate": 9.751015560832582e-05,
      "loss": 0.0953,
      "step": 509
    },
    {
      "epoch": 0.3825956489122281,
      "grad_norm": 0.16374067962169647,
      "learning_rate": 9.748971015594078e-05,
      "loss": 0.0602,
      "step": 510
    },
    {
      "epoch": 0.3833458364591148,
      "grad_norm": 0.15971016883850098,
      "learning_rate": 9.746918326337923e-05,
      "loss": 0.0506,
      "step": 511
    },
    {
      "epoch": 0.3840960240060015,
      "grad_norm": 0.20230558514595032,
      "learning_rate": 9.744857496584274e-05,
      "loss": 0.1119,
      "step": 512
    },
    {
      "epoch": 0.38484621155288823,
      "grad_norm": 0.18030567467212677,
      "learning_rate": 9.742788529867255e-05,
      "loss": 0.0736,
      "step": 513
    },
    {
      "epoch": 0.38559639909977494,
      "grad_norm": 0.12409501522779465,
      "learning_rate": 9.740711429734936e-05,
      "loss": 0.0677,
      "step": 514
    },
    {
      "epoch": 0.38634658664666166,
      "grad_norm": 0.19922702014446259,
      "learning_rate": 9.738626199749341e-05,
      "loss": 0.0703,
      "step": 515
    },
    {
      "epoch": 0.3870967741935484,
      "grad_norm": 0.22303518652915955,
      "learning_rate": 9.736532843486433e-05,
      "loss": 0.0985,
      "step": 516
    },
    {
      "epoch": 0.3878469617404351,
      "grad_norm": 0.259988397359848,
      "learning_rate": 9.734431364536114e-05,
      "loss": 0.0886,
      "step": 517
    },
    {
      "epoch": 0.3885971492873218,
      "grad_norm": 0.17871250212192535,
      "learning_rate": 9.732321766502213e-05,
      "loss": 0.0929,
      "step": 518
    },
    {
      "epoch": 0.3893473368342086,
      "grad_norm": 0.06613419950008392,
      "learning_rate": 9.730204053002481e-05,
      "loss": 0.0376,
      "step": 519
    },
    {
      "epoch": 0.3900975243810953,
      "grad_norm": 0.14904949069023132,
      "learning_rate": 9.728078227668588e-05,
      "loss": 0.0616,
      "step": 520
    },
    {
      "epoch": 0.390847711927982,
      "grad_norm": 0.15744099020957947,
      "learning_rate": 9.725944294146119e-05,
      "loss": 0.0696,
      "step": 521
    },
    {
      "epoch": 0.3915978994748687,
      "grad_norm": 0.19264377653598785,
      "learning_rate": 9.723802256094555e-05,
      "loss": 0.0765,
      "step": 522
    },
    {
      "epoch": 0.39234808702175544,
      "grad_norm": 0.17883890867233276,
      "learning_rate": 9.721652117187283e-05,
      "loss": 0.0666,
      "step": 523
    },
    {
      "epoch": 0.39309827456864216,
      "grad_norm": 0.1510065793991089,
      "learning_rate": 9.71949388111158e-05,
      "loss": 0.0796,
      "step": 524
    },
    {
      "epoch": 0.3938484621155289,
      "grad_norm": 0.16549058258533478,
      "learning_rate": 9.717327551568608e-05,
      "loss": 0.0761,
      "step": 525
    },
    {
      "epoch": 0.3945986496624156,
      "grad_norm": 0.10932381451129913,
      "learning_rate": 9.715153132273407e-05,
      "loss": 0.058,
      "step": 526
    },
    {
      "epoch": 0.3953488372093023,
      "grad_norm": 0.15327946841716766,
      "learning_rate": 9.712970626954893e-05,
      "loss": 0.0592,
      "step": 527
    },
    {
      "epoch": 0.396099024756189,
      "grad_norm": 0.1345522105693817,
      "learning_rate": 9.71078003935585e-05,
      "loss": 0.0752,
      "step": 528
    },
    {
      "epoch": 0.3968492123030758,
      "grad_norm": 0.10842501372098923,
      "learning_rate": 9.708581373232917e-05,
      "loss": 0.0607,
      "step": 529
    },
    {
      "epoch": 0.3975993998499625,
      "grad_norm": 0.15361064672470093,
      "learning_rate": 9.70637463235659e-05,
      "loss": 0.0648,
      "step": 530
    },
    {
      "epoch": 0.3983495873968492,
      "grad_norm": 0.13886478543281555,
      "learning_rate": 9.704159820511214e-05,
      "loss": 0.0888,
      "step": 531
    },
    {
      "epoch": 0.39909977494373594,
      "grad_norm": 0.15999573469161987,
      "learning_rate": 9.701936941494971e-05,
      "loss": 0.0651,
      "step": 532
    },
    {
      "epoch": 0.39984996249062266,
      "grad_norm": 0.23427529633045197,
      "learning_rate": 9.699705999119882e-05,
      "loss": 0.0672,
      "step": 533
    },
    {
      "epoch": 0.4006001500375094,
      "grad_norm": 0.10031108558177948,
      "learning_rate": 9.697466997211793e-05,
      "loss": 0.0414,
      "step": 534
    },
    {
      "epoch": 0.4013503375843961,
      "grad_norm": 0.1497843712568283,
      "learning_rate": 9.69521993961037e-05,
      "loss": 0.0784,
      "step": 535
    },
    {
      "epoch": 0.4021005251312828,
      "grad_norm": 0.14574141800403595,
      "learning_rate": 9.692964830169098e-05,
      "loss": 0.0683,
      "step": 536
    },
    {
      "epoch": 0.4028507126781695,
      "grad_norm": 0.12160151451826096,
      "learning_rate": 9.690701672755266e-05,
      "loss": 0.0692,
      "step": 537
    },
    {
      "epoch": 0.4036009002250563,
      "grad_norm": 0.15294575691223145,
      "learning_rate": 9.688430471249967e-05,
      "loss": 0.0645,
      "step": 538
    },
    {
      "epoch": 0.404351087771943,
      "grad_norm": 0.12379523366689682,
      "learning_rate": 9.686151229548088e-05,
      "loss": 0.0446,
      "step": 539
    },
    {
      "epoch": 0.4051012753188297,
      "grad_norm": 0.11460956931114197,
      "learning_rate": 9.683863951558301e-05,
      "loss": 0.068,
      "step": 540
    },
    {
      "epoch": 0.40585146286571644,
      "grad_norm": 0.11864083260297775,
      "learning_rate": 9.681568641203068e-05,
      "loss": 0.0263,
      "step": 541
    },
    {
      "epoch": 0.40660165041260315,
      "grad_norm": 0.1245616003870964,
      "learning_rate": 9.679265302418615e-05,
      "loss": 0.054,
      "step": 542
    },
    {
      "epoch": 0.40735183795948987,
      "grad_norm": 0.16922646760940552,
      "learning_rate": 9.676953939154945e-05,
      "loss": 0.0824,
      "step": 543
    },
    {
      "epoch": 0.4081020255063766,
      "grad_norm": 0.24694877862930298,
      "learning_rate": 9.674634555375817e-05,
      "loss": 0.0666,
      "step": 544
    },
    {
      "epoch": 0.4088522130532633,
      "grad_norm": 0.1859944462776184,
      "learning_rate": 9.672307155058744e-05,
      "loss": 0.0959,
      "step": 545
    },
    {
      "epoch": 0.40960240060015,
      "grad_norm": 0.21233129501342773,
      "learning_rate": 9.669971742194992e-05,
      "loss": 0.1242,
      "step": 546
    },
    {
      "epoch": 0.41035258814703673,
      "grad_norm": 0.19170650839805603,
      "learning_rate": 9.667628320789562e-05,
      "loss": 0.0795,
      "step": 547
    },
    {
      "epoch": 0.4111027756939235,
      "grad_norm": 0.16862432658672333,
      "learning_rate": 9.665276894861188e-05,
      "loss": 0.0835,
      "step": 548
    },
    {
      "epoch": 0.4118529632408102,
      "grad_norm": 0.12009342759847641,
      "learning_rate": 9.66291746844234e-05,
      "loss": 0.0483,
      "step": 549
    },
    {
      "epoch": 0.41260315078769694,
      "grad_norm": 0.16073642671108246,
      "learning_rate": 9.660550045579199e-05,
      "loss": 0.0631,
      "step": 550
    },
    {
      "epoch": 0.41335333833458365,
      "grad_norm": 0.2920999825000763,
      "learning_rate": 9.65817463033166e-05,
      "loss": 0.097,
      "step": 551
    },
    {
      "epoch": 0.41410352588147037,
      "grad_norm": 0.14352647960186005,
      "learning_rate": 9.655791226773331e-05,
      "loss": 0.0654,
      "step": 552
    },
    {
      "epoch": 0.4148537134283571,
      "grad_norm": 0.16379228234291077,
      "learning_rate": 9.65339983899151e-05,
      "loss": 0.0574,
      "step": 553
    },
    {
      "epoch": 0.4156039009752438,
      "grad_norm": 0.15645262598991394,
      "learning_rate": 9.651000471087193e-05,
      "loss": 0.0556,
      "step": 554
    },
    {
      "epoch": 0.4163540885221305,
      "grad_norm": 0.1435929536819458,
      "learning_rate": 9.64859312717506e-05,
      "loss": 0.0597,
      "step": 555
    },
    {
      "epoch": 0.41710427606901723,
      "grad_norm": 0.12999999523162842,
      "learning_rate": 9.64617781138347e-05,
      "loss": 0.0598,
      "step": 556
    },
    {
      "epoch": 0.41785446361590395,
      "grad_norm": 0.18979759514331818,
      "learning_rate": 9.643754527854451e-05,
      "loss": 0.0857,
      "step": 557
    },
    {
      "epoch": 0.4186046511627907,
      "grad_norm": 0.18712787330150604,
      "learning_rate": 9.641323280743693e-05,
      "loss": 0.0692,
      "step": 558
    },
    {
      "epoch": 0.41935483870967744,
      "grad_norm": 0.12155728787183762,
      "learning_rate": 9.638884074220548e-05,
      "loss": 0.0728,
      "step": 559
    },
    {
      "epoch": 0.42010502625656415,
      "grad_norm": 0.16611698269844055,
      "learning_rate": 9.636436912468015e-05,
      "loss": 0.0585,
      "step": 560
    },
    {
      "epoch": 0.42085521380345087,
      "grad_norm": 0.1172047033905983,
      "learning_rate": 9.633981799682735e-05,
      "loss": 0.082,
      "step": 561
    },
    {
      "epoch": 0.4216054013503376,
      "grad_norm": 0.09718162566423416,
      "learning_rate": 9.631518740074985e-05,
      "loss": 0.0281,
      "step": 562
    },
    {
      "epoch": 0.4223555888972243,
      "grad_norm": 0.10215144604444504,
      "learning_rate": 9.629047737868669e-05,
      "loss": 0.0544,
      "step": 563
    },
    {
      "epoch": 0.423105776444111,
      "grad_norm": 0.1456746608018875,
      "learning_rate": 9.626568797301311e-05,
      "loss": 0.0673,
      "step": 564
    },
    {
      "epoch": 0.42385596399099773,
      "grad_norm": 0.1138940155506134,
      "learning_rate": 9.624081922624053e-05,
      "loss": 0.0516,
      "step": 565
    },
    {
      "epoch": 0.42460615153788445,
      "grad_norm": 0.2256423681974411,
      "learning_rate": 9.621587118101638e-05,
      "loss": 0.0729,
      "step": 566
    },
    {
      "epoch": 0.4253563390847712,
      "grad_norm": 0.159426748752594,
      "learning_rate": 9.619084388012412e-05,
      "loss": 0.0676,
      "step": 567
    },
    {
      "epoch": 0.42610652663165793,
      "grad_norm": 0.21540193259716034,
      "learning_rate": 9.616573736648308e-05,
      "loss": 0.0879,
      "step": 568
    },
    {
      "epoch": 0.42685671417854465,
      "grad_norm": 0.12239127606153488,
      "learning_rate": 9.61405516831485e-05,
      "loss": 0.0502,
      "step": 569
    },
    {
      "epoch": 0.42760690172543137,
      "grad_norm": 0.15514914691448212,
      "learning_rate": 9.61152868733113e-05,
      "loss": 0.0592,
      "step": 570
    },
    {
      "epoch": 0.4283570892723181,
      "grad_norm": 0.10405272245407104,
      "learning_rate": 9.608994298029818e-05,
      "loss": 0.0546,
      "step": 571
    },
    {
      "epoch": 0.4291072768192048,
      "grad_norm": 0.1854027807712555,
      "learning_rate": 9.60645200475714e-05,
      "loss": 0.0711,
      "step": 572
    },
    {
      "epoch": 0.4298574643660915,
      "grad_norm": 0.14749664068222046,
      "learning_rate": 9.603901811872877e-05,
      "loss": 0.0709,
      "step": 573
    },
    {
      "epoch": 0.43060765191297823,
      "grad_norm": 0.1849459558725357,
      "learning_rate": 9.601343723750363e-05,
      "loss": 0.0826,
      "step": 574
    },
    {
      "epoch": 0.43135783945986494,
      "grad_norm": 0.10557045042514801,
      "learning_rate": 9.598777744776464e-05,
      "loss": 0.0506,
      "step": 575
    },
    {
      "epoch": 0.43210802700675166,
      "grad_norm": 0.13303524255752563,
      "learning_rate": 9.596203879351582e-05,
      "loss": 0.0526,
      "step": 576
    },
    {
      "epoch": 0.43285821455363843,
      "grad_norm": 0.1565026342868805,
      "learning_rate": 9.593622131889643e-05,
      "loss": 0.099,
      "step": 577
    },
    {
      "epoch": 0.43360840210052515,
      "grad_norm": 0.12550555169582367,
      "learning_rate": 9.591032506818089e-05,
      "loss": 0.0525,
      "step": 578
    },
    {
      "epoch": 0.43435858964741186,
      "grad_norm": 0.14403165876865387,
      "learning_rate": 9.588435008577873e-05,
      "loss": 0.0613,
      "step": 579
    },
    {
      "epoch": 0.4351087771942986,
      "grad_norm": 0.144205242395401,
      "learning_rate": 9.585829641623448e-05,
      "loss": 0.0596,
      "step": 580
    },
    {
      "epoch": 0.4358589647411853,
      "grad_norm": 0.14015857875347137,
      "learning_rate": 9.583216410422762e-05,
      "loss": 0.0773,
      "step": 581
    },
    {
      "epoch": 0.436609152288072,
      "grad_norm": 0.15716828405857086,
      "learning_rate": 9.580595319457249e-05,
      "loss": 0.0545,
      "step": 582
    },
    {
      "epoch": 0.4373593398349587,
      "grad_norm": 0.25865277647972107,
      "learning_rate": 9.577966373221823e-05,
      "loss": 0.0885,
      "step": 583
    },
    {
      "epoch": 0.43810952738184544,
      "grad_norm": 0.21027599275112152,
      "learning_rate": 9.575329576224868e-05,
      "loss": 0.0704,
      "step": 584
    },
    {
      "epoch": 0.43885971492873216,
      "grad_norm": 0.1413678526878357,
      "learning_rate": 9.572684932988227e-05,
      "loss": 0.0464,
      "step": 585
    },
    {
      "epoch": 0.43960990247561893,
      "grad_norm": 0.1475217193365097,
      "learning_rate": 9.570032448047208e-05,
      "loss": 0.0648,
      "step": 586
    },
    {
      "epoch": 0.44036009002250565,
      "grad_norm": 0.1679753214120865,
      "learning_rate": 9.567372125950559e-05,
      "loss": 0.0583,
      "step": 587
    },
    {
      "epoch": 0.44111027756939236,
      "grad_norm": 0.13862156867980957,
      "learning_rate": 9.564703971260472e-05,
      "loss": 0.0648,
      "step": 588
    },
    {
      "epoch": 0.4418604651162791,
      "grad_norm": 0.12006936222314835,
      "learning_rate": 9.562027988552567e-05,
      "loss": 0.0477,
      "step": 589
    },
    {
      "epoch": 0.4426106526631658,
      "grad_norm": 0.2421911656856537,
      "learning_rate": 9.559344182415891e-05,
      "loss": 0.0996,
      "step": 590
    },
    {
      "epoch": 0.4433608402100525,
      "grad_norm": 0.16158467531204224,
      "learning_rate": 9.55665255745291e-05,
      "loss": 0.0735,
      "step": 591
    },
    {
      "epoch": 0.4441110277569392,
      "grad_norm": 0.17239175736904144,
      "learning_rate": 9.553953118279496e-05,
      "loss": 0.0827,
      "step": 592
    },
    {
      "epoch": 0.44486121530382594,
      "grad_norm": 0.2204049825668335,
      "learning_rate": 9.551245869524916e-05,
      "loss": 0.0696,
      "step": 593
    },
    {
      "epoch": 0.44561140285071266,
      "grad_norm": 0.11255641281604767,
      "learning_rate": 9.54853081583184e-05,
      "loss": 0.048,
      "step": 594
    },
    {
      "epoch": 0.4463615903975994,
      "grad_norm": 0.11511281132698059,
      "learning_rate": 9.545807961856317e-05,
      "loss": 0.0569,
      "step": 595
    },
    {
      "epoch": 0.44711177794448614,
      "grad_norm": 0.18176564574241638,
      "learning_rate": 9.543077312267773e-05,
      "loss": 0.0637,
      "step": 596
    },
    {
      "epoch": 0.44786196549137286,
      "grad_norm": 0.1667564958333969,
      "learning_rate": 9.540338871749002e-05,
      "loss": 0.0808,
      "step": 597
    },
    {
      "epoch": 0.4486121530382596,
      "grad_norm": 0.2754107117652893,
      "learning_rate": 9.537592644996162e-05,
      "loss": 0.102,
      "step": 598
    },
    {
      "epoch": 0.4493623405851463,
      "grad_norm": 0.10601028054952621,
      "learning_rate": 9.534838636718759e-05,
      "loss": 0.0475,
      "step": 599
    },
    {
      "epoch": 0.450112528132033,
      "grad_norm": 0.13959158957004547,
      "learning_rate": 9.532076851639649e-05,
      "loss": 0.0649,
      "step": 600
    },
    {
      "epoch": 0.450112528132033,
      "eval_loss": 0.07095802575349808,
      "eval_runtime": 2.656,
      "eval_samples_per_second": 20.332,
      "eval_steps_per_second": 5.271,
      "step": 600
    },
    {
      "epoch": 0.4508627156789197,
      "grad_norm": 0.08408086746931076,
      "learning_rate": 9.529307294495018e-05,
      "loss": 0.0446,
      "step": 601
    },
    {
      "epoch": 0.45161290322580644,
      "grad_norm": 0.1700592041015625,
      "learning_rate": 9.526529970034386e-05,
      "loss": 0.1001,
      "step": 602
    },
    {
      "epoch": 0.45236309077269315,
      "grad_norm": 0.16827765107154846,
      "learning_rate": 9.52374488302059e-05,
      "loss": 0.0726,
      "step": 603
    },
    {
      "epoch": 0.45311327831957987,
      "grad_norm": 0.1284192055463791,
      "learning_rate": 9.52095203822978e-05,
      "loss": 0.0514,
      "step": 604
    },
    {
      "epoch": 0.45386346586646664,
      "grad_norm": 0.1511625349521637,
      "learning_rate": 9.518151440451411e-05,
      "loss": 0.0658,
      "step": 605
    },
    {
      "epoch": 0.45461365341335336,
      "grad_norm": 0.14977985620498657,
      "learning_rate": 9.515343094488232e-05,
      "loss": 0.0655,
      "step": 606
    },
    {
      "epoch": 0.4553638409602401,
      "grad_norm": 0.20794422924518585,
      "learning_rate": 9.51252700515628e-05,
      "loss": 0.0732,
      "step": 607
    },
    {
      "epoch": 0.4561140285071268,
      "grad_norm": 0.18329943716526031,
      "learning_rate": 9.509703177284869e-05,
      "loss": 0.0909,
      "step": 608
    },
    {
      "epoch": 0.4568642160540135,
      "grad_norm": 0.19080418348312378,
      "learning_rate": 9.506871615716587e-05,
      "loss": 0.0964,
      "step": 609
    },
    {
      "epoch": 0.4576144036009002,
      "grad_norm": 0.15063661336898804,
      "learning_rate": 9.504032325307284e-05,
      "loss": 0.0554,
      "step": 610
    },
    {
      "epoch": 0.45836459114778694,
      "grad_norm": 0.18503394722938538,
      "learning_rate": 9.501185310926062e-05,
      "loss": 0.0725,
      "step": 611
    },
    {
      "epoch": 0.45911477869467365,
      "grad_norm": 0.16494551301002502,
      "learning_rate": 9.498330577455273e-05,
      "loss": 0.0653,
      "step": 612
    },
    {
      "epoch": 0.45986496624156037,
      "grad_norm": 0.14555247128009796,
      "learning_rate": 9.495468129790499e-05,
      "loss": 0.062,
      "step": 613
    },
    {
      "epoch": 0.4606151537884471,
      "grad_norm": 0.13288968801498413,
      "learning_rate": 9.49259797284056e-05,
      "loss": 0.0678,
      "step": 614
    },
    {
      "epoch": 0.46136534133533386,
      "grad_norm": 0.17785833775997162,
      "learning_rate": 9.489720111527492e-05,
      "loss": 0.1018,
      "step": 615
    },
    {
      "epoch": 0.46211552888222057,
      "grad_norm": 0.156355082988739,
      "learning_rate": 9.486834550786543e-05,
      "loss": 0.0651,
      "step": 616
    },
    {
      "epoch": 0.4628657164291073,
      "grad_norm": 0.09945828467607498,
      "learning_rate": 9.483941295566165e-05,
      "loss": 0.0455,
      "step": 617
    },
    {
      "epoch": 0.463615903975994,
      "grad_norm": 0.15181989967823029,
      "learning_rate": 9.481040350828006e-05,
      "loss": 0.0414,
      "step": 618
    },
    {
      "epoch": 0.4643660915228807,
      "grad_norm": 0.1600968986749649,
      "learning_rate": 9.4781317215469e-05,
      "loss": 0.0601,
      "step": 619
    },
    {
      "epoch": 0.46511627906976744,
      "grad_norm": 0.14700649678707123,
      "learning_rate": 9.475215412710864e-05,
      "loss": 0.0588,
      "step": 620
    },
    {
      "epoch": 0.46586646661665415,
      "grad_norm": 0.11616702377796173,
      "learning_rate": 9.472291429321075e-05,
      "loss": 0.0521,
      "step": 621
    },
    {
      "epoch": 0.46661665416354087,
      "grad_norm": 0.1925794631242752,
      "learning_rate": 9.469359776391879e-05,
      "loss": 0.0885,
      "step": 622
    },
    {
      "epoch": 0.4673668417104276,
      "grad_norm": 0.22271235287189484,
      "learning_rate": 9.466420458950773e-05,
      "loss": 0.1039,
      "step": 623
    },
    {
      "epoch": 0.46811702925731435,
      "grad_norm": 0.13130241632461548,
      "learning_rate": 9.463473482038395e-05,
      "loss": 0.0556,
      "step": 624
    },
    {
      "epoch": 0.46886721680420107,
      "grad_norm": 0.20446011424064636,
      "learning_rate": 9.46051885070852e-05,
      "loss": 0.0842,
      "step": 625
    },
    {
      "epoch": 0.4696174043510878,
      "grad_norm": 0.17276065051555634,
      "learning_rate": 9.457556570028052e-05,
      "loss": 0.0651,
      "step": 626
    },
    {
      "epoch": 0.4703675918979745,
      "grad_norm": 0.14631524682044983,
      "learning_rate": 9.454586645077011e-05,
      "loss": 0.0638,
      "step": 627
    },
    {
      "epoch": 0.4711177794448612,
      "grad_norm": 0.3121795058250427,
      "learning_rate": 9.451609080948522e-05,
      "loss": 0.1204,
      "step": 628
    },
    {
      "epoch": 0.47186796699174793,
      "grad_norm": 0.12264740467071533,
      "learning_rate": 9.448623882748817e-05,
      "loss": 0.0624,
      "step": 629
    },
    {
      "epoch": 0.47261815453863465,
      "grad_norm": 0.09743326157331467,
      "learning_rate": 9.445631055597217e-05,
      "loss": 0.06,
      "step": 630
    },
    {
      "epoch": 0.47336834208552137,
      "grad_norm": 0.18968532979488373,
      "learning_rate": 9.442630604626126e-05,
      "loss": 0.1011,
      "step": 631
    },
    {
      "epoch": 0.4741185296324081,
      "grad_norm": 0.10481397807598114,
      "learning_rate": 9.43962253498102e-05,
      "loss": 0.0544,
      "step": 632
    },
    {
      "epoch": 0.4748687171792948,
      "grad_norm": 0.12469401955604553,
      "learning_rate": 9.436606851820444e-05,
      "loss": 0.0551,
      "step": 633
    },
    {
      "epoch": 0.47561890472618157,
      "grad_norm": 0.16171766817569733,
      "learning_rate": 9.433583560315999e-05,
      "loss": 0.073,
      "step": 634
    },
    {
      "epoch": 0.4763690922730683,
      "grad_norm": 0.12446840107440948,
      "learning_rate": 9.430552665652328e-05,
      "loss": 0.0601,
      "step": 635
    },
    {
      "epoch": 0.477119279819955,
      "grad_norm": 0.1368139535188675,
      "learning_rate": 9.427514173027121e-05,
      "loss": 0.0547,
      "step": 636
    },
    {
      "epoch": 0.4778694673668417,
      "grad_norm": 0.1330532282590866,
      "learning_rate": 9.424468087651092e-05,
      "loss": 0.0536,
      "step": 637
    },
    {
      "epoch": 0.47861965491372843,
      "grad_norm": 0.13870249688625336,
      "learning_rate": 9.421414414747978e-05,
      "loss": 0.0439,
      "step": 638
    },
    {
      "epoch": 0.47936984246061515,
      "grad_norm": 0.11789939552545547,
      "learning_rate": 9.418353159554526e-05,
      "loss": 0.0601,
      "step": 639
    },
    {
      "epoch": 0.48012003000750186,
      "grad_norm": 0.1677863597869873,
      "learning_rate": 9.415284327320489e-05,
      "loss": 0.0574,
      "step": 640
    },
    {
      "epoch": 0.4808702175543886,
      "grad_norm": 0.12923505902290344,
      "learning_rate": 9.41220792330861e-05,
      "loss": 0.0646,
      "step": 641
    },
    {
      "epoch": 0.4816204051012753,
      "grad_norm": 0.16255703568458557,
      "learning_rate": 9.40912395279462e-05,
      "loss": 0.1003,
      "step": 642
    },
    {
      "epoch": 0.48237059264816207,
      "grad_norm": 0.12949885427951813,
      "learning_rate": 9.406032421067224e-05,
      "loss": 0.0509,
      "step": 643
    },
    {
      "epoch": 0.4831207801950488,
      "grad_norm": 0.1707848310470581,
      "learning_rate": 9.402933333428097e-05,
      "loss": 0.0737,
      "step": 644
    },
    {
      "epoch": 0.4838709677419355,
      "grad_norm": 0.14833003282546997,
      "learning_rate": 9.399826695191868e-05,
      "loss": 0.0637,
      "step": 645
    },
    {
      "epoch": 0.4846211552888222,
      "grad_norm": 0.140524223446846,
      "learning_rate": 9.396712511686114e-05,
      "loss": 0.0627,
      "step": 646
    },
    {
      "epoch": 0.48537134283570893,
      "grad_norm": 0.13231976330280304,
      "learning_rate": 9.393590788251354e-05,
      "loss": 0.0627,
      "step": 647
    },
    {
      "epoch": 0.48612153038259565,
      "grad_norm": 0.1312929391860962,
      "learning_rate": 9.390461530241037e-05,
      "loss": 0.0447,
      "step": 648
    },
    {
      "epoch": 0.48687171792948236,
      "grad_norm": 0.2141118049621582,
      "learning_rate": 9.38732474302153e-05,
      "loss": 0.0775,
      "step": 649
    },
    {
      "epoch": 0.4876219054763691,
      "grad_norm": 0.19238243997097015,
      "learning_rate": 9.384180431972119e-05,
      "loss": 0.0575,
      "step": 650
    },
    {
      "epoch": 0.4883720930232558,
      "grad_norm": 0.13290603458881378,
      "learning_rate": 9.381028602484984e-05,
      "loss": 0.0645,
      "step": 651
    },
    {
      "epoch": 0.4891222805701425,
      "grad_norm": 0.1605457365512848,
      "learning_rate": 9.377869259965202e-05,
      "loss": 0.0623,
      "step": 652
    },
    {
      "epoch": 0.4898724681170293,
      "grad_norm": 0.17835864424705505,
      "learning_rate": 9.374702409830736e-05,
      "loss": 0.0992,
      "step": 653
    },
    {
      "epoch": 0.490622655663916,
      "grad_norm": 0.2935710549354553,
      "learning_rate": 9.37152805751242e-05,
      "loss": 0.0668,
      "step": 654
    },
    {
      "epoch": 0.4913728432108027,
      "grad_norm": 0.19382020831108093,
      "learning_rate": 9.36834620845396e-05,
      "loss": 0.0645,
      "step": 655
    },
    {
      "epoch": 0.49212303075768943,
      "grad_norm": 0.1444290429353714,
      "learning_rate": 9.365156868111908e-05,
      "loss": 0.0612,
      "step": 656
    },
    {
      "epoch": 0.49287321830457614,
      "grad_norm": 0.12895061075687408,
      "learning_rate": 9.361960041955672e-05,
      "loss": 0.0468,
      "step": 657
    },
    {
      "epoch": 0.49362340585146286,
      "grad_norm": 0.13367299735546112,
      "learning_rate": 9.358755735467494e-05,
      "loss": 0.0663,
      "step": 658
    },
    {
      "epoch": 0.4943735933983496,
      "grad_norm": 0.17381970584392548,
      "learning_rate": 9.355543954142446e-05,
      "loss": 0.0857,
      "step": 659
    },
    {
      "epoch": 0.4951237809452363,
      "grad_norm": 0.1889047920703888,
      "learning_rate": 9.352324703488412e-05,
      "loss": 0.0968,
      "step": 660
    },
    {
      "epoch": 0.495873968492123,
      "grad_norm": 0.16304278373718262,
      "learning_rate": 9.349097989026093e-05,
      "loss": 0.0682,
      "step": 661
    },
    {
      "epoch": 0.4966241560390098,
      "grad_norm": 0.19829104840755463,
      "learning_rate": 9.345863816288985e-05,
      "loss": 0.0684,
      "step": 662
    },
    {
      "epoch": 0.4973743435858965,
      "grad_norm": 0.15854620933532715,
      "learning_rate": 9.342622190823378e-05,
      "loss": 0.0664,
      "step": 663
    },
    {
      "epoch": 0.4981245311327832,
      "grad_norm": 0.25643864274024963,
      "learning_rate": 9.339373118188338e-05,
      "loss": 0.0805,
      "step": 664
    },
    {
      "epoch": 0.4988747186796699,
      "grad_norm": 0.13145574927330017,
      "learning_rate": 9.336116603955707e-05,
      "loss": 0.0526,
      "step": 665
    },
    {
      "epoch": 0.49962490622655664,
      "grad_norm": 0.13013754785060883,
      "learning_rate": 9.332852653710084e-05,
      "loss": 0.0505,
      "step": 666
    },
    {
      "epoch": 0.5003750937734434,
      "grad_norm": 0.1300058811903,
      "learning_rate": 9.329581273048822e-05,
      "loss": 0.048,
      "step": 667
    },
    {
      "epoch": 0.5011252813203301,
      "grad_norm": 0.15247949957847595,
      "learning_rate": 9.32630246758202e-05,
      "loss": 0.0657,
      "step": 668
    },
    {
      "epoch": 0.5018754688672168,
      "grad_norm": 0.1410437822341919,
      "learning_rate": 9.323016242932504e-05,
      "loss": 0.0936,
      "step": 669
    },
    {
      "epoch": 0.5026256564141035,
      "grad_norm": 0.12135988473892212,
      "learning_rate": 9.319722604735825e-05,
      "loss": 0.0641,
      "step": 670
    },
    {
      "epoch": 0.5033758439609902,
      "grad_norm": 0.15133972465991974,
      "learning_rate": 9.31642155864025e-05,
      "loss": 0.0675,
      "step": 671
    },
    {
      "epoch": 0.5041260315078769,
      "grad_norm": 0.1304636150598526,
      "learning_rate": 9.313113110306748e-05,
      "loss": 0.0479,
      "step": 672
    },
    {
      "epoch": 0.5048762190547637,
      "grad_norm": 0.1356845498085022,
      "learning_rate": 9.309797265408979e-05,
      "loss": 0.0588,
      "step": 673
    },
    {
      "epoch": 0.5056264066016504,
      "grad_norm": 0.1584395468235016,
      "learning_rate": 9.306474029633294e-05,
      "loss": 0.0543,
      "step": 674
    },
    {
      "epoch": 0.5063765941485371,
      "grad_norm": 0.16862989962100983,
      "learning_rate": 9.303143408678716e-05,
      "loss": 0.0693,
      "step": 675
    },
    {
      "epoch": 0.5071267816954238,
      "grad_norm": 0.12757883965969086,
      "learning_rate": 9.299805408256928e-05,
      "loss": 0.0376,
      "step": 676
    },
    {
      "epoch": 0.5078769692423106,
      "grad_norm": 0.1599583476781845,
      "learning_rate": 9.296460034092274e-05,
      "loss": 0.0593,
      "step": 677
    },
    {
      "epoch": 0.5086271567891973,
      "grad_norm": 0.12639591097831726,
      "learning_rate": 9.293107291921741e-05,
      "loss": 0.0675,
      "step": 678
    },
    {
      "epoch": 0.5093773443360841,
      "grad_norm": 0.10506059974431992,
      "learning_rate": 9.289747187494952e-05,
      "loss": 0.0365,
      "step": 679
    },
    {
      "epoch": 0.5101275318829708,
      "grad_norm": 0.1758994609117508,
      "learning_rate": 9.286379726574155e-05,
      "loss": 0.0484,
      "step": 680
    },
    {
      "epoch": 0.5108777194298575,
      "grad_norm": 0.1311517059803009,
      "learning_rate": 9.283004914934215e-05,
      "loss": 0.0524,
      "step": 681
    },
    {
      "epoch": 0.5116279069767442,
      "grad_norm": 0.1752699762582779,
      "learning_rate": 9.2796227583626e-05,
      "loss": 0.0897,
      "step": 682
    },
    {
      "epoch": 0.5123780945236309,
      "grad_norm": 0.15679432451725006,
      "learning_rate": 9.276233262659375e-05,
      "loss": 0.058,
      "step": 683
    },
    {
      "epoch": 0.5131282820705176,
      "grad_norm": 0.15536212921142578,
      "learning_rate": 9.272836433637193e-05,
      "loss": 0.0739,
      "step": 684
    },
    {
      "epoch": 0.5138784696174044,
      "grad_norm": 0.15193812549114227,
      "learning_rate": 9.269432277121281e-05,
      "loss": 0.055,
      "step": 685
    },
    {
      "epoch": 0.5146286571642911,
      "grad_norm": 0.1312589794397354,
      "learning_rate": 9.266020798949433e-05,
      "loss": 0.0463,
      "step": 686
    },
    {
      "epoch": 0.5153788447111778,
      "grad_norm": 0.16843152046203613,
      "learning_rate": 9.262602004971996e-05,
      "loss": 0.065,
      "step": 687
    },
    {
      "epoch": 0.5161290322580645,
      "grad_norm": 0.12552760541439056,
      "learning_rate": 9.259175901051867e-05,
      "loss": 0.0533,
      "step": 688
    },
    {
      "epoch": 0.5168792198049512,
      "grad_norm": 0.18700753152370453,
      "learning_rate": 9.255742493064474e-05,
      "loss": 0.0739,
      "step": 689
    },
    {
      "epoch": 0.5176294073518379,
      "grad_norm": 0.24483560025691986,
      "learning_rate": 9.252301786897776e-05,
      "loss": 0.0679,
      "step": 690
    },
    {
      "epoch": 0.5183795948987246,
      "grad_norm": 0.20147421956062317,
      "learning_rate": 9.248853788452247e-05,
      "loss": 0.0753,
      "step": 691
    },
    {
      "epoch": 0.5191297824456114,
      "grad_norm": 0.18182440102100372,
      "learning_rate": 9.24539850364086e-05,
      "loss": 0.0444,
      "step": 692
    },
    {
      "epoch": 0.5198799699924981,
      "grad_norm": 0.24294641613960266,
      "learning_rate": 9.241935938389093e-05,
      "loss": 0.0712,
      "step": 693
    },
    {
      "epoch": 0.5206301575393848,
      "grad_norm": 0.1470489203929901,
      "learning_rate": 9.238466098634902e-05,
      "loss": 0.0557,
      "step": 694
    },
    {
      "epoch": 0.5213803450862715,
      "grad_norm": 0.19487878680229187,
      "learning_rate": 9.234988990328719e-05,
      "loss": 0.0826,
      "step": 695
    },
    {
      "epoch": 0.5221305326331583,
      "grad_norm": 0.13695645332336426,
      "learning_rate": 9.231504619433445e-05,
      "loss": 0.0662,
      "step": 696
    },
    {
      "epoch": 0.522880720180045,
      "grad_norm": 0.2282414436340332,
      "learning_rate": 9.228012991924433e-05,
      "loss": 0.0939,
      "step": 697
    },
    {
      "epoch": 0.5236309077269318,
      "grad_norm": 0.154418483376503,
      "learning_rate": 9.224514113789477e-05,
      "loss": 0.0622,
      "step": 698
    },
    {
      "epoch": 0.5243810952738185,
      "grad_norm": 0.2042827010154724,
      "learning_rate": 9.221007991028814e-05,
      "loss": 0.0698,
      "step": 699
    },
    {
      "epoch": 0.5251312828207052,
      "grad_norm": 0.1465548425912857,
      "learning_rate": 9.217494629655094e-05,
      "loss": 0.0468,
      "step": 700
    },
    {
      "epoch": 0.5258814703675919,
      "grad_norm": 0.12249326705932617,
      "learning_rate": 9.213974035693389e-05,
      "loss": 0.0453,
      "step": 701
    },
    {
      "epoch": 0.5266316579144786,
      "grad_norm": 0.11008257418870926,
      "learning_rate": 9.21044621518117e-05,
      "loss": 0.0506,
      "step": 702
    },
    {
      "epoch": 0.5273818454613654,
      "grad_norm": 0.19281499087810516,
      "learning_rate": 9.206911174168301e-05,
      "loss": 0.0797,
      "step": 703
    },
    {
      "epoch": 0.5281320330082521,
      "grad_norm": 0.16175544261932373,
      "learning_rate": 9.20336891871703e-05,
      "loss": 0.0649,
      "step": 704
    },
    {
      "epoch": 0.5288822205551388,
      "grad_norm": 0.15684905648231506,
      "learning_rate": 9.199819454901977e-05,
      "loss": 0.0677,
      "step": 705
    },
    {
      "epoch": 0.5296324081020255,
      "grad_norm": 0.2674317955970764,
      "learning_rate": 9.196262788810121e-05,
      "loss": 0.1118,
      "step": 706
    },
    {
      "epoch": 0.5303825956489122,
      "grad_norm": 0.1423470377922058,
      "learning_rate": 9.192698926540795e-05,
      "loss": 0.0764,
      "step": 707
    },
    {
      "epoch": 0.5311327831957989,
      "grad_norm": 0.16944824159145355,
      "learning_rate": 9.189127874205674e-05,
      "loss": 0.0689,
      "step": 708
    },
    {
      "epoch": 0.5318829707426856,
      "grad_norm": 0.19155418872833252,
      "learning_rate": 9.185549637928758e-05,
      "loss": 0.0654,
      "step": 709
    },
    {
      "epoch": 0.5326331582895724,
      "grad_norm": 0.16647079586982727,
      "learning_rate": 9.181964223846371e-05,
      "loss": 0.0961,
      "step": 710
    },
    {
      "epoch": 0.5333833458364591,
      "grad_norm": 0.1647406965494156,
      "learning_rate": 9.178371638107146e-05,
      "loss": 0.0745,
      "step": 711
    },
    {
      "epoch": 0.5341335333833458,
      "grad_norm": 0.17723067104816437,
      "learning_rate": 9.174771886872011e-05,
      "loss": 0.0557,
      "step": 712
    },
    {
      "epoch": 0.5348837209302325,
      "grad_norm": 0.206720232963562,
      "learning_rate": 9.17116497631419e-05,
      "loss": 0.0733,
      "step": 713
    },
    {
      "epoch": 0.5356339084771192,
      "grad_norm": 0.16216231882572174,
      "learning_rate": 9.167550912619173e-05,
      "loss": 0.0576,
      "step": 714
    },
    {
      "epoch": 0.536384096024006,
      "grad_norm": 0.11692114174365997,
      "learning_rate": 9.16392970198473e-05,
      "loss": 0.0526,
      "step": 715
    },
    {
      "epoch": 0.5371342835708928,
      "grad_norm": 0.13061407208442688,
      "learning_rate": 9.160301350620875e-05,
      "loss": 0.0599,
      "step": 716
    },
    {
      "epoch": 0.5378844711177795,
      "grad_norm": 0.10589703917503357,
      "learning_rate": 9.156665864749876e-05,
      "loss": 0.0689,
      "step": 717
    },
    {
      "epoch": 0.5386346586646662,
      "grad_norm": 0.12243220210075378,
      "learning_rate": 9.153023250606234e-05,
      "loss": 0.0537,
      "step": 718
    },
    {
      "epoch": 0.5393848462115529,
      "grad_norm": 0.18423284590244293,
      "learning_rate": 9.14937351443667e-05,
      "loss": 0.1045,
      "step": 719
    },
    {
      "epoch": 0.5401350337584396,
      "grad_norm": 0.17368818819522858,
      "learning_rate": 9.145716662500126e-05,
      "loss": 0.0825,
      "step": 720
    },
    {
      "epoch": 0.5408852213053263,
      "grad_norm": 0.1663954257965088,
      "learning_rate": 9.142052701067741e-05,
      "loss": 0.1131,
      "step": 721
    },
    {
      "epoch": 0.5416354088522131,
      "grad_norm": 0.19684909284114838,
      "learning_rate": 9.13838163642285e-05,
      "loss": 0.0925,
      "step": 722
    },
    {
      "epoch": 0.5423855963990998,
      "grad_norm": 0.18679098784923553,
      "learning_rate": 9.134703474860963e-05,
      "loss": 0.0513,
      "step": 723
    },
    {
      "epoch": 0.5431357839459865,
      "grad_norm": 0.1188245639204979,
      "learning_rate": 9.13101822268977e-05,
      "loss": 0.0636,
      "step": 724
    },
    {
      "epoch": 0.5438859714928732,
      "grad_norm": 0.25207948684692383,
      "learning_rate": 9.127325886229115e-05,
      "loss": 0.0907,
      "step": 725
    },
    {
      "epoch": 0.5446361590397599,
      "grad_norm": 0.13716183602809906,
      "learning_rate": 9.123626471810988e-05,
      "loss": 0.0533,
      "step": 726
    },
    {
      "epoch": 0.5453863465866466,
      "grad_norm": 0.3036738634109497,
      "learning_rate": 9.119919985779521e-05,
      "loss": 0.1033,
      "step": 727
    },
    {
      "epoch": 0.5461365341335334,
      "grad_norm": 0.18584197759628296,
      "learning_rate": 9.116206434490976e-05,
      "loss": 0.1002,
      "step": 728
    },
    {
      "epoch": 0.5468867216804201,
      "grad_norm": 0.21791420876979828,
      "learning_rate": 9.112485824313726e-05,
      "loss": 0.0672,
      "step": 729
    },
    {
      "epoch": 0.5476369092273068,
      "grad_norm": 0.29374638199806213,
      "learning_rate": 9.10875816162825e-05,
      "loss": 0.0671,
      "step": 730
    },
    {
      "epoch": 0.5483870967741935,
      "grad_norm": 0.12101375311613083,
      "learning_rate": 9.105023452827121e-05,
      "loss": 0.0599,
      "step": 731
    },
    {
      "epoch": 0.5491372843210802,
      "grad_norm": 0.13263867795467377,
      "learning_rate": 9.101281704315002e-05,
      "loss": 0.0619,
      "step": 732
    },
    {
      "epoch": 0.5498874718679669,
      "grad_norm": 0.17979863286018372,
      "learning_rate": 9.097532922508619e-05,
      "loss": 0.0669,
      "step": 733
    },
    {
      "epoch": 0.5506376594148538,
      "grad_norm": 0.14049032330513,
      "learning_rate": 9.093777113836765e-05,
      "loss": 0.0643,
      "step": 734
    },
    {
      "epoch": 0.5513878469617405,
      "grad_norm": 0.1863878220319748,
      "learning_rate": 9.090014284740283e-05,
      "loss": 0.0684,
      "step": 735
    },
    {
      "epoch": 0.5521380345086272,
      "grad_norm": 0.13345694541931152,
      "learning_rate": 9.086244441672052e-05,
      "loss": 0.061,
      "step": 736
    },
    {
      "epoch": 0.5528882220555139,
      "grad_norm": 0.2068844437599182,
      "learning_rate": 9.082467591096982e-05,
      "loss": 0.0801,
      "step": 737
    },
    {
      "epoch": 0.5536384096024006,
      "grad_norm": 0.12110719829797745,
      "learning_rate": 9.078683739492002e-05,
      "loss": 0.0769,
      "step": 738
    },
    {
      "epoch": 0.5543885971492873,
      "grad_norm": 0.2093079388141632,
      "learning_rate": 9.074892893346043e-05,
      "loss": 0.0782,
      "step": 739
    },
    {
      "epoch": 0.5551387846961741,
      "grad_norm": 0.12246458232402802,
      "learning_rate": 9.071095059160035e-05,
      "loss": 0.0521,
      "step": 740
    },
    {
      "epoch": 0.5558889722430608,
      "grad_norm": 0.115948386490345,
      "learning_rate": 9.067290243446887e-05,
      "loss": 0.0497,
      "step": 741
    },
    {
      "epoch": 0.5566391597899475,
      "grad_norm": 0.16100749373435974,
      "learning_rate": 9.063478452731484e-05,
      "loss": 0.0615,
      "step": 742
    },
    {
      "epoch": 0.5573893473368342,
      "grad_norm": 0.1973392218351364,
      "learning_rate": 9.059659693550673e-05,
      "loss": 0.0736,
      "step": 743
    },
    {
      "epoch": 0.5581395348837209,
      "grad_norm": 0.12596265971660614,
      "learning_rate": 9.055833972453249e-05,
      "loss": 0.0725,
      "step": 744
    },
    {
      "epoch": 0.5588897224306076,
      "grad_norm": 0.22205737233161926,
      "learning_rate": 9.052001295999947e-05,
      "loss": 0.1457,
      "step": 745
    },
    {
      "epoch": 0.5596399099774944,
      "grad_norm": 0.15573537349700928,
      "learning_rate": 9.048161670763429e-05,
      "loss": 0.0668,
      "step": 746
    },
    {
      "epoch": 0.5603900975243811,
      "grad_norm": 0.1514582484960556,
      "learning_rate": 9.044315103328276e-05,
      "loss": 0.0452,
      "step": 747
    },
    {
      "epoch": 0.5611402850712678,
      "grad_norm": 0.12557372450828552,
      "learning_rate": 9.04046160029097e-05,
      "loss": 0.0483,
      "step": 748
    },
    {
      "epoch": 0.5618904726181545,
      "grad_norm": 0.188478484749794,
      "learning_rate": 9.036601168259893e-05,
      "loss": 0.0584,
      "step": 749
    },
    {
      "epoch": 0.5626406601650412,
      "grad_norm": 0.18197283148765564,
      "learning_rate": 9.032733813855301e-05,
      "loss": 0.078,
      "step": 750
    },
    {
      "epoch": 0.5633908477119279,
      "grad_norm": 0.20598235726356506,
      "learning_rate": 9.02885954370933e-05,
      "loss": 0.0575,
      "step": 751
    },
    {
      "epoch": 0.5641410352588146,
      "grad_norm": 0.2001180350780487,
      "learning_rate": 9.02497836446597e-05,
      "loss": 0.0912,
      "step": 752
    },
    {
      "epoch": 0.5648912228057015,
      "grad_norm": 0.12836728990077972,
      "learning_rate": 9.021090282781059e-05,
      "loss": 0.0592,
      "step": 753
    },
    {
      "epoch": 0.5656414103525882,
      "grad_norm": 0.15549710392951965,
      "learning_rate": 9.01719530532228e-05,
      "loss": 0.0955,
      "step": 754
    },
    {
      "epoch": 0.5663915978994749,
      "grad_norm": 0.10522754490375519,
      "learning_rate": 9.01329343876913e-05,
      "loss": 0.0417,
      "step": 755
    },
    {
      "epoch": 0.5671417854463616,
      "grad_norm": 0.17558157444000244,
      "learning_rate": 9.009384689812928e-05,
      "loss": 0.0686,
      "step": 756
    },
    {
      "epoch": 0.5678919729932483,
      "grad_norm": 0.129380464553833,
      "learning_rate": 9.005469065156795e-05,
      "loss": 0.0614,
      "step": 757
    },
    {
      "epoch": 0.568642160540135,
      "grad_norm": 0.2937568426132202,
      "learning_rate": 9.00154657151564e-05,
      "loss": 0.1039,
      "step": 758
    },
    {
      "epoch": 0.5693923480870218,
      "grad_norm": 0.10850602388381958,
      "learning_rate": 8.997617215616154e-05,
      "loss": 0.0518,
      "step": 759
    },
    {
      "epoch": 0.5701425356339085,
      "grad_norm": 0.11563457548618317,
      "learning_rate": 8.993681004196797e-05,
      "loss": 0.0513,
      "step": 760
    },
    {
      "epoch": 0.5708927231807952,
      "grad_norm": 0.12018301337957382,
      "learning_rate": 8.989737944007781e-05,
      "loss": 0.0354,
      "step": 761
    },
    {
      "epoch": 0.5716429107276819,
      "grad_norm": 0.16475611925125122,
      "learning_rate": 8.985788041811068e-05,
      "loss": 0.0801,
      "step": 762
    },
    {
      "epoch": 0.5723930982745686,
      "grad_norm": 0.1512751430273056,
      "learning_rate": 8.981831304380348e-05,
      "loss": 0.0735,
      "step": 763
    },
    {
      "epoch": 0.5731432858214554,
      "grad_norm": 0.12887828052043915,
      "learning_rate": 8.97786773850104e-05,
      "loss": 0.0511,
      "step": 764
    },
    {
      "epoch": 0.5738934733683421,
      "grad_norm": 0.11312253773212433,
      "learning_rate": 8.973897350970269e-05,
      "loss": 0.0501,
      "step": 765
    },
    {
      "epoch": 0.5746436609152288,
      "grad_norm": 0.20769958198070526,
      "learning_rate": 8.969920148596857e-05,
      "loss": 0.1016,
      "step": 766
    },
    {
      "epoch": 0.5753938484621155,
      "grad_norm": 0.15098562836647034,
      "learning_rate": 8.965936138201314e-05,
      "loss": 0.0648,
      "step": 767
    },
    {
      "epoch": 0.5761440360090022,
      "grad_norm": 0.16101424396038055,
      "learning_rate": 8.961945326615829e-05,
      "loss": 0.0829,
      "step": 768
    },
    {
      "epoch": 0.5768942235558889,
      "grad_norm": 0.14836345613002777,
      "learning_rate": 8.957947720684246e-05,
      "loss": 0.0691,
      "step": 769
    },
    {
      "epoch": 0.5776444111027756,
      "grad_norm": 0.19855999946594238,
      "learning_rate": 8.953943327262066e-05,
      "loss": 0.0712,
      "step": 770
    },
    {
      "epoch": 0.5783945986496624,
      "grad_norm": 0.14603960514068604,
      "learning_rate": 8.949932153216434e-05,
      "loss": 0.0685,
      "step": 771
    },
    {
      "epoch": 0.5791447861965492,
      "grad_norm": 0.09962477535009384,
      "learning_rate": 8.945914205426116e-05,
      "loss": 0.0453,
      "step": 772
    },
    {
      "epoch": 0.5798949737434359,
      "grad_norm": 0.17665497958660126,
      "learning_rate": 8.941889490781494e-05,
      "loss": 0.0738,
      "step": 773
    },
    {
      "epoch": 0.5806451612903226,
      "grad_norm": 0.16767451167106628,
      "learning_rate": 8.937858016184563e-05,
      "loss": 0.0766,
      "step": 774
    },
    {
      "epoch": 0.5813953488372093,
      "grad_norm": 0.14359025657176971,
      "learning_rate": 8.933819788548899e-05,
      "loss": 0.0755,
      "step": 775
    },
    {
      "epoch": 0.582145536384096,
      "grad_norm": 0.07473455369472504,
      "learning_rate": 8.92977481479967e-05,
      "loss": 0.0429,
      "step": 776
    },
    {
      "epoch": 0.5828957239309828,
      "grad_norm": 0.18343903124332428,
      "learning_rate": 8.925723101873603e-05,
      "loss": 0.0764,
      "step": 777
    },
    {
      "epoch": 0.5836459114778695,
      "grad_norm": 0.11543784290552139,
      "learning_rate": 8.92166465671899e-05,
      "loss": 0.0609,
      "step": 778
    },
    {
      "epoch": 0.5843960990247562,
      "grad_norm": 0.20404362678527832,
      "learning_rate": 8.917599486295664e-05,
      "loss": 0.0951,
      "step": 779
    },
    {
      "epoch": 0.5851462865716429,
      "grad_norm": 0.16599202156066895,
      "learning_rate": 8.913527597574991e-05,
      "loss": 0.0814,
      "step": 780
    },
    {
      "epoch": 0.5858964741185296,
      "grad_norm": 0.11812645941972733,
      "learning_rate": 8.90944899753986e-05,
      "loss": 0.042,
      "step": 781
    },
    {
      "epoch": 0.5866466616654163,
      "grad_norm": 0.11141446232795715,
      "learning_rate": 8.905363693184668e-05,
      "loss": 0.0387,
      "step": 782
    },
    {
      "epoch": 0.5873968492123031,
      "grad_norm": 0.22815898060798645,
      "learning_rate": 8.901271691515309e-05,
      "loss": 0.0869,
      "step": 783
    },
    {
      "epoch": 0.5881470367591898,
      "grad_norm": 0.09209627658128738,
      "learning_rate": 8.897172999549165e-05,
      "loss": 0.0434,
      "step": 784
    },
    {
      "epoch": 0.5888972243060765,
      "grad_norm": 0.12892037630081177,
      "learning_rate": 8.893067624315088e-05,
      "loss": 0.0413,
      "step": 785
    },
    {
      "epoch": 0.5896474118529632,
      "grad_norm": 0.16496118903160095,
      "learning_rate": 8.888955572853392e-05,
      "loss": 0.0676,
      "step": 786
    },
    {
      "epoch": 0.5903975993998499,
      "grad_norm": 0.12853941321372986,
      "learning_rate": 8.884836852215841e-05,
      "loss": 0.0605,
      "step": 787
    },
    {
      "epoch": 0.5911477869467366,
      "grad_norm": 0.1408003568649292,
      "learning_rate": 8.880711469465635e-05,
      "loss": 0.0572,
      "step": 788
    },
    {
      "epoch": 0.5918979744936234,
      "grad_norm": 0.1078130304813385,
      "learning_rate": 8.876579431677398e-05,
      "loss": 0.0444,
      "step": 789
    },
    {
      "epoch": 0.5926481620405101,
      "grad_norm": 0.1855546236038208,
      "learning_rate": 8.87244074593717e-05,
      "loss": 0.0624,
      "step": 790
    },
    {
      "epoch": 0.5933983495873969,
      "grad_norm": 0.15005631744861603,
      "learning_rate": 8.868295419342389e-05,
      "loss": 0.0554,
      "step": 791
    },
    {
      "epoch": 0.5941485371342836,
      "grad_norm": 0.18375401198863983,
      "learning_rate": 8.86414345900188e-05,
      "loss": 0.1016,
      "step": 792
    },
    {
      "epoch": 0.5948987246811703,
      "grad_norm": 0.1501947045326233,
      "learning_rate": 8.859984872035849e-05,
      "loss": 0.0645,
      "step": 793
    },
    {
      "epoch": 0.595648912228057,
      "grad_norm": 0.15210017561912537,
      "learning_rate": 8.85581966557586e-05,
      "loss": 0.0739,
      "step": 794
    },
    {
      "epoch": 0.5963990997749438,
      "grad_norm": 0.1773383766412735,
      "learning_rate": 8.851647846764835e-05,
      "loss": 0.0686,
      "step": 795
    },
    {
      "epoch": 0.5971492873218305,
      "grad_norm": 0.23530487716197968,
      "learning_rate": 8.847469422757031e-05,
      "loss": 0.0664,
      "step": 796
    },
    {
      "epoch": 0.5978994748687172,
      "grad_norm": 0.10172255337238312,
      "learning_rate": 8.843284400718033e-05,
      "loss": 0.0416,
      "step": 797
    },
    {
      "epoch": 0.5986496624156039,
      "grad_norm": 0.14675600826740265,
      "learning_rate": 8.839092787824743e-05,
      "loss": 0.0651,
      "step": 798
    },
    {
      "epoch": 0.5993998499624906,
      "grad_norm": 0.09715414047241211,
      "learning_rate": 8.834894591265364e-05,
      "loss": 0.0453,
      "step": 799
    },
    {
      "epoch": 0.6001500375093773,
      "grad_norm": 0.11148321628570557,
      "learning_rate": 8.830689818239388e-05,
      "loss": 0.0468,
      "step": 800
    },
    {
      "epoch": 0.6001500375093773,
      "eval_loss": 0.07001598179340363,
      "eval_runtime": 2.6621,
      "eval_samples_per_second": 20.285,
      "eval_steps_per_second": 5.259,
      "step": 800
    },
    {
      "epoch": 0.6009002250562641,
      "grad_norm": 0.11450861394405365,
      "learning_rate": 8.826478475957589e-05,
      "loss": 0.0401,
      "step": 801
    },
    {
      "epoch": 0.6016504126031508,
      "grad_norm": 0.10207905620336533,
      "learning_rate": 8.822260571642005e-05,
      "loss": 0.0511,
      "step": 802
    },
    {
      "epoch": 0.6024006001500375,
      "grad_norm": 0.09535292536020279,
      "learning_rate": 8.818036112525924e-05,
      "loss": 0.0613,
      "step": 803
    },
    {
      "epoch": 0.6031507876969242,
      "grad_norm": 0.18865959346294403,
      "learning_rate": 8.813805105853879e-05,
      "loss": 0.0744,
      "step": 804
    },
    {
      "epoch": 0.6039009752438109,
      "grad_norm": 0.11314369738101959,
      "learning_rate": 8.809567558881628e-05,
      "loss": 0.0449,
      "step": 805
    },
    {
      "epoch": 0.6046511627906976,
      "grad_norm": 0.1716671735048294,
      "learning_rate": 8.805323478876149e-05,
      "loss": 0.0764,
      "step": 806
    },
    {
      "epoch": 0.6054013503375844,
      "grad_norm": 0.15436431765556335,
      "learning_rate": 8.80107287311562e-05,
      "loss": 0.0697,
      "step": 807
    },
    {
      "epoch": 0.6061515378844711,
      "grad_norm": 0.13632182776927948,
      "learning_rate": 8.796815748889413e-05,
      "loss": 0.0456,
      "step": 808
    },
    {
      "epoch": 0.6069017254313578,
      "grad_norm": 0.10250575095415115,
      "learning_rate": 8.792552113498073e-05,
      "loss": 0.0502,
      "step": 809
    },
    {
      "epoch": 0.6076519129782446,
      "grad_norm": 0.22113870084285736,
      "learning_rate": 8.788281974253318e-05,
      "loss": 0.1038,
      "step": 810
    },
    {
      "epoch": 0.6084021005251313,
      "grad_norm": 0.14202985167503357,
      "learning_rate": 8.784005338478017e-05,
      "loss": 0.073,
      "step": 811
    },
    {
      "epoch": 0.609152288072018,
      "grad_norm": 0.15260817110538483,
      "learning_rate": 8.779722213506178e-05,
      "loss": 0.0769,
      "step": 812
    },
    {
      "epoch": 0.6099024756189048,
      "grad_norm": 0.20531697571277618,
      "learning_rate": 8.775432606682937e-05,
      "loss": 0.109,
      "step": 813
    },
    {
      "epoch": 0.6106526631657915,
      "grad_norm": 0.19820401072502136,
      "learning_rate": 8.77113652536455e-05,
      "loss": 0.0571,
      "step": 814
    },
    {
      "epoch": 0.6114028507126782,
      "grad_norm": 0.23314431309700012,
      "learning_rate": 8.766833976918371e-05,
      "loss": 0.0665,
      "step": 815
    },
    {
      "epoch": 0.6121530382595649,
      "grad_norm": 0.20076413452625275,
      "learning_rate": 8.76252496872285e-05,
      "loss": 0.076,
      "step": 816
    },
    {
      "epoch": 0.6129032258064516,
      "grad_norm": 0.11270924657583237,
      "learning_rate": 8.758209508167508e-05,
      "loss": 0.0573,
      "step": 817
    },
    {
      "epoch": 0.6136534133533383,
      "grad_norm": 0.16007684171199799,
      "learning_rate": 8.753887602652937e-05,
      "loss": 0.0711,
      "step": 818
    },
    {
      "epoch": 0.614403600900225,
      "grad_norm": 0.15355923771858215,
      "learning_rate": 8.74955925959078e-05,
      "loss": 0.0617,
      "step": 819
    },
    {
      "epoch": 0.6151537884471118,
      "grad_norm": 0.1625223010778427,
      "learning_rate": 8.745224486403718e-05,
      "loss": 0.0726,
      "step": 820
    },
    {
      "epoch": 0.6159039759939985,
      "grad_norm": 0.15932884812355042,
      "learning_rate": 8.74088329052546e-05,
      "loss": 0.0691,
      "step": 821
    },
    {
      "epoch": 0.6166541635408852,
      "grad_norm": 0.17074544727802277,
      "learning_rate": 8.73653567940073e-05,
      "loss": 0.0882,
      "step": 822
    },
    {
      "epoch": 0.6174043510877719,
      "grad_norm": 0.12639515101909637,
      "learning_rate": 8.732181660485252e-05,
      "loss": 0.0447,
      "step": 823
    },
    {
      "epoch": 0.6181545386346586,
      "grad_norm": 0.12108887732028961,
      "learning_rate": 8.727821241245742e-05,
      "loss": 0.0424,
      "step": 824
    },
    {
      "epoch": 0.6189047261815454,
      "grad_norm": 0.1238684207201004,
      "learning_rate": 8.723454429159888e-05,
      "loss": 0.0439,
      "step": 825
    },
    {
      "epoch": 0.6196549137284321,
      "grad_norm": 0.1405922919511795,
      "learning_rate": 8.719081231716341e-05,
      "loss": 0.0576,
      "step": 826
    },
    {
      "epoch": 0.6204051012753188,
      "grad_norm": 0.1311212182044983,
      "learning_rate": 8.714701656414708e-05,
      "loss": 0.0449,
      "step": 827
    },
    {
      "epoch": 0.6211552888222055,
      "grad_norm": 0.19940714538097382,
      "learning_rate": 8.710315710765526e-05,
      "loss": 0.077,
      "step": 828
    },
    {
      "epoch": 0.6219054763690923,
      "grad_norm": 0.1685575693845749,
      "learning_rate": 8.705923402290261e-05,
      "loss": 0.0493,
      "step": 829
    },
    {
      "epoch": 0.622655663915979,
      "grad_norm": 0.10744970291852951,
      "learning_rate": 8.701524738521291e-05,
      "loss": 0.0442,
      "step": 830
    },
    {
      "epoch": 0.6234058514628658,
      "grad_norm": 0.18083858489990234,
      "learning_rate": 8.697119727001887e-05,
      "loss": 0.0812,
      "step": 831
    },
    {
      "epoch": 0.6241560390097525,
      "grad_norm": 0.1295631229877472,
      "learning_rate": 8.692708375286217e-05,
      "loss": 0.0601,
      "step": 832
    },
    {
      "epoch": 0.6249062265566392,
      "grad_norm": 0.1965394765138626,
      "learning_rate": 8.688290690939307e-05,
      "loss": 0.0686,
      "step": 833
    },
    {
      "epoch": 0.6256564141035259,
      "grad_norm": 0.12678660452365875,
      "learning_rate": 8.683866681537054e-05,
      "loss": 0.0511,
      "step": 834
    },
    {
      "epoch": 0.6264066016504126,
      "grad_norm": 0.08941943943500519,
      "learning_rate": 8.679436354666202e-05,
      "loss": 0.0408,
      "step": 835
    },
    {
      "epoch": 0.6271567891972993,
      "grad_norm": 0.15167708694934845,
      "learning_rate": 8.67499971792432e-05,
      "loss": 0.0657,
      "step": 836
    },
    {
      "epoch": 0.627906976744186,
      "grad_norm": 0.15622377395629883,
      "learning_rate": 8.670556778919805e-05,
      "loss": 0.0667,
      "step": 837
    },
    {
      "epoch": 0.6286571642910728,
      "grad_norm": 0.4224050045013428,
      "learning_rate": 8.666107545271859e-05,
      "loss": 0.1789,
      "step": 838
    },
    {
      "epoch": 0.6294073518379595,
      "grad_norm": 0.24056875705718994,
      "learning_rate": 8.661652024610482e-05,
      "loss": 0.0992,
      "step": 839
    },
    {
      "epoch": 0.6301575393848462,
      "grad_norm": 0.10566816478967667,
      "learning_rate": 8.657190224576453e-05,
      "loss": 0.0384,
      "step": 840
    },
    {
      "epoch": 0.6309077269317329,
      "grad_norm": 0.11838135123252869,
      "learning_rate": 8.652722152821318e-05,
      "loss": 0.0555,
      "step": 841
    },
    {
      "epoch": 0.6316579144786196,
      "grad_norm": 0.10628216713666916,
      "learning_rate": 8.64824781700738e-05,
      "loss": 0.0542,
      "step": 842
    },
    {
      "epoch": 0.6324081020255063,
      "grad_norm": 0.17062321305274963,
      "learning_rate": 8.643767224807685e-05,
      "loss": 0.0836,
      "step": 843
    },
    {
      "epoch": 0.6331582895723931,
      "grad_norm": 0.1308564692735672,
      "learning_rate": 8.639280383906008e-05,
      "loss": 0.0835,
      "step": 844
    },
    {
      "epoch": 0.6339084771192798,
      "grad_norm": 0.2822250723838806,
      "learning_rate": 8.634787301996839e-05,
      "loss": 0.0664,
      "step": 845
    },
    {
      "epoch": 0.6346586646661665,
      "grad_norm": 0.1532263457775116,
      "learning_rate": 8.630287986785368e-05,
      "loss": 0.0753,
      "step": 846
    },
    {
      "epoch": 0.6354088522130532,
      "grad_norm": 0.1545627862215042,
      "learning_rate": 8.625782445987483e-05,
      "loss": 0.0623,
      "step": 847
    },
    {
      "epoch": 0.63615903975994,
      "grad_norm": 0.1597648561000824,
      "learning_rate": 8.621270687329738e-05,
      "loss": 0.0584,
      "step": 848
    },
    {
      "epoch": 0.6369092273068268,
      "grad_norm": 0.1512480229139328,
      "learning_rate": 8.616752718549359e-05,
      "loss": 0.0859,
      "step": 849
    },
    {
      "epoch": 0.6376594148537135,
      "grad_norm": 0.11838315427303314,
      "learning_rate": 8.612228547394215e-05,
      "loss": 0.0662,
      "step": 850
    },
    {
      "epoch": 0.6384096024006002,
      "grad_norm": 0.1927739381790161,
      "learning_rate": 8.607698181622814e-05,
      "loss": 0.0883,
      "step": 851
    },
    {
      "epoch": 0.6391597899474869,
      "grad_norm": 0.1669844686985016,
      "learning_rate": 8.603161629004287e-05,
      "loss": 0.0869,
      "step": 852
    },
    {
      "epoch": 0.6399099774943736,
      "grad_norm": 0.16778261959552765,
      "learning_rate": 8.598618897318375e-05,
      "loss": 0.0879,
      "step": 853
    },
    {
      "epoch": 0.6406601650412603,
      "grad_norm": 0.2178831249475479,
      "learning_rate": 8.594069994355419e-05,
      "loss": 0.0635,
      "step": 854
    },
    {
      "epoch": 0.641410352588147,
      "grad_norm": 0.19614070653915405,
      "learning_rate": 8.589514927916336e-05,
      "loss": 0.0836,
      "step": 855
    },
    {
      "epoch": 0.6421605401350338,
      "grad_norm": 0.12226955592632294,
      "learning_rate": 8.584953705812615e-05,
      "loss": 0.0714,
      "step": 856
    },
    {
      "epoch": 0.6429107276819205,
      "grad_norm": 0.1600448340177536,
      "learning_rate": 8.580386335866308e-05,
      "loss": 0.0753,
      "step": 857
    },
    {
      "epoch": 0.6436609152288072,
      "grad_norm": 0.12808331847190857,
      "learning_rate": 8.575812825909998e-05,
      "loss": 0.0583,
      "step": 858
    },
    {
      "epoch": 0.6444111027756939,
      "grad_norm": 0.13148951530456543,
      "learning_rate": 8.57123318378681e-05,
      "loss": 0.0575,
      "step": 859
    },
    {
      "epoch": 0.6451612903225806,
      "grad_norm": 0.13673394918441772,
      "learning_rate": 8.566647417350378e-05,
      "loss": 0.064,
      "step": 860
    },
    {
      "epoch": 0.6459114778694673,
      "grad_norm": 0.8954272866249084,
      "learning_rate": 8.562055534464838e-05,
      "loss": 0.0658,
      "step": 861
    },
    {
      "epoch": 0.6466616654163541,
      "grad_norm": 0.1643034815788269,
      "learning_rate": 8.557457543004819e-05,
      "loss": 0.0753,
      "step": 862
    },
    {
      "epoch": 0.6474118529632408,
      "grad_norm": 0.1936601996421814,
      "learning_rate": 8.552853450855422e-05,
      "loss": 0.0955,
      "step": 863
    },
    {
      "epoch": 0.6481620405101275,
      "grad_norm": 0.11946239322423935,
      "learning_rate": 8.548243265912213e-05,
      "loss": 0.0711,
      "step": 864
    },
    {
      "epoch": 0.6489122280570142,
      "grad_norm": 0.13554371893405914,
      "learning_rate": 8.543626996081202e-05,
      "loss": 0.0569,
      "step": 865
    },
    {
      "epoch": 0.6496624156039009,
      "grad_norm": 0.1824364960193634,
      "learning_rate": 8.539004649278841e-05,
      "loss": 0.0683,
      "step": 866
    },
    {
      "epoch": 0.6504126031507877,
      "grad_norm": 0.11970564723014832,
      "learning_rate": 8.534376233432e-05,
      "loss": 0.0503,
      "step": 867
    },
    {
      "epoch": 0.6511627906976745,
      "grad_norm": 0.16899152100086212,
      "learning_rate": 8.529741756477953e-05,
      "loss": 0.0775,
      "step": 868
    },
    {
      "epoch": 0.6519129782445612,
      "grad_norm": 0.13145393133163452,
      "learning_rate": 8.525101226364374e-05,
      "loss": 0.0784,
      "step": 869
    },
    {
      "epoch": 0.6526631657914479,
      "grad_norm": 0.11247047781944275,
      "learning_rate": 8.520454651049313e-05,
      "loss": 0.0493,
      "step": 870
    },
    {
      "epoch": 0.6534133533383346,
      "grad_norm": 0.1390233337879181,
      "learning_rate": 8.51580203850119e-05,
      "loss": 0.0507,
      "step": 871
    },
    {
      "epoch": 0.6541635408852213,
      "grad_norm": 0.14679768681526184,
      "learning_rate": 8.511143396698781e-05,
      "loss": 0.0698,
      "step": 872
    },
    {
      "epoch": 0.654913728432108,
      "grad_norm": 0.3411470949649811,
      "learning_rate": 8.506478733631193e-05,
      "loss": 0.1122,
      "step": 873
    },
    {
      "epoch": 0.6556639159789948,
      "grad_norm": 0.12128949165344238,
      "learning_rate": 8.501808057297866e-05,
      "loss": 0.0519,
      "step": 874
    },
    {
      "epoch": 0.6564141035258815,
      "grad_norm": 0.12452667206525803,
      "learning_rate": 8.49713137570855e-05,
      "loss": 0.0549,
      "step": 875
    },
    {
      "epoch": 0.6571642910727682,
      "grad_norm": 0.12882107496261597,
      "learning_rate": 8.492448696883292e-05,
      "loss": 0.0627,
      "step": 876
    },
    {
      "epoch": 0.6579144786196549,
      "grad_norm": 0.26537570357322693,
      "learning_rate": 8.487760028852427e-05,
      "loss": 0.0543,
      "step": 877
    },
    {
      "epoch": 0.6586646661665416,
      "grad_norm": 0.1377580612897873,
      "learning_rate": 8.483065379656558e-05,
      "loss": 0.0676,
      "step": 878
    },
    {
      "epoch": 0.6594148537134283,
      "grad_norm": 0.13235914707183838,
      "learning_rate": 8.478364757346546e-05,
      "loss": 0.0567,
      "step": 879
    },
    {
      "epoch": 0.660165041260315,
      "grad_norm": 0.23526223003864288,
      "learning_rate": 8.473658169983496e-05,
      "loss": 0.0602,
      "step": 880
    },
    {
      "epoch": 0.6609152288072018,
      "grad_norm": 0.1625460982322693,
      "learning_rate": 8.468945625638742e-05,
      "loss": 0.0723,
      "step": 881
    },
    {
      "epoch": 0.6616654163540885,
      "grad_norm": 0.19045262038707733,
      "learning_rate": 8.464227132393831e-05,
      "loss": 0.0901,
      "step": 882
    },
    {
      "epoch": 0.6624156039009752,
      "grad_norm": 0.20493094623088837,
      "learning_rate": 8.459502698340519e-05,
      "loss": 0.0856,
      "step": 883
    },
    {
      "epoch": 0.6631657914478619,
      "grad_norm": 0.1073690876364708,
      "learning_rate": 8.45477233158074e-05,
      "loss": 0.0397,
      "step": 884
    },
    {
      "epoch": 0.6639159789947486,
      "grad_norm": 0.12174858152866364,
      "learning_rate": 8.450036040226612e-05,
      "loss": 0.0567,
      "step": 885
    },
    {
      "epoch": 0.6646661665416355,
      "grad_norm": 0.10243424773216248,
      "learning_rate": 8.445293832400402e-05,
      "loss": 0.0391,
      "step": 886
    },
    {
      "epoch": 0.6654163540885222,
      "grad_norm": 0.10561061650514603,
      "learning_rate": 8.440545716234538e-05,
      "loss": 0.0341,
      "step": 887
    },
    {
      "epoch": 0.6661665416354089,
      "grad_norm": 0.09283889830112457,
      "learning_rate": 8.435791699871564e-05,
      "loss": 0.0487,
      "step": 888
    },
    {
      "epoch": 0.6669167291822956,
      "grad_norm": 0.17678828537464142,
      "learning_rate": 8.431031791464155e-05,
      "loss": 0.088,
      "step": 889
    },
    {
      "epoch": 0.6676669167291823,
      "grad_norm": 0.11029431223869324,
      "learning_rate": 8.426265999175081e-05,
      "loss": 0.0433,
      "step": 890
    },
    {
      "epoch": 0.668417104276069,
      "grad_norm": 0.22715550661087036,
      "learning_rate": 8.421494331177208e-05,
      "loss": 0.0859,
      "step": 891
    },
    {
      "epoch": 0.6691672918229558,
      "grad_norm": 0.1496887505054474,
      "learning_rate": 8.41671679565348e-05,
      "loss": 0.0584,
      "step": 892
    },
    {
      "epoch": 0.6699174793698425,
      "grad_norm": 0.20761065185070038,
      "learning_rate": 8.411933400796896e-05,
      "loss": 0.0691,
      "step": 893
    },
    {
      "epoch": 0.6706676669167292,
      "grad_norm": 0.17341768741607666,
      "learning_rate": 8.407144154810509e-05,
      "loss": 0.0869,
      "step": 894
    },
    {
      "epoch": 0.6714178544636159,
      "grad_norm": 0.13700233399868011,
      "learning_rate": 8.402349065907403e-05,
      "loss": 0.0644,
      "step": 895
    },
    {
      "epoch": 0.6721680420105026,
      "grad_norm": 0.19762715697288513,
      "learning_rate": 8.397548142310685e-05,
      "loss": 0.07,
      "step": 896
    },
    {
      "epoch": 0.6729182295573893,
      "grad_norm": 0.1878272444009781,
      "learning_rate": 8.392741392253465e-05,
      "loss": 0.0722,
      "step": 897
    },
    {
      "epoch": 0.673668417104276,
      "grad_norm": 0.18015135824680328,
      "learning_rate": 8.387928823978846e-05,
      "loss": 0.0544,
      "step": 898
    },
    {
      "epoch": 0.6744186046511628,
      "grad_norm": 0.10764296352863312,
      "learning_rate": 8.383110445739907e-05,
      "loss": 0.0369,
      "step": 899
    },
    {
      "epoch": 0.6751687921980495,
      "grad_norm": 0.09914194792509079,
      "learning_rate": 8.378286265799698e-05,
      "loss": 0.0452,
      "step": 900
    },
    {
      "epoch": 0.6759189797449362,
      "grad_norm": 0.17248530685901642,
      "learning_rate": 8.373456292431206e-05,
      "loss": 0.0662,
      "step": 901
    },
    {
      "epoch": 0.6766691672918229,
      "grad_norm": 0.18333736062049866,
      "learning_rate": 8.368620533917363e-05,
      "loss": 0.0724,
      "step": 902
    },
    {
      "epoch": 0.6774193548387096,
      "grad_norm": 0.11491480469703674,
      "learning_rate": 8.363778998551018e-05,
      "loss": 0.0645,
      "step": 903
    },
    {
      "epoch": 0.6781695423855963,
      "grad_norm": 0.13533350825309753,
      "learning_rate": 8.358931694634928e-05,
      "loss": 0.0541,
      "step": 904
    },
    {
      "epoch": 0.6789197299324832,
      "grad_norm": 0.15621417760849,
      "learning_rate": 8.35407863048174e-05,
      "loss": 0.0574,
      "step": 905
    },
    {
      "epoch": 0.6796699174793699,
      "grad_norm": 0.12412711977958679,
      "learning_rate": 8.349219814413984e-05,
      "loss": 0.0335,
      "step": 906
    },
    {
      "epoch": 0.6804201050262566,
      "grad_norm": 0.22609005868434906,
      "learning_rate": 8.344355254764049e-05,
      "loss": 0.0816,
      "step": 907
    },
    {
      "epoch": 0.6811702925731433,
      "grad_norm": 0.1081281304359436,
      "learning_rate": 8.339484959874178e-05,
      "loss": 0.0421,
      "step": 908
    },
    {
      "epoch": 0.68192048012003,
      "grad_norm": 0.15061849355697632,
      "learning_rate": 8.334608938096443e-05,
      "loss": 0.06,
      "step": 909
    },
    {
      "epoch": 0.6826706676669168,
      "grad_norm": 0.15337811410427094,
      "learning_rate": 8.329727197792744e-05,
      "loss": 0.0484,
      "step": 910
    },
    {
      "epoch": 0.6834208552138035,
      "grad_norm": 0.10216069221496582,
      "learning_rate": 8.324839747334787e-05,
      "loss": 0.0363,
      "step": 911
    },
    {
      "epoch": 0.6841710427606902,
      "grad_norm": 0.1537286341190338,
      "learning_rate": 8.319946595104065e-05,
      "loss": 0.0459,
      "step": 912
    },
    {
      "epoch": 0.6849212303075769,
      "grad_norm": 0.14444048702716827,
      "learning_rate": 8.315047749491851e-05,
      "loss": 0.0518,
      "step": 913
    },
    {
      "epoch": 0.6856714178544636,
      "grad_norm": 0.15967465937137604,
      "learning_rate": 8.310143218899187e-05,
      "loss": 0.0499,
      "step": 914
    },
    {
      "epoch": 0.6864216054013503,
      "grad_norm": 0.17263127863407135,
      "learning_rate": 8.305233011736857e-05,
      "loss": 0.087,
      "step": 915
    },
    {
      "epoch": 0.687171792948237,
      "grad_norm": 0.18618431687355042,
      "learning_rate": 8.300317136425385e-05,
      "loss": 0.0533,
      "step": 916
    },
    {
      "epoch": 0.6879219804951238,
      "grad_norm": 0.17359670996665955,
      "learning_rate": 8.295395601395011e-05,
      "loss": 0.0653,
      "step": 917
    },
    {
      "epoch": 0.6886721680420105,
      "grad_norm": 0.14314648509025574,
      "learning_rate": 8.290468415085683e-05,
      "loss": 0.0523,
      "step": 918
    },
    {
      "epoch": 0.6894223555888972,
      "grad_norm": 0.19745318591594696,
      "learning_rate": 8.285535585947042e-05,
      "loss": 0.0623,
      "step": 919
    },
    {
      "epoch": 0.6901725431357839,
      "grad_norm": 0.1895579844713211,
      "learning_rate": 8.280597122438404e-05,
      "loss": 0.0668,
      "step": 920
    },
    {
      "epoch": 0.6909227306826706,
      "grad_norm": 0.17338034510612488,
      "learning_rate": 8.275653033028745e-05,
      "loss": 0.0607,
      "step": 921
    },
    {
      "epoch": 0.6916729182295573,
      "grad_norm": 0.12494196742773056,
      "learning_rate": 8.270703326196696e-05,
      "loss": 0.0646,
      "step": 922
    },
    {
      "epoch": 0.6924231057764441,
      "grad_norm": 0.12629970908164978,
      "learning_rate": 8.265748010430513e-05,
      "loss": 0.0576,
      "step": 923
    },
    {
      "epoch": 0.6931732933233309,
      "grad_norm": 0.14575371146202087,
      "learning_rate": 8.260787094228076e-05,
      "loss": 0.0684,
      "step": 924
    },
    {
      "epoch": 0.6939234808702176,
      "grad_norm": 0.1634964495897293,
      "learning_rate": 8.255820586096867e-05,
      "loss": 0.0797,
      "step": 925
    },
    {
      "epoch": 0.6946736684171043,
      "grad_norm": 0.17009958624839783,
      "learning_rate": 8.25084849455396e-05,
      "loss": 0.0808,
      "step": 926
    },
    {
      "epoch": 0.695423855963991,
      "grad_norm": 0.17264492809772491,
      "learning_rate": 8.245870828126e-05,
      "loss": 0.0842,
      "step": 927
    },
    {
      "epoch": 0.6961740435108777,
      "grad_norm": 0.1700659990310669,
      "learning_rate": 8.240887595349197e-05,
      "loss": 0.0652,
      "step": 928
    },
    {
      "epoch": 0.6969242310577645,
      "grad_norm": 0.16623543202877045,
      "learning_rate": 8.235898804769303e-05,
      "loss": 0.0827,
      "step": 929
    },
    {
      "epoch": 0.6976744186046512,
      "grad_norm": 0.17181211709976196,
      "learning_rate": 8.230904464941604e-05,
      "loss": 0.0755,
      "step": 930
    },
    {
      "epoch": 0.6984246061515379,
      "grad_norm": 0.19417133927345276,
      "learning_rate": 8.225904584430901e-05,
      "loss": 0.09,
      "step": 931
    },
    {
      "epoch": 0.6991747936984246,
      "grad_norm": 0.14830316603183746,
      "learning_rate": 8.220899171811495e-05,
      "loss": 0.0559,
      "step": 932
    },
    {
      "epoch": 0.6999249812453113,
      "grad_norm": 0.11206447333097458,
      "learning_rate": 8.215888235667176e-05,
      "loss": 0.0567,
      "step": 933
    },
    {
      "epoch": 0.700675168792198,
      "grad_norm": 0.15120255947113037,
      "learning_rate": 8.210871784591207e-05,
      "loss": 0.059,
      "step": 934
    },
    {
      "epoch": 0.7014253563390848,
      "grad_norm": 0.152877077460289,
      "learning_rate": 8.205849827186308e-05,
      "loss": 0.0574,
      "step": 935
    },
    {
      "epoch": 0.7021755438859715,
      "grad_norm": 0.1164332926273346,
      "learning_rate": 8.200822372064641e-05,
      "loss": 0.0586,
      "step": 936
    },
    {
      "epoch": 0.7029257314328582,
      "grad_norm": 0.17080263793468475,
      "learning_rate": 8.195789427847796e-05,
      "loss": 0.0597,
      "step": 937
    },
    {
      "epoch": 0.7036759189797449,
      "grad_norm": 0.0957210585474968,
      "learning_rate": 8.190751003166778e-05,
      "loss": 0.039,
      "step": 938
    },
    {
      "epoch": 0.7044261065266316,
      "grad_norm": 0.15183965861797333,
      "learning_rate": 8.185707106661986e-05,
      "loss": 0.0742,
      "step": 939
    },
    {
      "epoch": 0.7051762940735183,
      "grad_norm": 0.15807807445526123,
      "learning_rate": 8.18065774698321e-05,
      "loss": 0.0585,
      "step": 940
    },
    {
      "epoch": 0.705926481620405,
      "grad_norm": 0.16685016453266144,
      "learning_rate": 8.175602932789601e-05,
      "loss": 0.0584,
      "step": 941
    },
    {
      "epoch": 0.7066766691672918,
      "grad_norm": 0.1323295682668686,
      "learning_rate": 8.17054267274967e-05,
      "loss": 0.0785,
      "step": 942
    },
    {
      "epoch": 0.7074268567141786,
      "grad_norm": 0.29115551710128784,
      "learning_rate": 8.165476975541264e-05,
      "loss": 0.0696,
      "step": 943
    },
    {
      "epoch": 0.7081770442610653,
      "grad_norm": 0.16488604247570038,
      "learning_rate": 8.160405849851556e-05,
      "loss": 0.0558,
      "step": 944
    },
    {
      "epoch": 0.708927231807952,
      "grad_norm": 0.16088944673538208,
      "learning_rate": 8.155329304377025e-05,
      "loss": 0.0616,
      "step": 945
    },
    {
      "epoch": 0.7096774193548387,
      "grad_norm": 0.17655105888843536,
      "learning_rate": 8.150247347823448e-05,
      "loss": 0.0749,
      "step": 946
    },
    {
      "epoch": 0.7104276069017255,
      "grad_norm": 0.15837235748767853,
      "learning_rate": 8.145159988905879e-05,
      "loss": 0.081,
      "step": 947
    },
    {
      "epoch": 0.7111777944486122,
      "grad_norm": 0.18028263747692108,
      "learning_rate": 8.140067236348638e-05,
      "loss": 0.0801,
      "step": 948
    },
    {
      "epoch": 0.7119279819954989,
      "grad_norm": 0.20195040106773376,
      "learning_rate": 8.134969098885294e-05,
      "loss": 0.1123,
      "step": 949
    },
    {
      "epoch": 0.7126781695423856,
      "grad_norm": 0.16333803534507751,
      "learning_rate": 8.129865585258653e-05,
      "loss": 0.0804,
      "step": 950
    },
    {
      "epoch": 0.7134283570892723,
      "grad_norm": 0.15731093287467957,
      "learning_rate": 8.124756704220735e-05,
      "loss": 0.0696,
      "step": 951
    },
    {
      "epoch": 0.714178544636159,
      "grad_norm": 0.1764248162508011,
      "learning_rate": 8.11964246453277e-05,
      "loss": 0.0805,
      "step": 952
    },
    {
      "epoch": 0.7149287321830458,
      "grad_norm": 0.17285150289535522,
      "learning_rate": 8.114522874965174e-05,
      "loss": 0.074,
      "step": 953
    },
    {
      "epoch": 0.7156789197299325,
      "grad_norm": 0.22047369182109833,
      "learning_rate": 8.10939794429754e-05,
      "loss": 0.0731,
      "step": 954
    },
    {
      "epoch": 0.7164291072768192,
      "grad_norm": 0.20119711756706238,
      "learning_rate": 8.10426768131862e-05,
      "loss": 0.0794,
      "step": 955
    },
    {
      "epoch": 0.7171792948237059,
      "grad_norm": 0.10244588553905487,
      "learning_rate": 8.099132094826308e-05,
      "loss": 0.0612,
      "step": 956
    },
    {
      "epoch": 0.7179294823705926,
      "grad_norm": 0.14183086156845093,
      "learning_rate": 8.093991193627631e-05,
      "loss": 0.0849,
      "step": 957
    },
    {
      "epoch": 0.7186796699174793,
      "grad_norm": 0.13276663422584534,
      "learning_rate": 8.088844986538727e-05,
      "loss": 0.0511,
      "step": 958
    },
    {
      "epoch": 0.719429857464366,
      "grad_norm": 0.1904866099357605,
      "learning_rate": 8.083693482384836e-05,
      "loss": 0.0661,
      "step": 959
    },
    {
      "epoch": 0.7201800450112528,
      "grad_norm": 0.20430415868759155,
      "learning_rate": 8.078536690000278e-05,
      "loss": 0.0936,
      "step": 960
    },
    {
      "epoch": 0.7209302325581395,
      "grad_norm": 0.13531947135925293,
      "learning_rate": 8.073374618228445e-05,
      "loss": 0.0594,
      "step": 961
    },
    {
      "epoch": 0.7216804201050263,
      "grad_norm": 0.14323526620864868,
      "learning_rate": 8.068207275921782e-05,
      "loss": 0.0659,
      "step": 962
    },
    {
      "epoch": 0.722430607651913,
      "grad_norm": 0.12277992069721222,
      "learning_rate": 8.063034671941774e-05,
      "loss": 0.0493,
      "step": 963
    },
    {
      "epoch": 0.7231807951987997,
      "grad_norm": 0.19475075602531433,
      "learning_rate": 8.057856815158924e-05,
      "loss": 0.0737,
      "step": 964
    },
    {
      "epoch": 0.7239309827456865,
      "grad_norm": 0.11780892312526703,
      "learning_rate": 8.05267371445275e-05,
      "loss": 0.0576,
      "step": 965
    },
    {
      "epoch": 0.7246811702925732,
      "grad_norm": 0.13752034306526184,
      "learning_rate": 8.047485378711756e-05,
      "loss": 0.0412,
      "step": 966
    },
    {
      "epoch": 0.7254313578394599,
      "grad_norm": 0.1040201336145401,
      "learning_rate": 8.042291816833429e-05,
      "loss": 0.0415,
      "step": 967
    },
    {
      "epoch": 0.7261815453863466,
      "grad_norm": 0.22411030530929565,
      "learning_rate": 8.037093037724216e-05,
      "loss": 0.0966,
      "step": 968
    },
    {
      "epoch": 0.7269317329332333,
      "grad_norm": 0.1516868621110916,
      "learning_rate": 8.031889050299511e-05,
      "loss": 0.0643,
      "step": 969
    },
    {
      "epoch": 0.72768192048012,
      "grad_norm": 0.18635967373847961,
      "learning_rate": 8.02667986348364e-05,
      "loss": 0.1136,
      "step": 970
    },
    {
      "epoch": 0.7284321080270068,
      "grad_norm": 0.16178272664546967,
      "learning_rate": 8.021465486209846e-05,
      "loss": 0.0615,
      "step": 971
    },
    {
      "epoch": 0.7291822955738935,
      "grad_norm": 0.15400740504264832,
      "learning_rate": 8.016245927420272e-05,
      "loss": 0.083,
      "step": 972
    },
    {
      "epoch": 0.7299324831207802,
      "grad_norm": 0.1883528083562851,
      "learning_rate": 8.011021196065946e-05,
      "loss": 0.0876,
      "step": 973
    },
    {
      "epoch": 0.7306826706676669,
      "grad_norm": 0.19876441359519958,
      "learning_rate": 8.005791301106769e-05,
      "loss": 0.088,
      "step": 974
    },
    {
      "epoch": 0.7314328582145536,
      "grad_norm": 0.1295093595981598,
      "learning_rate": 8.000556251511498e-05,
      "loss": 0.062,
      "step": 975
    },
    {
      "epoch": 0.7321830457614403,
      "grad_norm": 0.11591675877571106,
      "learning_rate": 7.995316056257723e-05,
      "loss": 0.0481,
      "step": 976
    },
    {
      "epoch": 0.732933233308327,
      "grad_norm": 0.13371731340885162,
      "learning_rate": 7.990070724331866e-05,
      "loss": 0.0555,
      "step": 977
    },
    {
      "epoch": 0.7336834208552138,
      "grad_norm": 0.18042734265327454,
      "learning_rate": 7.984820264729156e-05,
      "loss": 0.0673,
      "step": 978
    },
    {
      "epoch": 0.7344336084021005,
      "grad_norm": 0.1305835098028183,
      "learning_rate": 7.979564686453612e-05,
      "loss": 0.0699,
      "step": 979
    },
    {
      "epoch": 0.7351837959489872,
      "grad_norm": 0.1295088827610016,
      "learning_rate": 7.974303998518031e-05,
      "loss": 0.0522,
      "step": 980
    },
    {
      "epoch": 0.735933983495874,
      "grad_norm": 0.16895820200443268,
      "learning_rate": 7.96903820994398e-05,
      "loss": 0.0591,
      "step": 981
    },
    {
      "epoch": 0.7366841710427607,
      "grad_norm": 0.22246110439300537,
      "learning_rate": 7.963767329761762e-05,
      "loss": 0.1312,
      "step": 982
    },
    {
      "epoch": 0.7374343585896475,
      "grad_norm": 0.1366637945175171,
      "learning_rate": 7.958491367010423e-05,
      "loss": 0.0702,
      "step": 983
    },
    {
      "epoch": 0.7381845461365342,
      "grad_norm": 0.13772538304328918,
      "learning_rate": 7.953210330737718e-05,
      "loss": 0.0679,
      "step": 984
    },
    {
      "epoch": 0.7389347336834209,
      "grad_norm": 0.16669201850891113,
      "learning_rate": 7.947924230000102e-05,
      "loss": 0.0571,
      "step": 985
    },
    {
      "epoch": 0.7396849212303076,
      "grad_norm": 0.1642238348722458,
      "learning_rate": 7.942633073862718e-05,
      "loss": 0.0709,
      "step": 986
    },
    {
      "epoch": 0.7404351087771943,
      "grad_norm": 0.22212821245193481,
      "learning_rate": 7.937336871399379e-05,
      "loss": 0.0779,
      "step": 987
    },
    {
      "epoch": 0.741185296324081,
      "grad_norm": 0.19373202323913574,
      "learning_rate": 7.932035631692549e-05,
      "loss": 0.074,
      "step": 988
    },
    {
      "epoch": 0.7419354838709677,
      "grad_norm": 0.19060571491718292,
      "learning_rate": 7.926729363833335e-05,
      "loss": 0.0745,
      "step": 989
    },
    {
      "epoch": 0.7426856714178545,
      "grad_norm": 0.19428344070911407,
      "learning_rate": 7.921418076921461e-05,
      "loss": 0.078,
      "step": 990
    },
    {
      "epoch": 0.7434358589647412,
      "grad_norm": 0.1412833333015442,
      "learning_rate": 7.916101780065263e-05,
      "loss": 0.0634,
      "step": 991
    },
    {
      "epoch": 0.7441860465116279,
      "grad_norm": 0.11580026149749756,
      "learning_rate": 7.910780482381665e-05,
      "loss": 0.0496,
      "step": 992
    },
    {
      "epoch": 0.7449362340585146,
      "grad_norm": 0.14477257430553436,
      "learning_rate": 7.905454192996169e-05,
      "loss": 0.066,
      "step": 993
    },
    {
      "epoch": 0.7456864216054013,
      "grad_norm": 0.19042307138442993,
      "learning_rate": 7.900122921042837e-05,
      "loss": 0.0673,
      "step": 994
    },
    {
      "epoch": 0.746436609152288,
      "grad_norm": 0.18369849026203156,
      "learning_rate": 7.894786675664277e-05,
      "loss": 0.0697,
      "step": 995
    },
    {
      "epoch": 0.7471867966991748,
      "grad_norm": 0.17455372214317322,
      "learning_rate": 7.88944546601162e-05,
      "loss": 0.0833,
      "step": 996
    },
    {
      "epoch": 0.7479369842460615,
      "grad_norm": 0.11367595195770264,
      "learning_rate": 7.884099301244519e-05,
      "loss": 0.0459,
      "step": 997
    },
    {
      "epoch": 0.7486871717929482,
      "grad_norm": 0.09311968833208084,
      "learning_rate": 7.878748190531118e-05,
      "loss": 0.0397,
      "step": 998
    },
    {
      "epoch": 0.7494373593398349,
      "grad_norm": 0.14789381623268127,
      "learning_rate": 7.873392143048047e-05,
      "loss": 0.061,
      "step": 999
    },
    {
      "epoch": 0.7501875468867217,
      "grad_norm": 0.18568088114261627,
      "learning_rate": 7.868031167980397e-05,
      "loss": 0.1057,
      "step": 1000
    },
    {
      "epoch": 0.7501875468867217,
      "eval_loss": 0.06887388974428177,
      "eval_runtime": 2.654,
      "eval_samples_per_second": 20.347,
      "eval_steps_per_second": 5.275,
      "step": 1000
    },
    {
      "epoch": 0.7509377344336085,
      "grad_norm": 0.1491386443376541,
      "learning_rate": 7.862665274521712e-05,
      "loss": 0.0619,
      "step": 1001
    },
    {
      "epoch": 0.7516879219804952,
      "grad_norm": 0.1528184562921524,
      "learning_rate": 7.857294471873975e-05,
      "loss": 0.0653,
      "step": 1002
    },
    {
      "epoch": 0.7524381095273819,
      "grad_norm": 0.16220131516456604,
      "learning_rate": 7.851918769247582e-05,
      "loss": 0.0813,
      "step": 1003
    },
    {
      "epoch": 0.7531882970742686,
      "grad_norm": 0.1285926252603531,
      "learning_rate": 7.846538175861332e-05,
      "loss": 0.048,
      "step": 1004
    },
    {
      "epoch": 0.7539384846211553,
      "grad_norm": 0.12812086939811707,
      "learning_rate": 7.841152700942413e-05,
      "loss": 0.0726,
      "step": 1005
    },
    {
      "epoch": 0.754688672168042,
      "grad_norm": 0.17180190980434418,
      "learning_rate": 7.835762353726386e-05,
      "loss": 0.0835,
      "step": 1006
    },
    {
      "epoch": 0.7554388597149287,
      "grad_norm": 0.18833868205547333,
      "learning_rate": 7.830367143457165e-05,
      "loss": 0.0819,
      "step": 1007
    },
    {
      "epoch": 0.7561890472618155,
      "grad_norm": 0.36495018005371094,
      "learning_rate": 7.824967079387002e-05,
      "loss": 0.0986,
      "step": 1008
    },
    {
      "epoch": 0.7569392348087022,
      "grad_norm": 0.1798129677772522,
      "learning_rate": 7.81956217077648e-05,
      "loss": 0.0735,
      "step": 1009
    },
    {
      "epoch": 0.7576894223555889,
      "grad_norm": 0.12408173084259033,
      "learning_rate": 7.814152426894478e-05,
      "loss": 0.0545,
      "step": 1010
    },
    {
      "epoch": 0.7584396099024756,
      "grad_norm": 0.1563502848148346,
      "learning_rate": 7.808737857018182e-05,
      "loss": 0.0817,
      "step": 1011
    },
    {
      "epoch": 0.7591897974493623,
      "grad_norm": 0.1812286674976349,
      "learning_rate": 7.803318470433042e-05,
      "loss": 0.0619,
      "step": 1012
    },
    {
      "epoch": 0.759939984996249,
      "grad_norm": 0.12106587737798691,
      "learning_rate": 7.797894276432772e-05,
      "loss": 0.0651,
      "step": 1013
    },
    {
      "epoch": 0.7606901725431358,
      "grad_norm": 0.13680601119995117,
      "learning_rate": 7.792465284319332e-05,
      "loss": 0.0692,
      "step": 1014
    },
    {
      "epoch": 0.7614403600900225,
      "grad_norm": 0.13508430123329163,
      "learning_rate": 7.787031503402907e-05,
      "loss": 0.0568,
      "step": 1015
    },
    {
      "epoch": 0.7621905476369092,
      "grad_norm": 0.17294204235076904,
      "learning_rate": 7.781592943001899e-05,
      "loss": 0.0633,
      "step": 1016
    },
    {
      "epoch": 0.7629407351837959,
      "grad_norm": 0.12395866960287094,
      "learning_rate": 7.776149612442899e-05,
      "loss": 0.0702,
      "step": 1017
    },
    {
      "epoch": 0.7636909227306826,
      "grad_norm": 0.13402384519577026,
      "learning_rate": 7.770701521060688e-05,
      "loss": 0.0554,
      "step": 1018
    },
    {
      "epoch": 0.7644411102775694,
      "grad_norm": 0.11290405690670013,
      "learning_rate": 7.765248678198203e-05,
      "loss": 0.0429,
      "step": 1019
    },
    {
      "epoch": 0.7651912978244562,
      "grad_norm": 0.12147333472967148,
      "learning_rate": 7.759791093206534e-05,
      "loss": 0.0487,
      "step": 1020
    },
    {
      "epoch": 0.7659414853713429,
      "grad_norm": 0.13817934691905975,
      "learning_rate": 7.754328775444903e-05,
      "loss": 0.0625,
      "step": 1021
    },
    {
      "epoch": 0.7666916729182296,
      "grad_norm": 0.19256222248077393,
      "learning_rate": 7.748861734280643e-05,
      "loss": 0.079,
      "step": 1022
    },
    {
      "epoch": 0.7674418604651163,
      "grad_norm": 0.1472809612751007,
      "learning_rate": 7.743389979089196e-05,
      "loss": 0.0509,
      "step": 1023
    },
    {
      "epoch": 0.768192048012003,
      "grad_norm": 0.1204993948340416,
      "learning_rate": 7.737913519254079e-05,
      "loss": 0.0549,
      "step": 1024
    },
    {
      "epoch": 0.7689422355588897,
      "grad_norm": 0.22745735943317413,
      "learning_rate": 7.732432364166884e-05,
      "loss": 0.1078,
      "step": 1025
    },
    {
      "epoch": 0.7696924231057765,
      "grad_norm": 0.26715755462646484,
      "learning_rate": 7.726946523227251e-05,
      "loss": 0.0708,
      "step": 1026
    },
    {
      "epoch": 0.7704426106526632,
      "grad_norm": 0.14105719327926636,
      "learning_rate": 7.721456005842861e-05,
      "loss": 0.0634,
      "step": 1027
    },
    {
      "epoch": 0.7711927981995499,
      "grad_norm": 0.13659711182117462,
      "learning_rate": 7.715960821429404e-05,
      "loss": 0.0417,
      "step": 1028
    },
    {
      "epoch": 0.7719429857464366,
      "grad_norm": 0.16884845495224,
      "learning_rate": 7.710460979410585e-05,
      "loss": 0.0621,
      "step": 1029
    },
    {
      "epoch": 0.7726931732933233,
      "grad_norm": 0.20827940106391907,
      "learning_rate": 7.704956489218091e-05,
      "loss": 0.072,
      "step": 1030
    },
    {
      "epoch": 0.77344336084021,
      "grad_norm": 0.19093617796897888,
      "learning_rate": 7.699447360291576e-05,
      "loss": 0.0896,
      "step": 1031
    },
    {
      "epoch": 0.7741935483870968,
      "grad_norm": 0.14019818603992462,
      "learning_rate": 7.69393360207866e-05,
      "loss": 0.0437,
      "step": 1032
    },
    {
      "epoch": 0.7749437359339835,
      "grad_norm": 0.14744336903095245,
      "learning_rate": 7.688415224034893e-05,
      "loss": 0.056,
      "step": 1033
    },
    {
      "epoch": 0.7756939234808702,
      "grad_norm": 0.13031582534313202,
      "learning_rate": 7.682892235623749e-05,
      "loss": 0.0438,
      "step": 1034
    },
    {
      "epoch": 0.7764441110277569,
      "grad_norm": 0.10287275165319443,
      "learning_rate": 7.67736464631661e-05,
      "loss": 0.045,
      "step": 1035
    },
    {
      "epoch": 0.7771942985746436,
      "grad_norm": 0.16352872550487518,
      "learning_rate": 7.671832465592746e-05,
      "loss": 0.0718,
      "step": 1036
    },
    {
      "epoch": 0.7779444861215303,
      "grad_norm": 0.1713312715291977,
      "learning_rate": 7.666295702939305e-05,
      "loss": 0.0713,
      "step": 1037
    },
    {
      "epoch": 0.7786946736684172,
      "grad_norm": 0.18730027973651886,
      "learning_rate": 7.660754367851286e-05,
      "loss": 0.0698,
      "step": 1038
    },
    {
      "epoch": 0.7794448612153039,
      "grad_norm": 0.1919287145137787,
      "learning_rate": 7.655208469831536e-05,
      "loss": 0.0805,
      "step": 1039
    },
    {
      "epoch": 0.7801950487621906,
      "grad_norm": 0.20845071971416473,
      "learning_rate": 7.649658018390725e-05,
      "loss": 0.0723,
      "step": 1040
    },
    {
      "epoch": 0.7809452363090773,
      "grad_norm": 0.10697642713785172,
      "learning_rate": 7.644103023047327e-05,
      "loss": 0.0501,
      "step": 1041
    },
    {
      "epoch": 0.781695423855964,
      "grad_norm": 0.14027732610702515,
      "learning_rate": 7.638543493327613e-05,
      "loss": 0.0436,
      "step": 1042
    },
    {
      "epoch": 0.7824456114028507,
      "grad_norm": 0.15785855054855347,
      "learning_rate": 7.63297943876563e-05,
      "loss": 0.0549,
      "step": 1043
    },
    {
      "epoch": 0.7831957989497375,
      "grad_norm": 0.2016582190990448,
      "learning_rate": 7.627410868903184e-05,
      "loss": 0.07,
      "step": 1044
    },
    {
      "epoch": 0.7839459864966242,
      "grad_norm": 0.1890334188938141,
      "learning_rate": 7.621837793289824e-05,
      "loss": 0.0719,
      "step": 1045
    },
    {
      "epoch": 0.7846961740435109,
      "grad_norm": 0.1692233830690384,
      "learning_rate": 7.616260221482825e-05,
      "loss": 0.0546,
      "step": 1046
    },
    {
      "epoch": 0.7854463615903976,
      "grad_norm": 0.20178622007369995,
      "learning_rate": 7.610678163047174e-05,
      "loss": 0.0839,
      "step": 1047
    },
    {
      "epoch": 0.7861965491372843,
      "grad_norm": 0.12947288155555725,
      "learning_rate": 7.60509162755555e-05,
      "loss": 0.0592,
      "step": 1048
    },
    {
      "epoch": 0.786946736684171,
      "grad_norm": 0.1551869809627533,
      "learning_rate": 7.599500624588314e-05,
      "loss": 0.0643,
      "step": 1049
    },
    {
      "epoch": 0.7876969242310577,
      "grad_norm": 0.18167535960674286,
      "learning_rate": 7.593905163733484e-05,
      "loss": 0.0823,
      "step": 1050
    },
    {
      "epoch": 0.7884471117779445,
      "grad_norm": 0.18511007726192474,
      "learning_rate": 7.588305254586724e-05,
      "loss": 0.0805,
      "step": 1051
    },
    {
      "epoch": 0.7891972993248312,
      "grad_norm": 0.1279374212026596,
      "learning_rate": 7.582700906751325e-05,
      "loss": 0.0564,
      "step": 1052
    },
    {
      "epoch": 0.7899474868717179,
      "grad_norm": 0.1779521256685257,
      "learning_rate": 7.577092129838197e-05,
      "loss": 0.0659,
      "step": 1053
    },
    {
      "epoch": 0.7906976744186046,
      "grad_norm": 0.1136535033583641,
      "learning_rate": 7.571478933465836e-05,
      "loss": 0.0548,
      "step": 1054
    },
    {
      "epoch": 0.7914478619654913,
      "grad_norm": 0.17618438601493835,
      "learning_rate": 7.565861327260322e-05,
      "loss": 0.0725,
      "step": 1055
    },
    {
      "epoch": 0.792198049512378,
      "grad_norm": 0.23887811601161957,
      "learning_rate": 7.560239320855296e-05,
      "loss": 0.0956,
      "step": 1056
    },
    {
      "epoch": 0.7929482370592649,
      "grad_norm": 0.24576206505298615,
      "learning_rate": 7.554612923891946e-05,
      "loss": 0.1265,
      "step": 1057
    },
    {
      "epoch": 0.7936984246061516,
      "grad_norm": 0.17981268465518951,
      "learning_rate": 7.548982146018988e-05,
      "loss": 0.0779,
      "step": 1058
    },
    {
      "epoch": 0.7944486121530383,
      "grad_norm": 0.14327049255371094,
      "learning_rate": 7.543346996892654e-05,
      "loss": 0.0781,
      "step": 1059
    },
    {
      "epoch": 0.795198799699925,
      "grad_norm": 0.17094387114048004,
      "learning_rate": 7.537707486176667e-05,
      "loss": 0.0614,
      "step": 1060
    },
    {
      "epoch": 0.7959489872468117,
      "grad_norm": 0.10642191767692566,
      "learning_rate": 7.532063623542231e-05,
      "loss": 0.0572,
      "step": 1061
    },
    {
      "epoch": 0.7966991747936985,
      "grad_norm": 0.13238154351711273,
      "learning_rate": 7.52641541866802e-05,
      "loss": 0.0453,
      "step": 1062
    },
    {
      "epoch": 0.7974493623405852,
      "grad_norm": 0.19573621451854706,
      "learning_rate": 7.520762881240147e-05,
      "loss": 0.1012,
      "step": 1063
    },
    {
      "epoch": 0.7981995498874719,
      "grad_norm": 0.1620679497718811,
      "learning_rate": 7.515106020952156e-05,
      "loss": 0.0721,
      "step": 1064
    },
    {
      "epoch": 0.7989497374343586,
      "grad_norm": 0.1418398767709732,
      "learning_rate": 7.509444847505005e-05,
      "loss": 0.0822,
      "step": 1065
    },
    {
      "epoch": 0.7996999249812453,
      "grad_norm": 0.14470888674259186,
      "learning_rate": 7.503779370607049e-05,
      "loss": 0.0582,
      "step": 1066
    },
    {
      "epoch": 0.800450112528132,
      "grad_norm": 0.12305848300457001,
      "learning_rate": 7.498109599974024e-05,
      "loss": 0.0615,
      "step": 1067
    },
    {
      "epoch": 0.8012003000750187,
      "grad_norm": 0.12674108147621155,
      "learning_rate": 7.49243554532903e-05,
      "loss": 0.0693,
      "step": 1068
    },
    {
      "epoch": 0.8019504876219055,
      "grad_norm": 0.14826877415180206,
      "learning_rate": 7.486757216402509e-05,
      "loss": 0.0548,
      "step": 1069
    },
    {
      "epoch": 0.8027006751687922,
      "grad_norm": 0.12875312566757202,
      "learning_rate": 7.481074622932236e-05,
      "loss": 0.0556,
      "step": 1070
    },
    {
      "epoch": 0.8034508627156789,
      "grad_norm": 0.15808606147766113,
      "learning_rate": 7.475387774663302e-05,
      "loss": 0.0498,
      "step": 1071
    },
    {
      "epoch": 0.8042010502625656,
      "grad_norm": 0.11491523683071136,
      "learning_rate": 7.469696681348088e-05,
      "loss": 0.0487,
      "step": 1072
    },
    {
      "epoch": 0.8049512378094523,
      "grad_norm": 0.20159392058849335,
      "learning_rate": 7.464001352746263e-05,
      "loss": 0.1028,
      "step": 1073
    },
    {
      "epoch": 0.805701425356339,
      "grad_norm": 0.16383297741413116,
      "learning_rate": 7.45830179862475e-05,
      "loss": 0.0681,
      "step": 1074
    },
    {
      "epoch": 0.8064516129032258,
      "grad_norm": 0.12068310379981995,
      "learning_rate": 7.452598028757729e-05,
      "loss": 0.0816,
      "step": 1075
    },
    {
      "epoch": 0.8072018004501126,
      "grad_norm": 0.2644040584564209,
      "learning_rate": 7.446890052926598e-05,
      "loss": 0.0731,
      "step": 1076
    },
    {
      "epoch": 0.8079519879969993,
      "grad_norm": 0.13366550207138062,
      "learning_rate": 7.441177880919976e-05,
      "loss": 0.0596,
      "step": 1077
    },
    {
      "epoch": 0.808702175543886,
      "grad_norm": 0.22289660573005676,
      "learning_rate": 7.435461522533674e-05,
      "loss": 0.0567,
      "step": 1078
    },
    {
      "epoch": 0.8094523630907727,
      "grad_norm": 0.13923311233520508,
      "learning_rate": 7.429740987570686e-05,
      "loss": 0.0651,
      "step": 1079
    },
    {
      "epoch": 0.8102025506376594,
      "grad_norm": 0.15409795939922333,
      "learning_rate": 7.424016285841165e-05,
      "loss": 0.0556,
      "step": 1080
    },
    {
      "epoch": 0.8109527381845462,
      "grad_norm": 0.1327374130487442,
      "learning_rate": 7.41828742716241e-05,
      "loss": 0.0611,
      "step": 1081
    },
    {
      "epoch": 0.8117029257314329,
      "grad_norm": 0.16146451234817505,
      "learning_rate": 7.41255442135885e-05,
      "loss": 0.0681,
      "step": 1082
    },
    {
      "epoch": 0.8124531132783196,
      "grad_norm": 0.22603389620780945,
      "learning_rate": 7.406817278262027e-05,
      "loss": 0.0843,
      "step": 1083
    },
    {
      "epoch": 0.8132033008252063,
      "grad_norm": 0.11806265264749527,
      "learning_rate": 7.401076007710575e-05,
      "loss": 0.0535,
      "step": 1084
    },
    {
      "epoch": 0.813953488372093,
      "grad_norm": 0.1646219938993454,
      "learning_rate": 7.395330619550207e-05,
      "loss": 0.0631,
      "step": 1085
    },
    {
      "epoch": 0.8147036759189797,
      "grad_norm": 0.22715872526168823,
      "learning_rate": 7.3895811236337e-05,
      "loss": 0.0812,
      "step": 1086
    },
    {
      "epoch": 0.8154538634658665,
      "grad_norm": 0.24135173857212067,
      "learning_rate": 7.38382752982087e-05,
      "loss": 0.0916,
      "step": 1087
    },
    {
      "epoch": 0.8162040510127532,
      "grad_norm": 0.22600874304771423,
      "learning_rate": 7.378069847978568e-05,
      "loss": 0.0794,
      "step": 1088
    },
    {
      "epoch": 0.8169542385596399,
      "grad_norm": 0.21018345654010773,
      "learning_rate": 7.372308087980647e-05,
      "loss": 0.0811,
      "step": 1089
    },
    {
      "epoch": 0.8177044261065266,
      "grad_norm": 0.17604215443134308,
      "learning_rate": 7.366542259707962e-05,
      "loss": 0.058,
      "step": 1090
    },
    {
      "epoch": 0.8184546136534133,
      "grad_norm": 0.1911773532629013,
      "learning_rate": 7.360772373048338e-05,
      "loss": 0.0716,
      "step": 1091
    },
    {
      "epoch": 0.8192048012003,
      "grad_norm": 0.15144562721252441,
      "learning_rate": 7.354998437896565e-05,
      "loss": 0.058,
      "step": 1092
    },
    {
      "epoch": 0.8199549887471868,
      "grad_norm": 0.16863825917243958,
      "learning_rate": 7.349220464154371e-05,
      "loss": 0.0696,
      "step": 1093
    },
    {
      "epoch": 0.8207051762940735,
      "grad_norm": 0.11625709384679794,
      "learning_rate": 7.343438461730411e-05,
      "loss": 0.0497,
      "step": 1094
    },
    {
      "epoch": 0.8214553638409603,
      "grad_norm": 0.19467462599277496,
      "learning_rate": 7.337652440540252e-05,
      "loss": 0.0994,
      "step": 1095
    },
    {
      "epoch": 0.822205551387847,
      "grad_norm": 0.24418720602989197,
      "learning_rate": 7.331862410506353e-05,
      "loss": 0.0856,
      "step": 1096
    },
    {
      "epoch": 0.8229557389347337,
      "grad_norm": 0.14274224638938904,
      "learning_rate": 7.32606838155804e-05,
      "loss": 0.0516,
      "step": 1097
    },
    {
      "epoch": 0.8237059264816204,
      "grad_norm": 0.14197169244289398,
      "learning_rate": 7.320270363631505e-05,
      "loss": 0.0686,
      "step": 1098
    },
    {
      "epoch": 0.8244561140285072,
      "grad_norm": 0.17585918307304382,
      "learning_rate": 7.314468366669777e-05,
      "loss": 0.0778,
      "step": 1099
    },
    {
      "epoch": 0.8252063015753939,
      "grad_norm": 0.1533975601196289,
      "learning_rate": 7.30866240062271e-05,
      "loss": 0.0554,
      "step": 1100
    },
    {
      "epoch": 0.8259564891222806,
      "grad_norm": 0.21065160632133484,
      "learning_rate": 7.302852475446963e-05,
      "loss": 0.0919,
      "step": 1101
    },
    {
      "epoch": 0.8267066766691673,
      "grad_norm": 0.1751699149608612,
      "learning_rate": 7.297038601105988e-05,
      "loss": 0.0765,
      "step": 1102
    },
    {
      "epoch": 0.827456864216054,
      "grad_norm": 0.16724807024002075,
      "learning_rate": 7.291220787570005e-05,
      "loss": 0.0703,
      "step": 1103
    },
    {
      "epoch": 0.8282070517629407,
      "grad_norm": 0.16113238036632538,
      "learning_rate": 7.28539904481599e-05,
      "loss": 0.0718,
      "step": 1104
    },
    {
      "epoch": 0.8289572393098275,
      "grad_norm": 0.12530259788036346,
      "learning_rate": 7.279573382827662e-05,
      "loss": 0.0498,
      "step": 1105
    },
    {
      "epoch": 0.8297074268567142,
      "grad_norm": 0.13468191027641296,
      "learning_rate": 7.273743811595454e-05,
      "loss": 0.0514,
      "step": 1106
    },
    {
      "epoch": 0.8304576144036009,
      "grad_norm": 0.13877612352371216,
      "learning_rate": 7.267910341116512e-05,
      "loss": 0.0711,
      "step": 1107
    },
    {
      "epoch": 0.8312078019504876,
      "grad_norm": 0.11012528836727142,
      "learning_rate": 7.262072981394656e-05,
      "loss": 0.0458,
      "step": 1108
    },
    {
      "epoch": 0.8319579894973743,
      "grad_norm": 0.22262486815452576,
      "learning_rate": 7.256231742440389e-05,
      "loss": 0.1182,
      "step": 1109
    },
    {
      "epoch": 0.832708177044261,
      "grad_norm": 0.15494659543037415,
      "learning_rate": 7.25038663427086e-05,
      "loss": 0.0766,
      "step": 1110
    },
    {
      "epoch": 0.8334583645911477,
      "grad_norm": 0.14093343913555145,
      "learning_rate": 7.24453766690985e-05,
      "loss": 0.056,
      "step": 1111
    },
    {
      "epoch": 0.8342085521380345,
      "grad_norm": 0.15565334260463715,
      "learning_rate": 7.238684850387765e-05,
      "loss": 0.0565,
      "step": 1112
    },
    {
      "epoch": 0.8349587396849212,
      "grad_norm": 0.13291604816913605,
      "learning_rate": 7.232828194741611e-05,
      "loss": 0.0615,
      "step": 1113
    },
    {
      "epoch": 0.8357089272318079,
      "grad_norm": 0.1696081906557083,
      "learning_rate": 7.226967710014971e-05,
      "loss": 0.0601,
      "step": 1114
    },
    {
      "epoch": 0.8364591147786947,
      "grad_norm": 0.1792924702167511,
      "learning_rate": 7.221103406258003e-05,
      "loss": 0.0558,
      "step": 1115
    },
    {
      "epoch": 0.8372093023255814,
      "grad_norm": 0.11232032626867294,
      "learning_rate": 7.215235293527409e-05,
      "loss": 0.0473,
      "step": 1116
    },
    {
      "epoch": 0.8379594898724682,
      "grad_norm": 0.15998299419879913,
      "learning_rate": 7.209363381886423e-05,
      "loss": 0.0799,
      "step": 1117
    },
    {
      "epoch": 0.8387096774193549,
      "grad_norm": 0.12312288582324982,
      "learning_rate": 7.203487681404798e-05,
      "loss": 0.0385,
      "step": 1118
    },
    {
      "epoch": 0.8394598649662416,
      "grad_norm": 0.14762242138385773,
      "learning_rate": 7.19760820215878e-05,
      "loss": 0.0788,
      "step": 1119
    },
    {
      "epoch": 0.8402100525131283,
      "grad_norm": 0.15391835570335388,
      "learning_rate": 7.191724954231098e-05,
      "loss": 0.0556,
      "step": 1120
    },
    {
      "epoch": 0.840960240060015,
      "grad_norm": 0.1032281443476677,
      "learning_rate": 7.185837947710943e-05,
      "loss": 0.0601,
      "step": 1121
    },
    {
      "epoch": 0.8417104276069017,
      "grad_norm": 0.11759169399738312,
      "learning_rate": 7.17994719269395e-05,
      "loss": 0.0548,
      "step": 1122
    },
    {
      "epoch": 0.8424606151537885,
      "grad_norm": 0.15175266563892365,
      "learning_rate": 7.174052699282183e-05,
      "loss": 0.0829,
      "step": 1123
    },
    {
      "epoch": 0.8432108027006752,
      "grad_norm": 0.26938396692276,
      "learning_rate": 7.168154477584123e-05,
      "loss": 0.077,
      "step": 1124
    },
    {
      "epoch": 0.8439609902475619,
      "grad_norm": 0.13639144599437714,
      "learning_rate": 7.162252537714633e-05,
      "loss": 0.0516,
      "step": 1125
    },
    {
      "epoch": 0.8447111777944486,
      "grad_norm": 0.14507851004600525,
      "learning_rate": 7.156346889794962e-05,
      "loss": 0.0916,
      "step": 1126
    },
    {
      "epoch": 0.8454613653413353,
      "grad_norm": 0.21296165883541107,
      "learning_rate": 7.150437543952715e-05,
      "loss": 0.0742,
      "step": 1127
    },
    {
      "epoch": 0.846211552888222,
      "grad_norm": 0.09235338866710663,
      "learning_rate": 7.144524510321837e-05,
      "loss": 0.0501,
      "step": 1128
    },
    {
      "epoch": 0.8469617404351087,
      "grad_norm": 0.19731169939041138,
      "learning_rate": 7.138607799042598e-05,
      "loss": 0.0772,
      "step": 1129
    },
    {
      "epoch": 0.8477119279819955,
      "grad_norm": 0.1795894056558609,
      "learning_rate": 7.132687420261576e-05,
      "loss": 0.0757,
      "step": 1130
    },
    {
      "epoch": 0.8484621155288822,
      "grad_norm": 0.14649488031864166,
      "learning_rate": 7.126763384131638e-05,
      "loss": 0.055,
      "step": 1131
    },
    {
      "epoch": 0.8492123030757689,
      "grad_norm": 0.12544019520282745,
      "learning_rate": 7.120835700811923e-05,
      "loss": 0.0734,
      "step": 1132
    },
    {
      "epoch": 0.8499624906226556,
      "grad_norm": 0.1415441930294037,
      "learning_rate": 7.114904380467823e-05,
      "loss": 0.0542,
      "step": 1133
    },
    {
      "epoch": 0.8507126781695424,
      "grad_norm": 0.12028104066848755,
      "learning_rate": 7.108969433270968e-05,
      "loss": 0.0491,
      "step": 1134
    },
    {
      "epoch": 0.8514628657164292,
      "grad_norm": 0.12645025551319122,
      "learning_rate": 7.10303086939921e-05,
      "loss": 0.0677,
      "step": 1135
    },
    {
      "epoch": 0.8522130532633159,
      "grad_norm": 0.12939628958702087,
      "learning_rate": 7.097088699036598e-05,
      "loss": 0.089,
      "step": 1136
    },
    {
      "epoch": 0.8529632408102026,
      "grad_norm": 0.14228011667728424,
      "learning_rate": 7.091142932373371e-05,
      "loss": 0.066,
      "step": 1137
    },
    {
      "epoch": 0.8537134283570893,
      "grad_norm": 0.14560295641422272,
      "learning_rate": 7.085193579605935e-05,
      "loss": 0.0475,
      "step": 1138
    },
    {
      "epoch": 0.854463615903976,
      "grad_norm": 0.09960725158452988,
      "learning_rate": 7.079240650936843e-05,
      "loss": 0.0587,
      "step": 1139
    },
    {
      "epoch": 0.8552138034508627,
      "grad_norm": 0.13683639466762543,
      "learning_rate": 7.07328415657478e-05,
      "loss": 0.0502,
      "step": 1140
    },
    {
      "epoch": 0.8559639909977494,
      "grad_norm": 0.17216654121875763,
      "learning_rate": 7.067324106734548e-05,
      "loss": 0.0559,
      "step": 1141
    },
    {
      "epoch": 0.8567141785446362,
      "grad_norm": 0.16215908527374268,
      "learning_rate": 7.061360511637045e-05,
      "loss": 0.0453,
      "step": 1142
    },
    {
      "epoch": 0.8574643660915229,
      "grad_norm": 0.13577274978160858,
      "learning_rate": 7.055393381509253e-05,
      "loss": 0.0711,
      "step": 1143
    },
    {
      "epoch": 0.8582145536384096,
      "grad_norm": 0.14832454919815063,
      "learning_rate": 7.049422726584206e-05,
      "loss": 0.0644,
      "step": 1144
    },
    {
      "epoch": 0.8589647411852963,
      "grad_norm": 0.17870298027992249,
      "learning_rate": 7.043448557100995e-05,
      "loss": 0.0892,
      "step": 1145
    },
    {
      "epoch": 0.859714928732183,
      "grad_norm": 0.13199222087860107,
      "learning_rate": 7.037470883304731e-05,
      "loss": 0.0417,
      "step": 1146
    },
    {
      "epoch": 0.8604651162790697,
      "grad_norm": 0.11214780807495117,
      "learning_rate": 7.031489715446535e-05,
      "loss": 0.0474,
      "step": 1147
    },
    {
      "epoch": 0.8612153038259565,
      "grad_norm": 0.179098442196846,
      "learning_rate": 7.02550506378352e-05,
      "loss": 0.0788,
      "step": 1148
    },
    {
      "epoch": 0.8619654913728432,
      "grad_norm": 0.1804032027721405,
      "learning_rate": 7.019516938578777e-05,
      "loss": 0.0571,
      "step": 1149
    },
    {
      "epoch": 0.8627156789197299,
      "grad_norm": 0.1622643768787384,
      "learning_rate": 7.013525350101348e-05,
      "loss": 0.0626,
      "step": 1150
    },
    {
      "epoch": 0.8634658664666166,
      "grad_norm": 0.12790507078170776,
      "learning_rate": 7.00753030862622e-05,
      "loss": 0.0634,
      "step": 1151
    },
    {
      "epoch": 0.8642160540135033,
      "grad_norm": 0.11260277032852173,
      "learning_rate": 7.001531824434299e-05,
      "loss": 0.0523,
      "step": 1152
    },
    {
      "epoch": 0.8649662415603901,
      "grad_norm": 0.1972770243883133,
      "learning_rate": 6.995529907812391e-05,
      "loss": 0.0786,
      "step": 1153
    },
    {
      "epoch": 0.8657164291072769,
      "grad_norm": 0.14673151075839996,
      "learning_rate": 6.989524569053196e-05,
      "loss": 0.0658,
      "step": 1154
    },
    {
      "epoch": 0.8664666166541636,
      "grad_norm": 0.1285582333803177,
      "learning_rate": 6.983515818455275e-05,
      "loss": 0.0459,
      "step": 1155
    },
    {
      "epoch": 0.8672168042010503,
      "grad_norm": 0.1427638828754425,
      "learning_rate": 6.977503666323048e-05,
      "loss": 0.0809,
      "step": 1156
    },
    {
      "epoch": 0.867966991747937,
      "grad_norm": 0.19241690635681152,
      "learning_rate": 6.971488122966758e-05,
      "loss": 0.084,
      "step": 1157
    },
    {
      "epoch": 0.8687171792948237,
      "grad_norm": 0.1290270984172821,
      "learning_rate": 6.965469198702475e-05,
      "loss": 0.0541,
      "step": 1158
    },
    {
      "epoch": 0.8694673668417104,
      "grad_norm": 0.12479504197835922,
      "learning_rate": 6.95944690385206e-05,
      "loss": 0.0617,
      "step": 1159
    },
    {
      "epoch": 0.8702175543885972,
      "grad_norm": 0.157056525349617,
      "learning_rate": 6.953421248743154e-05,
      "loss": 0.0871,
      "step": 1160
    },
    {
      "epoch": 0.8709677419354839,
      "grad_norm": 0.14428332448005676,
      "learning_rate": 6.947392243709163e-05,
      "loss": 0.0512,
      "step": 1161
    },
    {
      "epoch": 0.8717179294823706,
      "grad_norm": 0.13996249437332153,
      "learning_rate": 6.941359899089238e-05,
      "loss": 0.0559,
      "step": 1162
    },
    {
      "epoch": 0.8724681170292573,
      "grad_norm": 0.17473511397838593,
      "learning_rate": 6.935324225228254e-05,
      "loss": 0.0899,
      "step": 1163
    },
    {
      "epoch": 0.873218304576144,
      "grad_norm": 0.17250919342041016,
      "learning_rate": 6.929285232476797e-05,
      "loss": 0.0793,
      "step": 1164
    },
    {
      "epoch": 0.8739684921230307,
      "grad_norm": 0.16943222284317017,
      "learning_rate": 6.923242931191148e-05,
      "loss": 0.0668,
      "step": 1165
    },
    {
      "epoch": 0.8747186796699175,
      "grad_norm": 0.175174281001091,
      "learning_rate": 6.917197331733257e-05,
      "loss": 0.0795,
      "step": 1166
    },
    {
      "epoch": 0.8754688672168042,
      "grad_norm": 0.16245006024837494,
      "learning_rate": 6.911148444470731e-05,
      "loss": 0.0641,
      "step": 1167
    },
    {
      "epoch": 0.8762190547636909,
      "grad_norm": 0.15257513523101807,
      "learning_rate": 6.905096279776819e-05,
      "loss": 0.0574,
      "step": 1168
    },
    {
      "epoch": 0.8769692423105776,
      "grad_norm": 0.10353401303291321,
      "learning_rate": 6.899040848030384e-05,
      "loss": 0.0571,
      "step": 1169
    },
    {
      "epoch": 0.8777194298574643,
      "grad_norm": 0.11673977226018906,
      "learning_rate": 6.892982159615895e-05,
      "loss": 0.076,
      "step": 1170
    },
    {
      "epoch": 0.878469617404351,
      "grad_norm": 0.17134547233581543,
      "learning_rate": 6.886920224923408e-05,
      "loss": 0.0907,
      "step": 1171
    },
    {
      "epoch": 0.8792198049512379,
      "grad_norm": 0.10142813622951508,
      "learning_rate": 6.880855054348543e-05,
      "loss": 0.0355,
      "step": 1172
    },
    {
      "epoch": 0.8799699924981246,
      "grad_norm": 0.12058088928461075,
      "learning_rate": 6.874786658292472e-05,
      "loss": 0.0499,
      "step": 1173
    },
    {
      "epoch": 0.8807201800450113,
      "grad_norm": 0.14051474630832672,
      "learning_rate": 6.868715047161896e-05,
      "loss": 0.0645,
      "step": 1174
    },
    {
      "epoch": 0.881470367591898,
      "grad_norm": 0.20736153423786163,
      "learning_rate": 6.862640231369029e-05,
      "loss": 0.0882,
      "step": 1175
    },
    {
      "epoch": 0.8822205551387847,
      "grad_norm": 0.13780175149440765,
      "learning_rate": 6.856562221331581e-05,
      "loss": 0.0587,
      "step": 1176
    },
    {
      "epoch": 0.8829707426856714,
      "grad_norm": 0.1332911252975464,
      "learning_rate": 6.850481027472743e-05,
      "loss": 0.071,
      "step": 1177
    },
    {
      "epoch": 0.8837209302325582,
      "grad_norm": 0.09639962762594223,
      "learning_rate": 6.844396660221164e-05,
      "loss": 0.0465,
      "step": 1178
    },
    {
      "epoch": 0.8844711177794449,
      "grad_norm": 0.124673031270504,
      "learning_rate": 6.838309130010933e-05,
      "loss": 0.0518,
      "step": 1179
    },
    {
      "epoch": 0.8852213053263316,
      "grad_norm": 0.12697821855545044,
      "learning_rate": 6.832218447281566e-05,
      "loss": 0.0516,
      "step": 1180
    },
    {
      "epoch": 0.8859714928732183,
      "grad_norm": 0.12598365545272827,
      "learning_rate": 6.826124622477981e-05,
      "loss": 0.072,
      "step": 1181
    },
    {
      "epoch": 0.886721680420105,
      "grad_norm": 0.18293868005275726,
      "learning_rate": 6.820027666050493e-05,
      "loss": 0.0527,
      "step": 1182
    },
    {
      "epoch": 0.8874718679669917,
      "grad_norm": 0.18333552777767181,
      "learning_rate": 6.813927588454778e-05,
      "loss": 0.0624,
      "step": 1183
    },
    {
      "epoch": 0.8882220555138785,
      "grad_norm": 0.14375266432762146,
      "learning_rate": 6.80782440015187e-05,
      "loss": 0.0648,
      "step": 1184
    },
    {
      "epoch": 0.8889722430607652,
      "grad_norm": 0.16927683353424072,
      "learning_rate": 6.801718111608133e-05,
      "loss": 0.0919,
      "step": 1185
    },
    {
      "epoch": 0.8897224306076519,
      "grad_norm": 0.1668524146080017,
      "learning_rate": 6.795608733295254e-05,
      "loss": 0.0633,
      "step": 1186
    },
    {
      "epoch": 0.8904726181545386,
      "grad_norm": 0.16813072562217712,
      "learning_rate": 6.789496275690215e-05,
      "loss": 0.0599,
      "step": 1187
    },
    {
      "epoch": 0.8912228057014253,
      "grad_norm": 0.1353646218776703,
      "learning_rate": 6.783380749275277e-05,
      "loss": 0.0529,
      "step": 1188
    },
    {
      "epoch": 0.891972993248312,
      "grad_norm": 0.12498775124549866,
      "learning_rate": 6.777262164537966e-05,
      "loss": 0.0537,
      "step": 1189
    },
    {
      "epoch": 0.8927231807951987,
      "grad_norm": 0.13221831619739532,
      "learning_rate": 6.771140531971054e-05,
      "loss": 0.0529,
      "step": 1190
    },
    {
      "epoch": 0.8934733683420856,
      "grad_norm": 0.16475389897823334,
      "learning_rate": 6.765015862072536e-05,
      "loss": 0.0509,
      "step": 1191
    },
    {
      "epoch": 0.8942235558889723,
      "grad_norm": 0.12393538653850555,
      "learning_rate": 6.758888165345619e-05,
      "loss": 0.048,
      "step": 1192
    },
    {
      "epoch": 0.894973743435859,
      "grad_norm": 0.1635979413986206,
      "learning_rate": 6.7527574522987e-05,
      "loss": 0.0661,
      "step": 1193
    },
    {
      "epoch": 0.8957239309827457,
      "grad_norm": 0.18363012373447418,
      "learning_rate": 6.746623733445346e-05,
      "loss": 0.0491,
      "step": 1194
    },
    {
      "epoch": 0.8964741185296324,
      "grad_norm": 0.11310746520757675,
      "learning_rate": 6.740487019304282e-05,
      "loss": 0.034,
      "step": 1195
    },
    {
      "epoch": 0.8972243060765192,
      "grad_norm": 0.09733211249113083,
      "learning_rate": 6.734347320399369e-05,
      "loss": 0.0337,
      "step": 1196
    },
    {
      "epoch": 0.8979744936234059,
      "grad_norm": 0.13894496858119965,
      "learning_rate": 6.728204647259586e-05,
      "loss": 0.0884,
      "step": 1197
    },
    {
      "epoch": 0.8987246811702926,
      "grad_norm": 0.11892880499362946,
      "learning_rate": 6.722059010419013e-05,
      "loss": 0.0565,
      "step": 1198
    },
    {
      "epoch": 0.8994748687171793,
      "grad_norm": 0.1761048585176468,
      "learning_rate": 6.715910420416809e-05,
      "loss": 0.0649,
      "step": 1199
    },
    {
      "epoch": 0.900225056264066,
      "grad_norm": 0.14857003092765808,
      "learning_rate": 6.709758887797205e-05,
      "loss": 0.0632,
      "step": 1200
    },
    {
      "epoch": 0.900225056264066,
      "eval_loss": 0.06695682555437088,
      "eval_runtime": 2.6653,
      "eval_samples_per_second": 20.26,
      "eval_steps_per_second": 5.253,
      "step": 1200
    },
    {
      "epoch": 0.9009752438109527,
      "grad_norm": 0.16598834097385406,
      "learning_rate": 6.703604423109468e-05,
      "loss": 0.0695,
      "step": 1201
    },
    {
      "epoch": 0.9017254313578394,
      "grad_norm": 0.13511797785758972,
      "learning_rate": 6.697447036907904e-05,
      "loss": 0.0508,
      "step": 1202
    },
    {
      "epoch": 0.9024756189047262,
      "grad_norm": 0.08580167591571808,
      "learning_rate": 6.691286739751824e-05,
      "loss": 0.0356,
      "step": 1203
    },
    {
      "epoch": 0.9032258064516129,
      "grad_norm": 0.12339236587285995,
      "learning_rate": 6.685123542205526e-05,
      "loss": 0.0414,
      "step": 1204
    },
    {
      "epoch": 0.9039759939984996,
      "grad_norm": 0.15861302614212036,
      "learning_rate": 6.678957454838292e-05,
      "loss": 0.0902,
      "step": 1205
    },
    {
      "epoch": 0.9047261815453863,
      "grad_norm": 0.15118162333965302,
      "learning_rate": 6.672788488224352e-05,
      "loss": 0.0665,
      "step": 1206
    },
    {
      "epoch": 0.905476369092273,
      "grad_norm": 0.17424209415912628,
      "learning_rate": 6.666616652942878e-05,
      "loss": 0.057,
      "step": 1207
    },
    {
      "epoch": 0.9062265566391597,
      "grad_norm": 0.1301327645778656,
      "learning_rate": 6.660441959577958e-05,
      "loss": 0.0744,
      "step": 1208
    },
    {
      "epoch": 0.9069767441860465,
      "grad_norm": 0.157297745347023,
      "learning_rate": 6.654264418718584e-05,
      "loss": 0.0746,
      "step": 1209
    },
    {
      "epoch": 0.9077269317329333,
      "grad_norm": 0.15477484464645386,
      "learning_rate": 6.64808404095863e-05,
      "loss": 0.069,
      "step": 1210
    },
    {
      "epoch": 0.90847711927982,
      "grad_norm": 0.20592740178108215,
      "learning_rate": 6.641900836896835e-05,
      "loss": 0.0669,
      "step": 1211
    },
    {
      "epoch": 0.9092273068267067,
      "grad_norm": 0.15215907990932465,
      "learning_rate": 6.635714817136785e-05,
      "loss": 0.0547,
      "step": 1212
    },
    {
      "epoch": 0.9099774943735934,
      "grad_norm": 0.13565507531166077,
      "learning_rate": 6.629525992286898e-05,
      "loss": 0.0609,
      "step": 1213
    },
    {
      "epoch": 0.9107276819204801,
      "grad_norm": 0.15176476538181305,
      "learning_rate": 6.623334372960393e-05,
      "loss": 0.0746,
      "step": 1214
    },
    {
      "epoch": 0.9114778694673669,
      "grad_norm": 0.14856791496276855,
      "learning_rate": 6.617139969775292e-05,
      "loss": 0.0648,
      "step": 1215
    },
    {
      "epoch": 0.9122280570142536,
      "grad_norm": 0.17555858194828033,
      "learning_rate": 6.610942793354387e-05,
      "loss": 0.0796,
      "step": 1216
    },
    {
      "epoch": 0.9129782445611403,
      "grad_norm": 0.1357504427433014,
      "learning_rate": 6.604742854325222e-05,
      "loss": 0.0612,
      "step": 1217
    },
    {
      "epoch": 0.913728432108027,
      "grad_norm": 0.14152507483959198,
      "learning_rate": 6.598540163320084e-05,
      "loss": 0.0517,
      "step": 1218
    },
    {
      "epoch": 0.9144786196549137,
      "grad_norm": 0.1403784304857254,
      "learning_rate": 6.592334730975975e-05,
      "loss": 0.0838,
      "step": 1219
    },
    {
      "epoch": 0.9152288072018004,
      "grad_norm": 0.1756121814250946,
      "learning_rate": 6.586126567934605e-05,
      "loss": 0.0949,
      "step": 1220
    },
    {
      "epoch": 0.9159789947486872,
      "grad_norm": 0.12650136649608612,
      "learning_rate": 6.57991568484236e-05,
      "loss": 0.0559,
      "step": 1221
    },
    {
      "epoch": 0.9167291822955739,
      "grad_norm": 0.11986692994832993,
      "learning_rate": 6.573702092350292e-05,
      "loss": 0.0527,
      "step": 1222
    },
    {
      "epoch": 0.9174793698424606,
      "grad_norm": 0.16555757820606232,
      "learning_rate": 6.567485801114099e-05,
      "loss": 0.0614,
      "step": 1223
    },
    {
      "epoch": 0.9182295573893473,
      "grad_norm": 0.08092650026082993,
      "learning_rate": 6.561266821794111e-05,
      "loss": 0.0388,
      "step": 1224
    },
    {
      "epoch": 0.918979744936234,
      "grad_norm": 0.1932453215122223,
      "learning_rate": 6.555045165055263e-05,
      "loss": 0.0803,
      "step": 1225
    },
    {
      "epoch": 0.9197299324831207,
      "grad_norm": 0.16399742662906647,
      "learning_rate": 6.548820841567086e-05,
      "loss": 0.0747,
      "step": 1226
    },
    {
      "epoch": 0.9204801200300075,
      "grad_norm": 0.13225746154785156,
      "learning_rate": 6.54259386200368e-05,
      "loss": 0.0602,
      "step": 1227
    },
    {
      "epoch": 0.9212303075768942,
      "grad_norm": 0.11068938672542572,
      "learning_rate": 6.536364237043703e-05,
      "loss": 0.0504,
      "step": 1228
    },
    {
      "epoch": 0.921980495123781,
      "grad_norm": 0.1256033480167389,
      "learning_rate": 6.530131977370348e-05,
      "loss": 0.05,
      "step": 1229
    },
    {
      "epoch": 0.9227306826706677,
      "grad_norm": 0.13208834826946259,
      "learning_rate": 6.523897093671326e-05,
      "loss": 0.0513,
      "step": 1230
    },
    {
      "epoch": 0.9234808702175544,
      "grad_norm": 0.12550483644008636,
      "learning_rate": 6.51765959663885e-05,
      "loss": 0.0597,
      "step": 1231
    },
    {
      "epoch": 0.9242310577644411,
      "grad_norm": 0.19324685633182526,
      "learning_rate": 6.511419496969612e-05,
      "loss": 0.0752,
      "step": 1232
    },
    {
      "epoch": 0.9249812453113279,
      "grad_norm": 0.1724669337272644,
      "learning_rate": 6.505176805364767e-05,
      "loss": 0.0825,
      "step": 1233
    },
    {
      "epoch": 0.9257314328582146,
      "grad_norm": 0.10835179686546326,
      "learning_rate": 6.498931532529921e-05,
      "loss": 0.0633,
      "step": 1234
    },
    {
      "epoch": 0.9264816204051013,
      "grad_norm": 0.15429191291332245,
      "learning_rate": 6.492683689175098e-05,
      "loss": 0.0492,
      "step": 1235
    },
    {
      "epoch": 0.927231807951988,
      "grad_norm": 0.1609717160463333,
      "learning_rate": 6.486433286014734e-05,
      "loss": 0.0638,
      "step": 1236
    },
    {
      "epoch": 0.9279819954988747,
      "grad_norm": 0.17381221055984497,
      "learning_rate": 6.480180333767658e-05,
      "loss": 0.0784,
      "step": 1237
    },
    {
      "epoch": 0.9287321830457614,
      "grad_norm": 0.14453348517417908,
      "learning_rate": 6.473924843157065e-05,
      "loss": 0.0557,
      "step": 1238
    },
    {
      "epoch": 0.9294823705926482,
      "grad_norm": 0.12073624134063721,
      "learning_rate": 6.467666824910505e-05,
      "loss": 0.0625,
      "step": 1239
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 0.16522017121315002,
      "learning_rate": 6.461406289759862e-05,
      "loss": 0.0746,
      "step": 1240
    },
    {
      "epoch": 0.9309827456864216,
      "grad_norm": 0.17048360407352448,
      "learning_rate": 6.455143248441342e-05,
      "loss": 0.0855,
      "step": 1241
    },
    {
      "epoch": 0.9317329332333083,
      "grad_norm": 0.12252648919820786,
      "learning_rate": 6.44887771169544e-05,
      "loss": 0.0524,
      "step": 1242
    },
    {
      "epoch": 0.932483120780195,
      "grad_norm": 0.14796991646289825,
      "learning_rate": 6.442609690266937e-05,
      "loss": 0.068,
      "step": 1243
    },
    {
      "epoch": 0.9332333083270817,
      "grad_norm": 0.1838654726743698,
      "learning_rate": 6.436339194904872e-05,
      "loss": 0.0478,
      "step": 1244
    },
    {
      "epoch": 0.9339834958739685,
      "grad_norm": 0.12232690304517746,
      "learning_rate": 6.430066236362524e-05,
      "loss": 0.0537,
      "step": 1245
    },
    {
      "epoch": 0.9347336834208552,
      "grad_norm": 0.18929807841777802,
      "learning_rate": 6.423790825397404e-05,
      "loss": 0.0735,
      "step": 1246
    },
    {
      "epoch": 0.9354838709677419,
      "grad_norm": 0.12835638225078583,
      "learning_rate": 6.417512972771219e-05,
      "loss": 0.0472,
      "step": 1247
    },
    {
      "epoch": 0.9362340585146287,
      "grad_norm": 0.3227969706058502,
      "learning_rate": 6.411232689249873e-05,
      "loss": 0.0836,
      "step": 1248
    },
    {
      "epoch": 0.9369842460615154,
      "grad_norm": 0.13637347519397736,
      "learning_rate": 6.40494998560343e-05,
      "loss": 0.0417,
      "step": 1249
    },
    {
      "epoch": 0.9377344336084021,
      "grad_norm": 0.18338750302791595,
      "learning_rate": 6.39866487260611e-05,
      "loss": 0.063,
      "step": 1250
    },
    {
      "epoch": 0.9384846211552889,
      "grad_norm": 0.08587335050106049,
      "learning_rate": 6.392377361036262e-05,
      "loss": 0.0332,
      "step": 1251
    },
    {
      "epoch": 0.9392348087021756,
      "grad_norm": 0.1209770143032074,
      "learning_rate": 6.386087461676351e-05,
      "loss": 0.0523,
      "step": 1252
    },
    {
      "epoch": 0.9399849962490623,
      "grad_norm": 0.11411428451538086,
      "learning_rate": 6.379795185312933e-05,
      "loss": 0.0397,
      "step": 1253
    },
    {
      "epoch": 0.940735183795949,
      "grad_norm": 0.21214881539344788,
      "learning_rate": 6.373500542736643e-05,
      "loss": 0.0745,
      "step": 1254
    },
    {
      "epoch": 0.9414853713428357,
      "grad_norm": 0.1349548101425171,
      "learning_rate": 6.367203544742171e-05,
      "loss": 0.0468,
      "step": 1255
    },
    {
      "epoch": 0.9422355588897224,
      "grad_norm": 0.188161700963974,
      "learning_rate": 6.360904202128252e-05,
      "loss": 0.0771,
      "step": 1256
    },
    {
      "epoch": 0.9429857464366092,
      "grad_norm": 0.2221253365278244,
      "learning_rate": 6.354602525697638e-05,
      "loss": 0.0785,
      "step": 1257
    },
    {
      "epoch": 0.9437359339834959,
      "grad_norm": 0.19272452592849731,
      "learning_rate": 6.348298526257082e-05,
      "loss": 0.0738,
      "step": 1258
    },
    {
      "epoch": 0.9444861215303826,
      "grad_norm": 0.1681603491306305,
      "learning_rate": 6.341992214617323e-05,
      "loss": 0.0609,
      "step": 1259
    },
    {
      "epoch": 0.9452363090772693,
      "grad_norm": 0.1713087409734726,
      "learning_rate": 6.335683601593062e-05,
      "loss": 0.0962,
      "step": 1260
    },
    {
      "epoch": 0.945986496624156,
      "grad_norm": 0.13711421191692352,
      "learning_rate": 6.329372698002954e-05,
      "loss": 0.0569,
      "step": 1261
    },
    {
      "epoch": 0.9467366841710427,
      "grad_norm": 0.3579675555229187,
      "learning_rate": 6.323059514669571e-05,
      "loss": 0.1269,
      "step": 1262
    },
    {
      "epoch": 0.9474868717179294,
      "grad_norm": 0.20477552711963654,
      "learning_rate": 6.316744062419409e-05,
      "loss": 0.0785,
      "step": 1263
    },
    {
      "epoch": 0.9482370592648162,
      "grad_norm": 0.166034996509552,
      "learning_rate": 6.310426352082838e-05,
      "loss": 0.0651,
      "step": 1264
    },
    {
      "epoch": 0.9489872468117029,
      "grad_norm": 0.1207178607583046,
      "learning_rate": 6.304106394494116e-05,
      "loss": 0.0461,
      "step": 1265
    },
    {
      "epoch": 0.9497374343585896,
      "grad_norm": 0.21342095732688904,
      "learning_rate": 6.297784200491343e-05,
      "loss": 0.0815,
      "step": 1266
    },
    {
      "epoch": 0.9504876219054764,
      "grad_norm": 0.11947363615036011,
      "learning_rate": 6.291459780916463e-05,
      "loss": 0.0342,
      "step": 1267
    },
    {
      "epoch": 0.9512378094523631,
      "grad_norm": 0.15550976991653442,
      "learning_rate": 6.285133146615228e-05,
      "loss": 0.0526,
      "step": 1268
    },
    {
      "epoch": 0.9519879969992499,
      "grad_norm": 0.14392493665218353,
      "learning_rate": 6.278804308437198e-05,
      "loss": 0.0523,
      "step": 1269
    },
    {
      "epoch": 0.9527381845461366,
      "grad_norm": 0.13637441396713257,
      "learning_rate": 6.272473277235703e-05,
      "loss": 0.0539,
      "step": 1270
    },
    {
      "epoch": 0.9534883720930233,
      "grad_norm": 0.20482169091701508,
      "learning_rate": 6.266140063867843e-05,
      "loss": 0.0784,
      "step": 1271
    },
    {
      "epoch": 0.95423855963991,
      "grad_norm": 0.19494204223155975,
      "learning_rate": 6.25980467919445e-05,
      "loss": 0.115,
      "step": 1272
    },
    {
      "epoch": 0.9549887471867967,
      "grad_norm": 0.16727079451084137,
      "learning_rate": 6.253467134080088e-05,
      "loss": 0.0623,
      "step": 1273
    },
    {
      "epoch": 0.9557389347336834,
      "grad_norm": 0.13359226286411285,
      "learning_rate": 6.247127439393023e-05,
      "loss": 0.0501,
      "step": 1274
    },
    {
      "epoch": 0.9564891222805701,
      "grad_norm": 0.20405220985412598,
      "learning_rate": 6.240785606005206e-05,
      "loss": 0.0644,
      "step": 1275
    },
    {
      "epoch": 0.9572393098274569,
      "grad_norm": 0.19779805839061737,
      "learning_rate": 6.234441644792256e-05,
      "loss": 0.0424,
      "step": 1276
    },
    {
      "epoch": 0.9579894973743436,
      "grad_norm": 0.1623539924621582,
      "learning_rate": 6.228095566633443e-05,
      "loss": 0.0743,
      "step": 1277
    },
    {
      "epoch": 0.9587396849212303,
      "grad_norm": 0.1125299483537674,
      "learning_rate": 6.221747382411667e-05,
      "loss": 0.0498,
      "step": 1278
    },
    {
      "epoch": 0.959489872468117,
      "grad_norm": 0.15883377194404602,
      "learning_rate": 6.215397103013436e-05,
      "loss": 0.056,
      "step": 1279
    },
    {
      "epoch": 0.9602400600150037,
      "grad_norm": 0.13185837864875793,
      "learning_rate": 6.209044739328858e-05,
      "loss": 0.0581,
      "step": 1280
    },
    {
      "epoch": 0.9609902475618904,
      "grad_norm": 0.14542366564273834,
      "learning_rate": 6.202690302251606e-05,
      "loss": 0.0692,
      "step": 1281
    },
    {
      "epoch": 0.9617404351087772,
      "grad_norm": 0.12652628123760223,
      "learning_rate": 6.196333802678918e-05,
      "loss": 0.0656,
      "step": 1282
    },
    {
      "epoch": 0.9624906226556639,
      "grad_norm": 0.26526379585266113,
      "learning_rate": 6.189975251511562e-05,
      "loss": 0.0915,
      "step": 1283
    },
    {
      "epoch": 0.9632408102025506,
      "grad_norm": 0.18798717856407166,
      "learning_rate": 6.18361465965383e-05,
      "loss": 0.0685,
      "step": 1284
    },
    {
      "epoch": 0.9639909977494373,
      "grad_norm": 0.12187514454126358,
      "learning_rate": 6.177252038013509e-05,
      "loss": 0.0455,
      "step": 1285
    },
    {
      "epoch": 0.9647411852963241,
      "grad_norm": 0.09327728301286697,
      "learning_rate": 6.170887397501868e-05,
      "loss": 0.0389,
      "step": 1286
    },
    {
      "epoch": 0.9654913728432108,
      "grad_norm": 0.2176682949066162,
      "learning_rate": 6.16452074903364e-05,
      "loss": 0.0718,
      "step": 1287
    },
    {
      "epoch": 0.9662415603900976,
      "grad_norm": 0.24573443830013275,
      "learning_rate": 6.158152103527e-05,
      "loss": 0.1097,
      "step": 1288
    },
    {
      "epoch": 0.9669917479369843,
      "grad_norm": 0.17032325267791748,
      "learning_rate": 6.151781471903548e-05,
      "loss": 0.076,
      "step": 1289
    },
    {
      "epoch": 0.967741935483871,
      "grad_norm": 0.22234180569648743,
      "learning_rate": 6.14540886508829e-05,
      "loss": 0.1093,
      "step": 1290
    },
    {
      "epoch": 0.9684921230307577,
      "grad_norm": 0.15120472013950348,
      "learning_rate": 6.139034294009617e-05,
      "loss": 0.0637,
      "step": 1291
    },
    {
      "epoch": 0.9692423105776444,
      "grad_norm": 0.14784853160381317,
      "learning_rate": 6.132657769599293e-05,
      "loss": 0.0476,
      "step": 1292
    },
    {
      "epoch": 0.9699924981245311,
      "grad_norm": 0.20848771929740906,
      "learning_rate": 6.126279302792429e-05,
      "loss": 0.086,
      "step": 1293
    },
    {
      "epoch": 0.9707426856714179,
      "grad_norm": 0.12628328800201416,
      "learning_rate": 6.119898904527468e-05,
      "loss": 0.0324,
      "step": 1294
    },
    {
      "epoch": 0.9714928732183046,
      "grad_norm": 0.16405990719795227,
      "learning_rate": 6.113516585746164e-05,
      "loss": 0.1027,
      "step": 1295
    },
    {
      "epoch": 0.9722430607651913,
      "grad_norm": 0.1743471920490265,
      "learning_rate": 6.107132357393563e-05,
      "loss": 0.0805,
      "step": 1296
    },
    {
      "epoch": 0.972993248312078,
      "grad_norm": 0.12977734208106995,
      "learning_rate": 6.100746230417993e-05,
      "loss": 0.0467,
      "step": 1297
    },
    {
      "epoch": 0.9737434358589647,
      "grad_norm": 0.1675756424665451,
      "learning_rate": 6.0943582157710285e-05,
      "loss": 0.0446,
      "step": 1298
    },
    {
      "epoch": 0.9744936234058514,
      "grad_norm": 0.23262391984462738,
      "learning_rate": 6.0879683244074894e-05,
      "loss": 0.0699,
      "step": 1299
    },
    {
      "epoch": 0.9752438109527382,
      "grad_norm": 0.12139663100242615,
      "learning_rate": 6.0815765672854065e-05,
      "loss": 0.0521,
      "step": 1300
    },
    {
      "epoch": 0.9759939984996249,
      "grad_norm": 0.12815728783607483,
      "learning_rate": 6.0751829553660155e-05,
      "loss": 0.044,
      "step": 1301
    },
    {
      "epoch": 0.9767441860465116,
      "grad_norm": 0.1259387880563736,
      "learning_rate": 6.06878749961373e-05,
      "loss": 0.0634,
      "step": 1302
    },
    {
      "epoch": 0.9774943735933983,
      "grad_norm": 0.1548275500535965,
      "learning_rate": 6.0623902109961295e-05,
      "loss": 0.0569,
      "step": 1303
    },
    {
      "epoch": 0.978244561140285,
      "grad_norm": 0.25065237283706665,
      "learning_rate": 6.055991100483932e-05,
      "loss": 0.0915,
      "step": 1304
    },
    {
      "epoch": 0.9789947486871718,
      "grad_norm": 0.18947431445121765,
      "learning_rate": 6.0495901790509836e-05,
      "loss": 0.08,
      "step": 1305
    },
    {
      "epoch": 0.9797449362340586,
      "grad_norm": 0.1380397528409958,
      "learning_rate": 6.043187457674231e-05,
      "loss": 0.0528,
      "step": 1306
    },
    {
      "epoch": 0.9804951237809453,
      "grad_norm": 0.1665130853652954,
      "learning_rate": 6.0367829473337136e-05,
      "loss": 0.0752,
      "step": 1307
    },
    {
      "epoch": 0.981245311327832,
      "grad_norm": 0.11333315819501877,
      "learning_rate": 6.0303766590125365e-05,
      "loss": 0.0368,
      "step": 1308
    },
    {
      "epoch": 0.9819954988747187,
      "grad_norm": 0.1431434005498886,
      "learning_rate": 6.02396860369685e-05,
      "loss": 0.0669,
      "step": 1309
    },
    {
      "epoch": 0.9827456864216054,
      "grad_norm": 0.1723565310239792,
      "learning_rate": 6.0175587923758416e-05,
      "loss": 0.0724,
      "step": 1310
    },
    {
      "epoch": 0.9834958739684921,
      "grad_norm": 0.15027496218681335,
      "learning_rate": 6.0111472360417044e-05,
      "loss": 0.0457,
      "step": 1311
    },
    {
      "epoch": 0.9842460615153789,
      "grad_norm": 0.169467493891716,
      "learning_rate": 6.004733945689628e-05,
      "loss": 0.0643,
      "step": 1312
    },
    {
      "epoch": 0.9849962490622656,
      "grad_norm": 0.14806106686592102,
      "learning_rate": 5.998318932317771e-05,
      "loss": 0.0707,
      "step": 1313
    },
    {
      "epoch": 0.9857464366091523,
      "grad_norm": 0.15697737038135529,
      "learning_rate": 5.991902206927252e-05,
      "loss": 0.0622,
      "step": 1314
    },
    {
      "epoch": 0.986496624156039,
      "grad_norm": 0.10275126993656158,
      "learning_rate": 5.985483780522122e-05,
      "loss": 0.0465,
      "step": 1315
    },
    {
      "epoch": 0.9872468117029257,
      "grad_norm": 0.12026581913232803,
      "learning_rate": 5.9790636641093523e-05,
      "loss": 0.0379,
      "step": 1316
    },
    {
      "epoch": 0.9879969992498124,
      "grad_norm": 0.16514074802398682,
      "learning_rate": 5.972641868698805e-05,
      "loss": 0.0525,
      "step": 1317
    },
    {
      "epoch": 0.9887471867966992,
      "grad_norm": 0.18575632572174072,
      "learning_rate": 5.966218405303234e-05,
      "loss": 0.073,
      "step": 1318
    },
    {
      "epoch": 0.9894973743435859,
      "grad_norm": 0.12170587480068207,
      "learning_rate": 5.959793284938242e-05,
      "loss": 0.0568,
      "step": 1319
    },
    {
      "epoch": 0.9902475618904726,
      "grad_norm": 0.18904197216033936,
      "learning_rate": 5.953366518622279e-05,
      "loss": 0.0771,
      "step": 1320
    },
    {
      "epoch": 0.9909977494373593,
      "grad_norm": 0.13215796649456024,
      "learning_rate": 5.946938117376616e-05,
      "loss": 0.0524,
      "step": 1321
    },
    {
      "epoch": 0.991747936984246,
      "grad_norm": 0.13645458221435547,
      "learning_rate": 5.940508092225328e-05,
      "loss": 0.0735,
      "step": 1322
    },
    {
      "epoch": 0.9924981245311327,
      "grad_norm": 0.16823965311050415,
      "learning_rate": 5.9340764541952755e-05,
      "loss": 0.0525,
      "step": 1323
    },
    {
      "epoch": 0.9932483120780196,
      "grad_norm": 0.13981732726097107,
      "learning_rate": 5.9276432143160835e-05,
      "loss": 0.0487,
      "step": 1324
    },
    {
      "epoch": 0.9939984996249063,
      "grad_norm": 0.1347687840461731,
      "learning_rate": 5.921208383620126e-05,
      "loss": 0.0375,
      "step": 1325
    },
    {
      "epoch": 0.994748687171793,
      "grad_norm": 0.2246084362268448,
      "learning_rate": 5.9147719731425034e-05,
      "loss": 0.0788,
      "step": 1326
    },
    {
      "epoch": 0.9954988747186797,
      "grad_norm": 0.13317465782165527,
      "learning_rate": 5.908333993921027e-05,
      "loss": 0.0718,
      "step": 1327
    },
    {
      "epoch": 0.9962490622655664,
      "grad_norm": 0.11800669133663177,
      "learning_rate": 5.901894456996196e-05,
      "loss": 0.053,
      "step": 1328
    },
    {
      "epoch": 0.9969992498124531,
      "grad_norm": 0.23010219633579254,
      "learning_rate": 5.895453373411182e-05,
      "loss": 0.1,
      "step": 1329
    },
    {
      "epoch": 0.9977494373593399,
      "grad_norm": 0.2343241274356842,
      "learning_rate": 5.889010754211809e-05,
      "loss": 0.0774,
      "step": 1330
    },
    {
      "epoch": 0.9984996249062266,
      "grad_norm": 0.16504386067390442,
      "learning_rate": 5.882566610446534e-05,
      "loss": 0.0441,
      "step": 1331
    },
    {
      "epoch": 0.9992498124531133,
      "grad_norm": 0.18011681735515594,
      "learning_rate": 5.8761209531664306e-05,
      "loss": 0.0792,
      "step": 1332
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1674385368824005,
      "learning_rate": 5.869673793425168e-05,
      "loss": 0.067,
      "step": 1333
    },
    {
      "epoch": 1.0007501875468867,
      "grad_norm": 0.15250438451766968,
      "learning_rate": 5.863225142278985e-05,
      "loss": 0.0647,
      "step": 1334
    },
    {
      "epoch": 1.0015003750937734,
      "grad_norm": 0.11759188026189804,
      "learning_rate": 5.856775010786687e-05,
      "loss": 0.0414,
      "step": 1335
    },
    {
      "epoch": 1.0022505626406601,
      "grad_norm": 0.13916079699993134,
      "learning_rate": 5.850323410009614e-05,
      "loss": 0.0442,
      "step": 1336
    },
    {
      "epoch": 1.0030007501875469,
      "grad_norm": 0.17896828055381775,
      "learning_rate": 5.8438703510116256e-05,
      "loss": 0.0503,
      "step": 1337
    },
    {
      "epoch": 1.0037509377344336,
      "grad_norm": 0.1747381091117859,
      "learning_rate": 5.8374158448590823e-05,
      "loss": 0.0475,
      "step": 1338
    },
    {
      "epoch": 1.0045011252813203,
      "grad_norm": 0.12167813628911972,
      "learning_rate": 5.830959902620826e-05,
      "loss": 0.0425,
      "step": 1339
    },
    {
      "epoch": 1.005251312828207,
      "grad_norm": 0.13961336016654968,
      "learning_rate": 5.824502535368164e-05,
      "loss": 0.048,
      "step": 1340
    },
    {
      "epoch": 1.0060015003750937,
      "grad_norm": 0.1229967400431633,
      "learning_rate": 5.818043754174843e-05,
      "loss": 0.0498,
      "step": 1341
    },
    {
      "epoch": 1.0067516879219804,
      "grad_norm": 0.20583300292491913,
      "learning_rate": 5.81158357011704e-05,
      "loss": 0.0608,
      "step": 1342
    },
    {
      "epoch": 1.0075018754688672,
      "grad_norm": 0.07669761776924133,
      "learning_rate": 5.80512199427333e-05,
      "loss": 0.0243,
      "step": 1343
    },
    {
      "epoch": 1.0082520630157539,
      "grad_norm": 0.11683350801467896,
      "learning_rate": 5.798659037724683e-05,
      "loss": 0.0374,
      "step": 1344
    },
    {
      "epoch": 1.0090022505626406,
      "grad_norm": 0.137917622923851,
      "learning_rate": 5.792194711554429e-05,
      "loss": 0.0536,
      "step": 1345
    },
    {
      "epoch": 1.0097524381095273,
      "grad_norm": 0.1384626179933548,
      "learning_rate": 5.7857290268482555e-05,
      "loss": 0.0779,
      "step": 1346
    },
    {
      "epoch": 1.010502625656414,
      "grad_norm": 0.12315264344215393,
      "learning_rate": 5.779261994694173e-05,
      "loss": 0.0271,
      "step": 1347
    },
    {
      "epoch": 1.0112528132033007,
      "grad_norm": 0.23275774717330933,
      "learning_rate": 5.772793626182506e-05,
      "loss": 0.0513,
      "step": 1348
    },
    {
      "epoch": 1.0120030007501875,
      "grad_norm": 0.12899671494960785,
      "learning_rate": 5.766323932405866e-05,
      "loss": 0.0489,
      "step": 1349
    },
    {
      "epoch": 1.0127531882970742,
      "grad_norm": 0.1920711249113083,
      "learning_rate": 5.7598529244591436e-05,
      "loss": 0.0495,
      "step": 1350
    },
    {
      "epoch": 1.0135033758439609,
      "grad_norm": 0.20560826361179352,
      "learning_rate": 5.7533806134394806e-05,
      "loss": 0.0483,
      "step": 1351
    },
    {
      "epoch": 1.0142535633908478,
      "grad_norm": 0.11408060789108276,
      "learning_rate": 5.746907010446252e-05,
      "loss": 0.0394,
      "step": 1352
    },
    {
      "epoch": 1.0150037509377345,
      "grad_norm": 0.1268560141324997,
      "learning_rate": 5.740432126581049e-05,
      "loss": 0.0371,
      "step": 1353
    },
    {
      "epoch": 1.0157539384846213,
      "grad_norm": 0.11626963317394257,
      "learning_rate": 5.73395597294766e-05,
      "loss": 0.0246,
      "step": 1354
    },
    {
      "epoch": 1.016504126031508,
      "grad_norm": 0.10146208852529526,
      "learning_rate": 5.727478560652053e-05,
      "loss": 0.0364,
      "step": 1355
    },
    {
      "epoch": 1.0172543135783947,
      "grad_norm": 0.17881205677986145,
      "learning_rate": 5.7209999008023496e-05,
      "loss": 0.0805,
      "step": 1356
    },
    {
      "epoch": 1.0180045011252814,
      "grad_norm": 0.16511453688144684,
      "learning_rate": 5.7145200045088156e-05,
      "loss": 0.0626,
      "step": 1357
    },
    {
      "epoch": 1.0187546886721681,
      "grad_norm": 0.11539970338344574,
      "learning_rate": 5.7080388828838324e-05,
      "loss": 0.0403,
      "step": 1358
    },
    {
      "epoch": 1.0195048762190548,
      "grad_norm": 0.17582516372203827,
      "learning_rate": 5.701556547041888e-05,
      "loss": 0.0615,
      "step": 1359
    },
    {
      "epoch": 1.0202550637659416,
      "grad_norm": 0.2376180738210678,
      "learning_rate": 5.695073008099547e-05,
      "loss": 0.0491,
      "step": 1360
    },
    {
      "epoch": 1.0210052513128283,
      "grad_norm": 0.2256619781255722,
      "learning_rate": 5.688588277175444e-05,
      "loss": 0.0403,
      "step": 1361
    },
    {
      "epoch": 1.021755438859715,
      "grad_norm": 0.14331862330436707,
      "learning_rate": 5.6821023653902517e-05,
      "loss": 0.0411,
      "step": 1362
    },
    {
      "epoch": 1.0225056264066017,
      "grad_norm": 0.12017953395843506,
      "learning_rate": 5.675615283866671e-05,
      "loss": 0.0421,
      "step": 1363
    },
    {
      "epoch": 1.0232558139534884,
      "grad_norm": 0.16904471814632416,
      "learning_rate": 5.669127043729406e-05,
      "loss": 0.0608,
      "step": 1364
    },
    {
      "epoch": 1.0240060015003751,
      "grad_norm": 0.13869859278202057,
      "learning_rate": 5.662637656105152e-05,
      "loss": 0.0443,
      "step": 1365
    },
    {
      "epoch": 1.0247561890472618,
      "grad_norm": 0.1698876917362213,
      "learning_rate": 5.6561471321225676e-05,
      "loss": 0.0489,
      "step": 1366
    },
    {
      "epoch": 1.0255063765941486,
      "grad_norm": 0.2037367969751358,
      "learning_rate": 5.649655482912265e-05,
      "loss": 0.0716,
      "step": 1367
    },
    {
      "epoch": 1.0262565641410353,
      "grad_norm": 0.1325848251581192,
      "learning_rate": 5.6431627196067816e-05,
      "loss": 0.0332,
      "step": 1368
    },
    {
      "epoch": 1.027006751687922,
      "grad_norm": 0.13476696610450745,
      "learning_rate": 5.636668853340567e-05,
      "loss": 0.0397,
      "step": 1369
    },
    {
      "epoch": 1.0277569392348087,
      "grad_norm": 0.0886571854352951,
      "learning_rate": 5.6301738952499636e-05,
      "loss": 0.0344,
      "step": 1370
    },
    {
      "epoch": 1.0285071267816954,
      "grad_norm": 0.1936550736427307,
      "learning_rate": 5.623677856473183e-05,
      "loss": 0.0761,
      "step": 1371
    },
    {
      "epoch": 1.0292573143285821,
      "grad_norm": 0.16046838462352753,
      "learning_rate": 5.617180748150295e-05,
      "loss": 0.0409,
      "step": 1372
    },
    {
      "epoch": 1.0300075018754689,
      "grad_norm": 0.10154321044683456,
      "learning_rate": 5.6106825814231953e-05,
      "loss": 0.0335,
      "step": 1373
    },
    {
      "epoch": 1.0307576894223556,
      "grad_norm": 0.2043825089931488,
      "learning_rate": 5.604183367435606e-05,
      "loss": 0.0494,
      "step": 1374
    },
    {
      "epoch": 1.0315078769692423,
      "grad_norm": 0.21420907974243164,
      "learning_rate": 5.597683117333036e-05,
      "loss": 0.0599,
      "step": 1375
    },
    {
      "epoch": 1.032258064516129,
      "grad_norm": 0.16880439221858978,
      "learning_rate": 5.591181842262776e-05,
      "loss": 0.0421,
      "step": 1376
    },
    {
      "epoch": 1.0330082520630157,
      "grad_norm": 0.2568267285823822,
      "learning_rate": 5.584679553373869e-05,
      "loss": 0.0533,
      "step": 1377
    },
    {
      "epoch": 1.0337584396099024,
      "grad_norm": 0.12662667036056519,
      "learning_rate": 5.578176261817104e-05,
      "loss": 0.0405,
      "step": 1378
    },
    {
      "epoch": 1.0345086271567892,
      "grad_norm": 0.17100553214550018,
      "learning_rate": 5.571671978744983e-05,
      "loss": 0.0481,
      "step": 1379
    },
    {
      "epoch": 1.0352588147036759,
      "grad_norm": 0.13645890355110168,
      "learning_rate": 5.565166715311711e-05,
      "loss": 0.0419,
      "step": 1380
    },
    {
      "epoch": 1.0360090022505626,
      "grad_norm": 0.10729946196079254,
      "learning_rate": 5.558660482673177e-05,
      "loss": 0.0393,
      "step": 1381
    },
    {
      "epoch": 1.0367591897974493,
      "grad_norm": 0.15839707851409912,
      "learning_rate": 5.552153291986927e-05,
      "loss": 0.0487,
      "step": 1382
    },
    {
      "epoch": 1.037509377344336,
      "grad_norm": 0.12748658657073975,
      "learning_rate": 5.5456451544121523e-05,
      "loss": 0.0492,
      "step": 1383
    },
    {
      "epoch": 1.0382595648912227,
      "grad_norm": 0.19807954132556915,
      "learning_rate": 5.5391360811096684e-05,
      "loss": 0.0418,
      "step": 1384
    },
    {
      "epoch": 1.0390097524381094,
      "grad_norm": 0.12736938893795013,
      "learning_rate": 5.5326260832418955e-05,
      "loss": 0.0441,
      "step": 1385
    },
    {
      "epoch": 1.0397599399849962,
      "grad_norm": 0.16423308849334717,
      "learning_rate": 5.526115171972838e-05,
      "loss": 0.0501,
      "step": 1386
    },
    {
      "epoch": 1.0405101275318829,
      "grad_norm": 0.15221011638641357,
      "learning_rate": 5.5196033584680675e-05,
      "loss": 0.0548,
      "step": 1387
    },
    {
      "epoch": 1.0412603150787696,
      "grad_norm": 0.14308512210845947,
      "learning_rate": 5.5130906538947034e-05,
      "loss": 0.0416,
      "step": 1388
    },
    {
      "epoch": 1.0420105026256563,
      "grad_norm": 0.17685441672801971,
      "learning_rate": 5.506577069421395e-05,
      "loss": 0.0358,
      "step": 1389
    },
    {
      "epoch": 1.042760690172543,
      "grad_norm": 0.20115597546100616,
      "learning_rate": 5.5000626162182944e-05,
      "loss": 0.043,
      "step": 1390
    },
    {
      "epoch": 1.04351087771943,
      "grad_norm": 0.17785103619098663,
      "learning_rate": 5.49354730545705e-05,
      "loss": 0.0536,
      "step": 1391
    },
    {
      "epoch": 1.0442610652663167,
      "grad_norm": 0.1535765528678894,
      "learning_rate": 5.487031148310775e-05,
      "loss": 0.0333,
      "step": 1392
    },
    {
      "epoch": 1.0450112528132034,
      "grad_norm": 0.14303374290466309,
      "learning_rate": 5.480514155954042e-05,
      "loss": 0.0469,
      "step": 1393
    },
    {
      "epoch": 1.04576144036009,
      "grad_norm": 0.11257001012563705,
      "learning_rate": 5.4739963395628456e-05,
      "loss": 0.0307,
      "step": 1394
    },
    {
      "epoch": 1.0465116279069768,
      "grad_norm": 0.16565249860286713,
      "learning_rate": 5.4674777103146045e-05,
      "loss": 0.0516,
      "step": 1395
    },
    {
      "epoch": 1.0472618154538635,
      "grad_norm": 0.16822345554828644,
      "learning_rate": 5.460958279388122e-05,
      "loss": 0.0518,
      "step": 1396
    },
    {
      "epoch": 1.0480120030007503,
      "grad_norm": 0.1288178563117981,
      "learning_rate": 5.4544380579635824e-05,
      "loss": 0.0395,
      "step": 1397
    },
    {
      "epoch": 1.048762190547637,
      "grad_norm": 0.1471712738275528,
      "learning_rate": 5.447917057222523e-05,
      "loss": 0.0543,
      "step": 1398
    },
    {
      "epoch": 1.0495123780945237,
      "grad_norm": 0.13822641968727112,
      "learning_rate": 5.441395288347818e-05,
      "loss": 0.044,
      "step": 1399
    },
    {
      "epoch": 1.0502625656414104,
      "grad_norm": 0.113441601395607,
      "learning_rate": 5.434872762523658e-05,
      "loss": 0.0446,
      "step": 1400
    },
    {
      "epoch": 1.0502625656414104,
      "eval_loss": 0.0680478885769844,
      "eval_runtime": 2.6592,
      "eval_samples_per_second": 20.307,
      "eval_steps_per_second": 5.265,
      "step": 1400
    },
    {
      "epoch": 1.0510127531882971,
      "grad_norm": 0.1047082468867302,
      "learning_rate": 5.4283494909355314e-05,
      "loss": 0.0316,
      "step": 1401
    },
    {
      "epoch": 1.0517629407351838,
      "grad_norm": 0.2028200924396515,
      "learning_rate": 5.42182548477021e-05,
      "loss": 0.0337,
      "step": 1402
    },
    {
      "epoch": 1.0525131282820706,
      "grad_norm": 0.13617637753486633,
      "learning_rate": 5.41530075521572e-05,
      "loss": 0.0345,
      "step": 1403
    },
    {
      "epoch": 1.0532633158289573,
      "grad_norm": 0.18374083936214447,
      "learning_rate": 5.4087753134613294e-05,
      "loss": 0.0575,
      "step": 1404
    },
    {
      "epoch": 1.054013503375844,
      "grad_norm": 0.18506267666816711,
      "learning_rate": 5.40224917069753e-05,
      "loss": 0.0466,
      "step": 1405
    },
    {
      "epoch": 1.0547636909227307,
      "grad_norm": 0.15255096554756165,
      "learning_rate": 5.3957223381160126e-05,
      "loss": 0.0337,
      "step": 1406
    },
    {
      "epoch": 1.0555138784696174,
      "grad_norm": 0.18384674191474915,
      "learning_rate": 5.389194826909653e-05,
      "loss": 0.0491,
      "step": 1407
    },
    {
      "epoch": 1.0562640660165041,
      "grad_norm": 0.1939430981874466,
      "learning_rate": 5.382666648272489e-05,
      "loss": 0.0941,
      "step": 1408
    },
    {
      "epoch": 1.0570142535633908,
      "grad_norm": 0.1425727754831314,
      "learning_rate": 5.3761378133997044e-05,
      "loss": 0.0483,
      "step": 1409
    },
    {
      "epoch": 1.0577644411102776,
      "grad_norm": 0.1924968659877777,
      "learning_rate": 5.3696083334876105e-05,
      "loss": 0.0747,
      "step": 1410
    },
    {
      "epoch": 1.0585146286571643,
      "grad_norm": 0.1328912377357483,
      "learning_rate": 5.363078219733619e-05,
      "loss": 0.0401,
      "step": 1411
    },
    {
      "epoch": 1.059264816204051,
      "grad_norm": 0.16654489934444427,
      "learning_rate": 5.3565474833362353e-05,
      "loss": 0.0504,
      "step": 1412
    },
    {
      "epoch": 1.0600150037509377,
      "grad_norm": 0.11094582080841064,
      "learning_rate": 5.3500161354950274e-05,
      "loss": 0.0375,
      "step": 1413
    },
    {
      "epoch": 1.0607651912978244,
      "grad_norm": 0.20550234615802765,
      "learning_rate": 5.3434841874106124e-05,
      "loss": 0.0512,
      "step": 1414
    },
    {
      "epoch": 1.0615153788447111,
      "grad_norm": 0.12731848657131195,
      "learning_rate": 5.3369516502846396e-05,
      "loss": 0.0648,
      "step": 1415
    },
    {
      "epoch": 1.0622655663915979,
      "grad_norm": 0.14043542742729187,
      "learning_rate": 5.330418535319768e-05,
      "loss": 0.0588,
      "step": 1416
    },
    {
      "epoch": 1.0630157539384846,
      "grad_norm": 0.1502482295036316,
      "learning_rate": 5.323884853719645e-05,
      "loss": 0.0488,
      "step": 1417
    },
    {
      "epoch": 1.0637659414853713,
      "grad_norm": 0.11024313420057297,
      "learning_rate": 5.31735061668889e-05,
      "loss": 0.0261,
      "step": 1418
    },
    {
      "epoch": 1.064516129032258,
      "grad_norm": 0.15195825695991516,
      "learning_rate": 5.3108158354330795e-05,
      "loss": 0.0403,
      "step": 1419
    },
    {
      "epoch": 1.0652663165791447,
      "grad_norm": 0.11762233078479767,
      "learning_rate": 5.304280521158716e-05,
      "loss": 0.0276,
      "step": 1420
    },
    {
      "epoch": 1.0660165041260314,
      "grad_norm": 0.23322537541389465,
      "learning_rate": 5.2977446850732236e-05,
      "loss": 0.0384,
      "step": 1421
    },
    {
      "epoch": 1.0667666916729182,
      "grad_norm": 0.18611235916614532,
      "learning_rate": 5.291208338384913e-05,
      "loss": 0.0606,
      "step": 1422
    },
    {
      "epoch": 1.0675168792198049,
      "grad_norm": 0.20848964154720306,
      "learning_rate": 5.2846714923029795e-05,
      "loss": 0.0468,
      "step": 1423
    },
    {
      "epoch": 1.0682670667666916,
      "grad_norm": 0.24529072642326355,
      "learning_rate": 5.278134158037469e-05,
      "loss": 0.0634,
      "step": 1424
    },
    {
      "epoch": 1.0690172543135783,
      "grad_norm": 0.10874413698911667,
      "learning_rate": 5.2715963467992656e-05,
      "loss": 0.0273,
      "step": 1425
    },
    {
      "epoch": 1.069767441860465,
      "grad_norm": 0.18923908472061157,
      "learning_rate": 5.265058069800072e-05,
      "loss": 0.0757,
      "step": 1426
    },
    {
      "epoch": 1.0705176294073517,
      "grad_norm": 0.12763912975788116,
      "learning_rate": 5.258519338252389e-05,
      "loss": 0.049,
      "step": 1427
    },
    {
      "epoch": 1.0712678169542387,
      "grad_norm": 0.13909712433815002,
      "learning_rate": 5.251980163369499e-05,
      "loss": 0.0453,
      "step": 1428
    },
    {
      "epoch": 1.0720180045011252,
      "grad_norm": 0.1828707605600357,
      "learning_rate": 5.24544055636544e-05,
      "loss": 0.0451,
      "step": 1429
    },
    {
      "epoch": 1.072768192048012,
      "grad_norm": 0.14683784544467926,
      "learning_rate": 5.2389005284549954e-05,
      "loss": 0.048,
      "step": 1430
    },
    {
      "epoch": 1.0735183795948988,
      "grad_norm": 0.19041985273361206,
      "learning_rate": 5.232360090853671e-05,
      "loss": 0.0724,
      "step": 1431
    },
    {
      "epoch": 1.0742685671417855,
      "grad_norm": 0.2441704124212265,
      "learning_rate": 5.225819254777671e-05,
      "loss": 0.0564,
      "step": 1432
    },
    {
      "epoch": 1.0750187546886723,
      "grad_norm": 0.12937527894973755,
      "learning_rate": 5.219278031443886e-05,
      "loss": 0.0267,
      "step": 1433
    },
    {
      "epoch": 1.075768942235559,
      "grad_norm": 0.2045406699180603,
      "learning_rate": 5.21273643206987e-05,
      "loss": 0.0451,
      "step": 1434
    },
    {
      "epoch": 1.0765191297824457,
      "grad_norm": 0.18630985915660858,
      "learning_rate": 5.206194467873822e-05,
      "loss": 0.0597,
      "step": 1435
    },
    {
      "epoch": 1.0772693173293324,
      "grad_norm": 0.17303740978240967,
      "learning_rate": 5.1996521500745645e-05,
      "loss": 0.0718,
      "step": 1436
    },
    {
      "epoch": 1.0780195048762191,
      "grad_norm": 0.17732733488082886,
      "learning_rate": 5.19310948989153e-05,
      "loss": 0.041,
      "step": 1437
    },
    {
      "epoch": 1.0787696924231058,
      "grad_norm": 0.12811049818992615,
      "learning_rate": 5.186566498544737e-05,
      "loss": 0.0331,
      "step": 1438
    },
    {
      "epoch": 1.0795198799699925,
      "grad_norm": 0.15231534838676453,
      "learning_rate": 5.18002318725477e-05,
      "loss": 0.0295,
      "step": 1439
    },
    {
      "epoch": 1.0802700675168793,
      "grad_norm": 0.16985756158828735,
      "learning_rate": 5.173479567242765e-05,
      "loss": 0.0426,
      "step": 1440
    },
    {
      "epoch": 1.081020255063766,
      "grad_norm": 0.19819964468479156,
      "learning_rate": 5.1669356497303835e-05,
      "loss": 0.062,
      "step": 1441
    },
    {
      "epoch": 1.0817704426106527,
      "grad_norm": 0.15997275710105896,
      "learning_rate": 5.1603914459398016e-05,
      "loss": 0.0445,
      "step": 1442
    },
    {
      "epoch": 1.0825206301575394,
      "grad_norm": 0.1522863507270813,
      "learning_rate": 5.153846967093684e-05,
      "loss": 0.0602,
      "step": 1443
    },
    {
      "epoch": 1.0832708177044261,
      "grad_norm": 0.16248387098312378,
      "learning_rate": 5.1473022244151684e-05,
      "loss": 0.0386,
      "step": 1444
    },
    {
      "epoch": 1.0840210052513128,
      "grad_norm": 0.16288204491138458,
      "learning_rate": 5.140757229127842e-05,
      "loss": 0.0442,
      "step": 1445
    },
    {
      "epoch": 1.0847711927981996,
      "grad_norm": 0.150307759642601,
      "learning_rate": 5.1342119924557275e-05,
      "loss": 0.0484,
      "step": 1446
    },
    {
      "epoch": 1.0855213803450863,
      "grad_norm": 0.19405823945999146,
      "learning_rate": 5.127666525623264e-05,
      "loss": 0.0516,
      "step": 1447
    },
    {
      "epoch": 1.086271567891973,
      "grad_norm": 0.20875540375709534,
      "learning_rate": 5.121120839855279e-05,
      "loss": 0.0476,
      "step": 1448
    },
    {
      "epoch": 1.0870217554388597,
      "grad_norm": 0.24014073610305786,
      "learning_rate": 5.114574946376982e-05,
      "loss": 0.0563,
      "step": 1449
    },
    {
      "epoch": 1.0877719429857464,
      "grad_norm": 0.15809394419193268,
      "learning_rate": 5.1080288564139325e-05,
      "loss": 0.0455,
      "step": 1450
    },
    {
      "epoch": 1.0885221305326331,
      "grad_norm": 0.17998574674129486,
      "learning_rate": 5.101482581192033e-05,
      "loss": 0.0366,
      "step": 1451
    },
    {
      "epoch": 1.0892723180795199,
      "grad_norm": 0.16414082050323486,
      "learning_rate": 5.0949361319374996e-05,
      "loss": 0.0471,
      "step": 1452
    },
    {
      "epoch": 1.0900225056264066,
      "grad_norm": 0.18297888338565826,
      "learning_rate": 5.0883895198768494e-05,
      "loss": 0.0468,
      "step": 1453
    },
    {
      "epoch": 1.0907726931732933,
      "grad_norm": 0.11977231502532959,
      "learning_rate": 5.0818427562368764e-05,
      "loss": 0.0339,
      "step": 1454
    },
    {
      "epoch": 1.09152288072018,
      "grad_norm": 0.13561047613620758,
      "learning_rate": 5.0752958522446356e-05,
      "loss": 0.0643,
      "step": 1455
    },
    {
      "epoch": 1.0922730682670667,
      "grad_norm": 0.1655491441488266,
      "learning_rate": 5.0687488191274215e-05,
      "loss": 0.0632,
      "step": 1456
    },
    {
      "epoch": 1.0930232558139534,
      "grad_norm": 0.2185972034931183,
      "learning_rate": 5.0622016681127526e-05,
      "loss": 0.0395,
      "step": 1457
    },
    {
      "epoch": 1.0937734433608401,
      "grad_norm": 0.13363471627235413,
      "learning_rate": 5.055654410428349e-05,
      "loss": 0.0413,
      "step": 1458
    },
    {
      "epoch": 1.0945236309077269,
      "grad_norm": 0.17551277577877045,
      "learning_rate": 5.0491070573021116e-05,
      "loss": 0.051,
      "step": 1459
    },
    {
      "epoch": 1.0952738184546136,
      "grad_norm": 0.16505476832389832,
      "learning_rate": 5.0425596199621064e-05,
      "loss": 0.0614,
      "step": 1460
    },
    {
      "epoch": 1.0960240060015003,
      "grad_norm": 0.12644611299037933,
      "learning_rate": 5.036012109636543e-05,
      "loss": 0.0346,
      "step": 1461
    },
    {
      "epoch": 1.096774193548387,
      "grad_norm": 0.10962353646755219,
      "learning_rate": 5.0294645375537594e-05,
      "loss": 0.0376,
      "step": 1462
    },
    {
      "epoch": 1.0975243810952737,
      "grad_norm": 0.18371117115020752,
      "learning_rate": 5.022916914942195e-05,
      "loss": 0.0449,
      "step": 1463
    },
    {
      "epoch": 1.0982745686421604,
      "grad_norm": 0.22092092037200928,
      "learning_rate": 5.0163692530303774e-05,
      "loss": 0.0473,
      "step": 1464
    },
    {
      "epoch": 1.0990247561890472,
      "grad_norm": 0.18737155199050903,
      "learning_rate": 5.009821563046903e-05,
      "loss": 0.0371,
      "step": 1465
    },
    {
      "epoch": 1.0997749437359339,
      "grad_norm": 0.1893552988767624,
      "learning_rate": 5.003273856220415e-05,
      "loss": 0.0457,
      "step": 1466
    },
    {
      "epoch": 1.1005251312828208,
      "grad_norm": 0.14333701133728027,
      "learning_rate": 4.996726143779586e-05,
      "loss": 0.0648,
      "step": 1467
    },
    {
      "epoch": 1.1012753188297075,
      "grad_norm": 0.17025060951709747,
      "learning_rate": 4.990178436953099e-05,
      "loss": 0.0396,
      "step": 1468
    },
    {
      "epoch": 1.1020255063765942,
      "grad_norm": 0.14343062043190002,
      "learning_rate": 4.9836307469696244e-05,
      "loss": 0.0333,
      "step": 1469
    },
    {
      "epoch": 1.102775693923481,
      "grad_norm": 0.13673539459705353,
      "learning_rate": 4.9770830850578075e-05,
      "loss": 0.0391,
      "step": 1470
    },
    {
      "epoch": 1.1035258814703677,
      "grad_norm": 0.1872038096189499,
      "learning_rate": 4.9705354624462424e-05,
      "loss": 0.0645,
      "step": 1471
    },
    {
      "epoch": 1.1042760690172544,
      "grad_norm": 0.1406128704547882,
      "learning_rate": 4.963987890363458e-05,
      "loss": 0.0367,
      "step": 1472
    },
    {
      "epoch": 1.105026256564141,
      "grad_norm": 0.12920933961868286,
      "learning_rate": 4.957440380037896e-05,
      "loss": 0.0391,
      "step": 1473
    },
    {
      "epoch": 1.1057764441110278,
      "grad_norm": 0.15628013014793396,
      "learning_rate": 4.9508929426978896e-05,
      "loss": 0.0442,
      "step": 1474
    },
    {
      "epoch": 1.1065266316579145,
      "grad_norm": 0.13842900097370148,
      "learning_rate": 4.944345589571651e-05,
      "loss": 0.0298,
      "step": 1475
    },
    {
      "epoch": 1.1072768192048013,
      "grad_norm": 0.2180444896221161,
      "learning_rate": 4.937798331887248e-05,
      "loss": 0.0593,
      "step": 1476
    },
    {
      "epoch": 1.108027006751688,
      "grad_norm": 0.1547778993844986,
      "learning_rate": 4.931251180872579e-05,
      "loss": 0.0493,
      "step": 1477
    },
    {
      "epoch": 1.1087771942985747,
      "grad_norm": 0.2607737183570862,
      "learning_rate": 4.9247041477553656e-05,
      "loss": 0.0712,
      "step": 1478
    },
    {
      "epoch": 1.1095273818454614,
      "grad_norm": 0.17707867920398712,
      "learning_rate": 4.9181572437631255e-05,
      "loss": 0.07,
      "step": 1479
    },
    {
      "epoch": 1.1102775693923481,
      "grad_norm": 0.17139020562171936,
      "learning_rate": 4.911610480123151e-05,
      "loss": 0.061,
      "step": 1480
    },
    {
      "epoch": 1.1110277569392348,
      "grad_norm": 0.19931213557720184,
      "learning_rate": 4.905063868062501e-05,
      "loss": 0.0607,
      "step": 1481
    },
    {
      "epoch": 1.1117779444861215,
      "grad_norm": 0.17375808954238892,
      "learning_rate": 4.898517418807968e-05,
      "loss": 0.0444,
      "step": 1482
    },
    {
      "epoch": 1.1125281320330083,
      "grad_norm": 0.12767089903354645,
      "learning_rate": 4.891971143586069e-05,
      "loss": 0.0486,
      "step": 1483
    },
    {
      "epoch": 1.113278319579895,
      "grad_norm": 0.1534682959318161,
      "learning_rate": 4.88542505362302e-05,
      "loss": 0.0458,
      "step": 1484
    },
    {
      "epoch": 1.1140285071267817,
      "grad_norm": 0.17411962151527405,
      "learning_rate": 4.878879160144723e-05,
      "loss": 0.0464,
      "step": 1485
    },
    {
      "epoch": 1.1147786946736684,
      "grad_norm": 0.12393703311681747,
      "learning_rate": 4.872333474376739e-05,
      "loss": 0.0543,
      "step": 1486
    },
    {
      "epoch": 1.1155288822205551,
      "grad_norm": 0.11996886134147644,
      "learning_rate": 4.865788007544274e-05,
      "loss": 0.0248,
      "step": 1487
    },
    {
      "epoch": 1.1162790697674418,
      "grad_norm": 0.14848671853542328,
      "learning_rate": 4.859242770872158e-05,
      "loss": 0.0312,
      "step": 1488
    },
    {
      "epoch": 1.1170292573143286,
      "grad_norm": 0.16110281646251678,
      "learning_rate": 4.852697775584833e-05,
      "loss": 0.04,
      "step": 1489
    },
    {
      "epoch": 1.1177794448612153,
      "grad_norm": 0.1755913347005844,
      "learning_rate": 4.846153032906316e-05,
      "loss": 0.0422,
      "step": 1490
    },
    {
      "epoch": 1.118529632408102,
      "grad_norm": 0.15683995187282562,
      "learning_rate": 4.8396085540601995e-05,
      "loss": 0.0575,
      "step": 1491
    },
    {
      "epoch": 1.1192798199549887,
      "grad_norm": 0.15512287616729736,
      "learning_rate": 4.833064350269617e-05,
      "loss": 0.0551,
      "step": 1492
    },
    {
      "epoch": 1.1200300075018754,
      "grad_norm": 0.16375933587551117,
      "learning_rate": 4.826520432757236e-05,
      "loss": 0.05,
      "step": 1493
    },
    {
      "epoch": 1.1207801950487621,
      "grad_norm": 0.19421043992042542,
      "learning_rate": 4.8199768127452314e-05,
      "loss": 0.0733,
      "step": 1494
    },
    {
      "epoch": 1.1215303825956489,
      "grad_norm": 0.2077593207359314,
      "learning_rate": 4.813433501455264e-05,
      "loss": 0.0483,
      "step": 1495
    },
    {
      "epoch": 1.1222805701425356,
      "grad_norm": 0.19140911102294922,
      "learning_rate": 4.806890510108471e-05,
      "loss": 0.0659,
      "step": 1496
    },
    {
      "epoch": 1.1230307576894223,
      "grad_norm": 0.23248641192913055,
      "learning_rate": 4.800347849925437e-05,
      "loss": 0.0512,
      "step": 1497
    },
    {
      "epoch": 1.123780945236309,
      "grad_norm": 0.20153175294399261,
      "learning_rate": 4.793805532126181e-05,
      "loss": 0.0604,
      "step": 1498
    },
    {
      "epoch": 1.1245311327831957,
      "grad_norm": 0.17858366668224335,
      "learning_rate": 4.787263567930132e-05,
      "loss": 0.0291,
      "step": 1499
    },
    {
      "epoch": 1.1252813203300824,
      "grad_norm": 0.16331790387630463,
      "learning_rate": 4.780721968556115e-05,
      "loss": 0.0485,
      "step": 1500
    },
    {
      "epoch": 1.1260315078769692,
      "grad_norm": 0.15837393701076508,
      "learning_rate": 4.774180745222331e-05,
      "loss": 0.0385,
      "step": 1501
    },
    {
      "epoch": 1.1267816954238559,
      "grad_norm": 0.13282233476638794,
      "learning_rate": 4.7676399091463296e-05,
      "loss": 0.0324,
      "step": 1502
    },
    {
      "epoch": 1.1275318829707426,
      "grad_norm": 0.26786598563194275,
      "learning_rate": 4.7610994715450044e-05,
      "loss": 0.0394,
      "step": 1503
    },
    {
      "epoch": 1.1282820705176295,
      "grad_norm": 0.19126728177070618,
      "learning_rate": 4.754559443634561e-05,
      "loss": 0.0477,
      "step": 1504
    },
    {
      "epoch": 1.129032258064516,
      "grad_norm": 0.1592726856470108,
      "learning_rate": 4.748019836630503e-05,
      "loss": 0.0593,
      "step": 1505
    },
    {
      "epoch": 1.129782445611403,
      "grad_norm": 0.19169463217258453,
      "learning_rate": 4.7414806617476124e-05,
      "loss": 0.0602,
      "step": 1506
    },
    {
      "epoch": 1.1305326331582897,
      "grad_norm": 0.2752186357975006,
      "learning_rate": 4.7349419301999294e-05,
      "loss": 0.071,
      "step": 1507
    },
    {
      "epoch": 1.1312828207051764,
      "grad_norm": 0.22131672501564026,
      "learning_rate": 4.7284036532007356e-05,
      "loss": 0.082,
      "step": 1508
    },
    {
      "epoch": 1.132033008252063,
      "grad_norm": 0.15666484832763672,
      "learning_rate": 4.721865841962533e-05,
      "loss": 0.0487,
      "step": 1509
    },
    {
      "epoch": 1.1327831957989498,
      "grad_norm": 0.11834803223609924,
      "learning_rate": 4.715328507697021e-05,
      "loss": 0.0524,
      "step": 1510
    },
    {
      "epoch": 1.1335333833458365,
      "grad_norm": 0.19373413920402527,
      "learning_rate": 4.7087916616150886e-05,
      "loss": 0.0618,
      "step": 1511
    },
    {
      "epoch": 1.1342835708927232,
      "grad_norm": 0.15369850397109985,
      "learning_rate": 4.702255314926779e-05,
      "loss": 0.0386,
      "step": 1512
    },
    {
      "epoch": 1.13503375843961,
      "grad_norm": 0.2039160132408142,
      "learning_rate": 4.695719478841286e-05,
      "loss": 0.0641,
      "step": 1513
    },
    {
      "epoch": 1.1357839459864967,
      "grad_norm": 0.1577986180782318,
      "learning_rate": 4.6891841645669224e-05,
      "loss": 0.0466,
      "step": 1514
    },
    {
      "epoch": 1.1365341335333834,
      "grad_norm": 0.17635010182857513,
      "learning_rate": 4.6826493833111104e-05,
      "loss": 0.0432,
      "step": 1515
    },
    {
      "epoch": 1.13728432108027,
      "grad_norm": 0.1743796020746231,
      "learning_rate": 4.676115146280356e-05,
      "loss": 0.0405,
      "step": 1516
    },
    {
      "epoch": 1.1380345086271568,
      "grad_norm": 0.1372276097536087,
      "learning_rate": 4.669581464680233e-05,
      "loss": 0.0345,
      "step": 1517
    },
    {
      "epoch": 1.1387846961740435,
      "grad_norm": 0.1724540740251541,
      "learning_rate": 4.66304834971536e-05,
      "loss": 0.0515,
      "step": 1518
    },
    {
      "epoch": 1.1395348837209303,
      "grad_norm": 0.10708152502775192,
      "learning_rate": 4.656515812589389e-05,
      "loss": 0.0317,
      "step": 1519
    },
    {
      "epoch": 1.140285071267817,
      "grad_norm": 0.1259574443101883,
      "learning_rate": 4.6499838645049744e-05,
      "loss": 0.0436,
      "step": 1520
    },
    {
      "epoch": 1.1410352588147037,
      "grad_norm": 0.1755182147026062,
      "learning_rate": 4.643452516663766e-05,
      "loss": 0.0411,
      "step": 1521
    },
    {
      "epoch": 1.1417854463615904,
      "grad_norm": 0.1899542510509491,
      "learning_rate": 4.636921780266381e-05,
      "loss": 0.0617,
      "step": 1522
    },
    {
      "epoch": 1.1425356339084771,
      "grad_norm": 0.1305563747882843,
      "learning_rate": 4.63039166651239e-05,
      "loss": 0.0475,
      "step": 1523
    },
    {
      "epoch": 1.1432858214553638,
      "grad_norm": 0.11840610206127167,
      "learning_rate": 4.623862186600297e-05,
      "loss": 0.0378,
      "step": 1524
    },
    {
      "epoch": 1.1440360090022506,
      "grad_norm": 0.13866670429706573,
      "learning_rate": 4.617333351727513e-05,
      "loss": 0.0361,
      "step": 1525
    },
    {
      "epoch": 1.1447861965491373,
      "grad_norm": 0.2099863588809967,
      "learning_rate": 4.61080517309035e-05,
      "loss": 0.0741,
      "step": 1526
    },
    {
      "epoch": 1.145536384096024,
      "grad_norm": 0.21586652100086212,
      "learning_rate": 4.604277661883989e-05,
      "loss": 0.058,
      "step": 1527
    },
    {
      "epoch": 1.1462865716429107,
      "grad_norm": 0.09848996251821518,
      "learning_rate": 4.5977508293024726e-05,
      "loss": 0.0238,
      "step": 1528
    },
    {
      "epoch": 1.1470367591897974,
      "grad_norm": 0.14757753908634186,
      "learning_rate": 4.591224686538672e-05,
      "loss": 0.0321,
      "step": 1529
    },
    {
      "epoch": 1.1477869467366841,
      "grad_norm": 0.18732796609401703,
      "learning_rate": 4.584699244784281e-05,
      "loss": 0.0584,
      "step": 1530
    },
    {
      "epoch": 1.1485371342835708,
      "grad_norm": 0.21146154403686523,
      "learning_rate": 4.578174515229789e-05,
      "loss": 0.0973,
      "step": 1531
    },
    {
      "epoch": 1.1492873218304576,
      "grad_norm": 0.1362430453300476,
      "learning_rate": 4.5716505090644684e-05,
      "loss": 0.0449,
      "step": 1532
    },
    {
      "epoch": 1.1500375093773443,
      "grad_norm": 0.1803388148546219,
      "learning_rate": 4.5651272374763423e-05,
      "loss": 0.0354,
      "step": 1533
    },
    {
      "epoch": 1.150787696924231,
      "grad_norm": 0.1937418282032013,
      "learning_rate": 4.558604711652183e-05,
      "loss": 0.0531,
      "step": 1534
    },
    {
      "epoch": 1.1515378844711177,
      "grad_norm": 0.15416602790355682,
      "learning_rate": 4.552082942777478e-05,
      "loss": 0.04,
      "step": 1535
    },
    {
      "epoch": 1.1522880720180044,
      "grad_norm": 0.1575358361005783,
      "learning_rate": 4.545561942036418e-05,
      "loss": 0.0378,
      "step": 1536
    },
    {
      "epoch": 1.1530382595648911,
      "grad_norm": 0.171066552400589,
      "learning_rate": 4.5390417206118784e-05,
      "loss": 0.054,
      "step": 1537
    },
    {
      "epoch": 1.1537884471117779,
      "grad_norm": 0.15054123103618622,
      "learning_rate": 4.5325222896853966e-05,
      "loss": 0.0647,
      "step": 1538
    },
    {
      "epoch": 1.1545386346586646,
      "grad_norm": 0.23045825958251953,
      "learning_rate": 4.5260036604371556e-05,
      "loss": 0.076,
      "step": 1539
    },
    {
      "epoch": 1.1552888222055513,
      "grad_norm": 0.17275682091712952,
      "learning_rate": 4.51948584404596e-05,
      "loss": 0.042,
      "step": 1540
    },
    {
      "epoch": 1.1560390097524382,
      "grad_norm": 0.15596450865268707,
      "learning_rate": 4.5129688516892264e-05,
      "loss": 0.0429,
      "step": 1541
    },
    {
      "epoch": 1.1567891972993247,
      "grad_norm": 0.16053304076194763,
      "learning_rate": 4.506452694542953e-05,
      "loss": 0.0463,
      "step": 1542
    },
    {
      "epoch": 1.1575393848462117,
      "grad_norm": 0.1733517348766327,
      "learning_rate": 4.499937383781708e-05,
      "loss": 0.0432,
      "step": 1543
    },
    {
      "epoch": 1.1582895723930982,
      "grad_norm": 0.14224635064601898,
      "learning_rate": 4.493422930578605e-05,
      "loss": 0.0334,
      "step": 1544
    },
    {
      "epoch": 1.159039759939985,
      "grad_norm": 0.1554994285106659,
      "learning_rate": 4.486909346105296e-05,
      "loss": 0.0477,
      "step": 1545
    },
    {
      "epoch": 1.1597899474868718,
      "grad_norm": 0.1568094938993454,
      "learning_rate": 4.480396641531932e-05,
      "loss": 0.0415,
      "step": 1546
    },
    {
      "epoch": 1.1605401350337585,
      "grad_norm": 0.15656742453575134,
      "learning_rate": 4.4738848280271626e-05,
      "loss": 0.0466,
      "step": 1547
    },
    {
      "epoch": 1.1612903225806452,
      "grad_norm": 0.10462535917758942,
      "learning_rate": 4.467373916758105e-05,
      "loss": 0.0191,
      "step": 1548
    },
    {
      "epoch": 1.162040510127532,
      "grad_norm": 0.2560919225215912,
      "learning_rate": 4.460863918890333e-05,
      "loss": 0.0794,
      "step": 1549
    },
    {
      "epoch": 1.1627906976744187,
      "grad_norm": 0.1778658926486969,
      "learning_rate": 4.454354845587849e-05,
      "loss": 0.0448,
      "step": 1550
    },
    {
      "epoch": 1.1635408852213054,
      "grad_norm": 0.13617105782032013,
      "learning_rate": 4.4478467080130734e-05,
      "loss": 0.0437,
      "step": 1551
    },
    {
      "epoch": 1.164291072768192,
      "grad_norm": 0.22036074101924896,
      "learning_rate": 4.4413395173268243e-05,
      "loss": 0.0498,
      "step": 1552
    },
    {
      "epoch": 1.1650412603150788,
      "grad_norm": 0.23211778700351715,
      "learning_rate": 4.43483328468829e-05,
      "loss": 0.0546,
      "step": 1553
    },
    {
      "epoch": 1.1657914478619655,
      "grad_norm": 0.18650752305984497,
      "learning_rate": 4.4283280212550194e-05,
      "loss": 0.0471,
      "step": 1554
    },
    {
      "epoch": 1.1665416354088523,
      "grad_norm": 0.14625145494937897,
      "learning_rate": 4.421823738182898e-05,
      "loss": 0.0341,
      "step": 1555
    },
    {
      "epoch": 1.167291822955739,
      "grad_norm": 0.17568133771419525,
      "learning_rate": 4.4153204466261334e-05,
      "loss": 0.0489,
      "step": 1556
    },
    {
      "epoch": 1.1680420105026257,
      "grad_norm": 0.15340299904346466,
      "learning_rate": 4.408818157737227e-05,
      "loss": 0.0426,
      "step": 1557
    },
    {
      "epoch": 1.1687921980495124,
      "grad_norm": 0.17339184880256653,
      "learning_rate": 4.402316882666964e-05,
      "loss": 0.0457,
      "step": 1558
    },
    {
      "epoch": 1.1695423855963991,
      "grad_norm": 0.25148463249206543,
      "learning_rate": 4.395816632564393e-05,
      "loss": 0.0479,
      "step": 1559
    },
    {
      "epoch": 1.1702925731432858,
      "grad_norm": 0.1664632111787796,
      "learning_rate": 4.3893174185768045e-05,
      "loss": 0.0466,
      "step": 1560
    },
    {
      "epoch": 1.1710427606901725,
      "grad_norm": 0.14164632558822632,
      "learning_rate": 4.382819251849707e-05,
      "loss": 0.0288,
      "step": 1561
    },
    {
      "epoch": 1.1717929482370593,
      "grad_norm": 0.11925794184207916,
      "learning_rate": 4.376322143526818e-05,
      "loss": 0.0349,
      "step": 1562
    },
    {
      "epoch": 1.172543135783946,
      "grad_norm": 0.1766175925731659,
      "learning_rate": 4.3698261047500376e-05,
      "loss": 0.0427,
      "step": 1563
    },
    {
      "epoch": 1.1732933233308327,
      "grad_norm": 0.18459202349185944,
      "learning_rate": 4.3633311466594345e-05,
      "loss": 0.0401,
      "step": 1564
    },
    {
      "epoch": 1.1740435108777194,
      "grad_norm": 0.20463117957115173,
      "learning_rate": 4.3568372803932195e-05,
      "loss": 0.0382,
      "step": 1565
    },
    {
      "epoch": 1.1747936984246061,
      "grad_norm": 0.20283189415931702,
      "learning_rate": 4.3503445170877354e-05,
      "loss": 0.0423,
      "step": 1566
    },
    {
      "epoch": 1.1755438859714928,
      "grad_norm": 0.1398739516735077,
      "learning_rate": 4.343852867877433e-05,
      "loss": 0.0371,
      "step": 1567
    },
    {
      "epoch": 1.1762940735183796,
      "grad_norm": 0.19692690670490265,
      "learning_rate": 4.3373623438948496e-05,
      "loss": 0.044,
      "step": 1568
    },
    {
      "epoch": 1.1770442610652663,
      "grad_norm": 0.11689583212137222,
      "learning_rate": 4.330872956270596e-05,
      "loss": 0.0393,
      "step": 1569
    },
    {
      "epoch": 1.177794448612153,
      "grad_norm": 0.17780661582946777,
      "learning_rate": 4.324384716133332e-05,
      "loss": 0.0449,
      "step": 1570
    },
    {
      "epoch": 1.1785446361590397,
      "grad_norm": 0.2003779113292694,
      "learning_rate": 4.317897634609751e-05,
      "loss": 0.0537,
      "step": 1571
    },
    {
      "epoch": 1.1792948237059264,
      "grad_norm": 0.20001736283302307,
      "learning_rate": 4.3114117228245565e-05,
      "loss": 0.0445,
      "step": 1572
    },
    {
      "epoch": 1.1800450112528131,
      "grad_norm": 0.19485583901405334,
      "learning_rate": 4.304926991900453e-05,
      "loss": 0.0503,
      "step": 1573
    },
    {
      "epoch": 1.1807951987996999,
      "grad_norm": 0.20719118416309357,
      "learning_rate": 4.298443452958113e-05,
      "loss": 0.0432,
      "step": 1574
    },
    {
      "epoch": 1.1815453863465866,
      "grad_norm": 0.21073326468467712,
      "learning_rate": 4.291961117116168e-05,
      "loss": 0.0429,
      "step": 1575
    },
    {
      "epoch": 1.1822955738934733,
      "grad_norm": 0.11755035817623138,
      "learning_rate": 4.285479995491185e-05,
      "loss": 0.03,
      "step": 1576
    },
    {
      "epoch": 1.18304576144036,
      "grad_norm": 0.1673261970281601,
      "learning_rate": 4.279000099197651e-05,
      "loss": 0.0404,
      "step": 1577
    },
    {
      "epoch": 1.1837959489872467,
      "grad_norm": 0.13132010400295258,
      "learning_rate": 4.272521439347947e-05,
      "loss": 0.0326,
      "step": 1578
    },
    {
      "epoch": 1.1845461365341334,
      "grad_norm": 0.18214571475982666,
      "learning_rate": 4.26604402705234e-05,
      "loss": 0.0395,
      "step": 1579
    },
    {
      "epoch": 1.1852963240810204,
      "grad_norm": 0.15369313955307007,
      "learning_rate": 4.259567873418952e-05,
      "loss": 0.0404,
      "step": 1580
    },
    {
      "epoch": 1.1860465116279069,
      "grad_norm": 0.21026532351970673,
      "learning_rate": 4.25309298955375e-05,
      "loss": 0.0619,
      "step": 1581
    },
    {
      "epoch": 1.1867966991747938,
      "grad_norm": 0.2855604290962219,
      "learning_rate": 4.246619386560521e-05,
      "loss": 0.0477,
      "step": 1582
    },
    {
      "epoch": 1.1875468867216805,
      "grad_norm": 0.10382580757141113,
      "learning_rate": 4.240147075540858e-05,
      "loss": 0.0271,
      "step": 1583
    },
    {
      "epoch": 1.1882970742685672,
      "grad_norm": 0.11604207754135132,
      "learning_rate": 4.233676067594137e-05,
      "loss": 0.0383,
      "step": 1584
    },
    {
      "epoch": 1.189047261815454,
      "grad_norm": 0.18044514954090118,
      "learning_rate": 4.227206373817497e-05,
      "loss": 0.0517,
      "step": 1585
    },
    {
      "epoch": 1.1897974493623407,
      "grad_norm": 0.1902741938829422,
      "learning_rate": 4.220738005305827e-05,
      "loss": 0.0564,
      "step": 1586
    },
    {
      "epoch": 1.1905476369092274,
      "grad_norm": 0.1472162902355194,
      "learning_rate": 4.214270973151745e-05,
      "loss": 0.0502,
      "step": 1587
    },
    {
      "epoch": 1.191297824456114,
      "grad_norm": 0.17314302921295166,
      "learning_rate": 4.207805288445571e-05,
      "loss": 0.0493,
      "step": 1588
    },
    {
      "epoch": 1.1920480120030008,
      "grad_norm": 0.17650650441646576,
      "learning_rate": 4.201340962275318e-05,
      "loss": 0.0666,
      "step": 1589
    },
    {
      "epoch": 1.1927981995498875,
      "grad_norm": 0.12752214074134827,
      "learning_rate": 4.194878005726671e-05,
      "loss": 0.0296,
      "step": 1590
    },
    {
      "epoch": 1.1935483870967742,
      "grad_norm": 0.16013704240322113,
      "learning_rate": 4.1884164298829615e-05,
      "loss": 0.0344,
      "step": 1591
    },
    {
      "epoch": 1.194298574643661,
      "grad_norm": 0.1943400651216507,
      "learning_rate": 4.181956245825158e-05,
      "loss": 0.0401,
      "step": 1592
    },
    {
      "epoch": 1.1950487621905477,
      "grad_norm": 0.1719129979610443,
      "learning_rate": 4.1754974646318365e-05,
      "loss": 0.0461,
      "step": 1593
    },
    {
      "epoch": 1.1957989497374344,
      "grad_norm": 0.1805904060602188,
      "learning_rate": 4.1690400973791756e-05,
      "loss": 0.0496,
      "step": 1594
    },
    {
      "epoch": 1.196549137284321,
      "grad_norm": 0.1407364308834076,
      "learning_rate": 4.1625841551409195e-05,
      "loss": 0.045,
      "step": 1595
    },
    {
      "epoch": 1.1972993248312078,
      "grad_norm": 0.21922746300697327,
      "learning_rate": 4.156129648988376e-05,
      "loss": 0.0575,
      "step": 1596
    },
    {
      "epoch": 1.1980495123780945,
      "grad_norm": 0.08665847778320312,
      "learning_rate": 4.149676589990388e-05,
      "loss": 0.0275,
      "step": 1597
    },
    {
      "epoch": 1.1987996999249813,
      "grad_norm": 0.30365556478500366,
      "learning_rate": 4.143224989213315e-05,
      "loss": 0.0697,
      "step": 1598
    },
    {
      "epoch": 1.199549887471868,
      "grad_norm": 0.1910962462425232,
      "learning_rate": 4.136774857721017e-05,
      "loss": 0.0534,
      "step": 1599
    },
    {
      "epoch": 1.2003000750187547,
      "grad_norm": 0.14763104915618896,
      "learning_rate": 4.130326206574834e-05,
      "loss": 0.0486,
      "step": 1600
    },
    {
      "epoch": 1.2003000750187547,
      "eval_loss": 0.06795958429574966,
      "eval_runtime": 2.6514,
      "eval_samples_per_second": 20.366,
      "eval_steps_per_second": 5.28,
      "step": 1600
    },
    {
      "epoch": 1.2010502625656414,
      "grad_norm": 0.14768727123737335,
      "learning_rate": 4.1238790468335685e-05,
      "loss": 0.0339,
      "step": 1601
    },
    {
      "epoch": 1.2018004501125281,
      "grad_norm": 0.17933112382888794,
      "learning_rate": 4.117433389553466e-05,
      "loss": 0.041,
      "step": 1602
    },
    {
      "epoch": 1.2025506376594148,
      "grad_norm": 0.1807253360748291,
      "learning_rate": 4.1109892457881924e-05,
      "loss": 0.049,
      "step": 1603
    },
    {
      "epoch": 1.2033008252063015,
      "grad_norm": 0.13722018897533417,
      "learning_rate": 4.1045466265888195e-05,
      "loss": 0.0468,
      "step": 1604
    },
    {
      "epoch": 1.2040510127531883,
      "grad_norm": 0.18306781351566315,
      "learning_rate": 4.0981055430038055e-05,
      "loss": 0.0366,
      "step": 1605
    },
    {
      "epoch": 1.204801200300075,
      "grad_norm": 0.17573761940002441,
      "learning_rate": 4.091666006078974e-05,
      "loss": 0.0466,
      "step": 1606
    },
    {
      "epoch": 1.2055513878469617,
      "grad_norm": 0.15146800875663757,
      "learning_rate": 4.085228026857498e-05,
      "loss": 0.0456,
      "step": 1607
    },
    {
      "epoch": 1.2063015753938484,
      "grad_norm": 0.18410256505012512,
      "learning_rate": 4.0787916163798743e-05,
      "loss": 0.0438,
      "step": 1608
    },
    {
      "epoch": 1.2070517629407351,
      "grad_norm": 0.24622641503810883,
      "learning_rate": 4.0723567856839184e-05,
      "loss": 0.073,
      "step": 1609
    },
    {
      "epoch": 1.2078019504876218,
      "grad_norm": 0.15817995369434357,
      "learning_rate": 4.0659235458047264e-05,
      "loss": 0.0362,
      "step": 1610
    },
    {
      "epoch": 1.2085521380345086,
      "grad_norm": 0.1328483670949936,
      "learning_rate": 4.0594919077746734e-05,
      "loss": 0.0281,
      "step": 1611
    },
    {
      "epoch": 1.2093023255813953,
      "grad_norm": 0.2026902437210083,
      "learning_rate": 4.053061882623386e-05,
      "loss": 0.0549,
      "step": 1612
    },
    {
      "epoch": 1.210052513128282,
      "grad_norm": 0.15139414370059967,
      "learning_rate": 4.0466334813777216e-05,
      "loss": 0.0439,
      "step": 1613
    },
    {
      "epoch": 1.2108027006751687,
      "grad_norm": 0.11512304097414017,
      "learning_rate": 4.040206715061758e-05,
      "loss": 0.0335,
      "step": 1614
    },
    {
      "epoch": 1.2115528882220554,
      "grad_norm": 0.1752549260854721,
      "learning_rate": 4.033781594696767e-05,
      "loss": 0.0481,
      "step": 1615
    },
    {
      "epoch": 1.2123030757689421,
      "grad_norm": 0.13294178247451782,
      "learning_rate": 4.027358131301194e-05,
      "loss": 0.03,
      "step": 1616
    },
    {
      "epoch": 1.213053263315829,
      "grad_norm": 0.3617671728134155,
      "learning_rate": 4.0209363358906495e-05,
      "loss": 0.0489,
      "step": 1617
    },
    {
      "epoch": 1.2138034508627156,
      "grad_norm": 0.20684273540973663,
      "learning_rate": 4.014516219477878e-05,
      "loss": 0.0548,
      "step": 1618
    },
    {
      "epoch": 1.2145536384096025,
      "grad_norm": 0.16169099509716034,
      "learning_rate": 4.008097793072749e-05,
      "loss": 0.0528,
      "step": 1619
    },
    {
      "epoch": 1.215303825956489,
      "grad_norm": 0.16483749449253082,
      "learning_rate": 4.00168106768223e-05,
      "loss": 0.055,
      "step": 1620
    },
    {
      "epoch": 1.216054013503376,
      "grad_norm": 0.11685378104448318,
      "learning_rate": 3.9952660543103734e-05,
      "loss": 0.0384,
      "step": 1621
    },
    {
      "epoch": 1.2168042010502627,
      "grad_norm": 0.2155417501926422,
      "learning_rate": 3.988852763958297e-05,
      "loss": 0.0437,
      "step": 1622
    },
    {
      "epoch": 1.2175543885971494,
      "grad_norm": 0.21236859261989594,
      "learning_rate": 3.9824412076241595e-05,
      "loss": 0.0675,
      "step": 1623
    },
    {
      "epoch": 1.218304576144036,
      "grad_norm": 0.18504956364631653,
      "learning_rate": 3.9760313963031516e-05,
      "loss": 0.0522,
      "step": 1624
    },
    {
      "epoch": 1.2190547636909228,
      "grad_norm": 0.12959396839141846,
      "learning_rate": 3.9696233409874654e-05,
      "loss": 0.0456,
      "step": 1625
    },
    {
      "epoch": 1.2198049512378095,
      "grad_norm": 0.1511463224887848,
      "learning_rate": 3.963217052666287e-05,
      "loss": 0.0311,
      "step": 1626
    },
    {
      "epoch": 1.2205551387846962,
      "grad_norm": 0.1537894457578659,
      "learning_rate": 3.956812542325769e-05,
      "loss": 0.0418,
      "step": 1627
    },
    {
      "epoch": 1.221305326331583,
      "grad_norm": 0.19139213860034943,
      "learning_rate": 3.950409820949018e-05,
      "loss": 0.0635,
      "step": 1628
    },
    {
      "epoch": 1.2220555138784697,
      "grad_norm": 0.28957399725914,
      "learning_rate": 3.9440088995160676e-05,
      "loss": 0.0867,
      "step": 1629
    },
    {
      "epoch": 1.2228057014253564,
      "grad_norm": 0.15784677863121033,
      "learning_rate": 3.937609789003871e-05,
      "loss": 0.0456,
      "step": 1630
    },
    {
      "epoch": 1.223555888972243,
      "grad_norm": 0.11543436348438263,
      "learning_rate": 3.93121250038627e-05,
      "loss": 0.0415,
      "step": 1631
    },
    {
      "epoch": 1.2243060765191298,
      "grad_norm": 0.12811176478862762,
      "learning_rate": 3.924817044633985e-05,
      "loss": 0.0303,
      "step": 1632
    },
    {
      "epoch": 1.2250562640660165,
      "grad_norm": 0.1936388462781906,
      "learning_rate": 3.9184234327145954e-05,
      "loss": 0.0455,
      "step": 1633
    },
    {
      "epoch": 1.2258064516129032,
      "grad_norm": 0.14532004296779633,
      "learning_rate": 3.912031675592512e-05,
      "loss": 0.0456,
      "step": 1634
    },
    {
      "epoch": 1.22655663915979,
      "grad_norm": 0.1502491980791092,
      "learning_rate": 3.905641784228972e-05,
      "loss": 0.029,
      "step": 1635
    },
    {
      "epoch": 1.2273068267066767,
      "grad_norm": 0.17692214250564575,
      "learning_rate": 3.899253769582008e-05,
      "loss": 0.0432,
      "step": 1636
    },
    {
      "epoch": 1.2280570142535634,
      "grad_norm": 0.18477126955986023,
      "learning_rate": 3.8928676426064376e-05,
      "loss": 0.0387,
      "step": 1637
    },
    {
      "epoch": 1.22880720180045,
      "grad_norm": 0.2677137851715088,
      "learning_rate": 3.886483414253838e-05,
      "loss": 0.0412,
      "step": 1638
    },
    {
      "epoch": 1.2295573893473368,
      "grad_norm": 0.1280679851770401,
      "learning_rate": 3.880101095472535e-05,
      "loss": 0.0333,
      "step": 1639
    },
    {
      "epoch": 1.2303075768942235,
      "grad_norm": 0.18574625253677368,
      "learning_rate": 3.873720697207572e-05,
      "loss": 0.0414,
      "step": 1640
    },
    {
      "epoch": 1.2310577644411103,
      "grad_norm": 0.13856203854084015,
      "learning_rate": 3.867342230400707e-05,
      "loss": 0.0311,
      "step": 1641
    },
    {
      "epoch": 1.231807951987997,
      "grad_norm": 0.13447070121765137,
      "learning_rate": 3.860965705990383e-05,
      "loss": 0.0324,
      "step": 1642
    },
    {
      "epoch": 1.2325581395348837,
      "grad_norm": 0.24870045483112335,
      "learning_rate": 3.8545911349117114e-05,
      "loss": 0.0619,
      "step": 1643
    },
    {
      "epoch": 1.2333083270817704,
      "grad_norm": 0.1902378350496292,
      "learning_rate": 3.848218528096452e-05,
      "loss": 0.0339,
      "step": 1644
    },
    {
      "epoch": 1.2340585146286571,
      "grad_norm": 0.18414323031902313,
      "learning_rate": 3.841847896473001e-05,
      "loss": 0.0388,
      "step": 1645
    },
    {
      "epoch": 1.2348087021755438,
      "grad_norm": 0.14399470388889313,
      "learning_rate": 3.83547925096636e-05,
      "loss": 0.0362,
      "step": 1646
    },
    {
      "epoch": 1.2355588897224306,
      "grad_norm": 0.1719599962234497,
      "learning_rate": 3.829112602498132e-05,
      "loss": 0.0626,
      "step": 1647
    },
    {
      "epoch": 1.2363090772693173,
      "grad_norm": 0.1415622979402542,
      "learning_rate": 3.822747961986493e-05,
      "loss": 0.0371,
      "step": 1648
    },
    {
      "epoch": 1.237059264816204,
      "grad_norm": 0.15044745802879333,
      "learning_rate": 3.816385340346171e-05,
      "loss": 0.0311,
      "step": 1649
    },
    {
      "epoch": 1.2378094523630907,
      "grad_norm": 0.18542815744876862,
      "learning_rate": 3.81002474848844e-05,
      "loss": 0.0399,
      "step": 1650
    },
    {
      "epoch": 1.2385596399099774,
      "grad_norm": 0.17152507603168488,
      "learning_rate": 3.803666197321084e-05,
      "loss": 0.0436,
      "step": 1651
    },
    {
      "epoch": 1.2393098274568641,
      "grad_norm": 0.19231060147285461,
      "learning_rate": 3.797309697748396e-05,
      "loss": 0.0457,
      "step": 1652
    },
    {
      "epoch": 1.2400600150037508,
      "grad_norm": 0.1155783087015152,
      "learning_rate": 3.7909552606711454e-05,
      "loss": 0.0245,
      "step": 1653
    },
    {
      "epoch": 1.2408102025506376,
      "grad_norm": 0.14751829206943512,
      "learning_rate": 3.784602896986566e-05,
      "loss": 0.0289,
      "step": 1654
    },
    {
      "epoch": 1.2415603900975243,
      "grad_norm": 0.17149440944194794,
      "learning_rate": 3.778252617588334e-05,
      "loss": 0.0265,
      "step": 1655
    },
    {
      "epoch": 1.2423105776444112,
      "grad_norm": 0.16274116933345795,
      "learning_rate": 3.771904433366557e-05,
      "loss": 0.0379,
      "step": 1656
    },
    {
      "epoch": 1.2430607651912977,
      "grad_norm": 0.17734788358211517,
      "learning_rate": 3.7655583552077446e-05,
      "loss": 0.0609,
      "step": 1657
    },
    {
      "epoch": 1.2438109527381846,
      "grad_norm": 0.16764874756336212,
      "learning_rate": 3.7592143939947955e-05,
      "loss": 0.0342,
      "step": 1658
    },
    {
      "epoch": 1.2445611402850711,
      "grad_norm": 0.14420807361602783,
      "learning_rate": 3.7528725606069774e-05,
      "loss": 0.0497,
      "step": 1659
    },
    {
      "epoch": 1.245311327831958,
      "grad_norm": 0.17944395542144775,
      "learning_rate": 3.746532865919913e-05,
      "loss": 0.0358,
      "step": 1660
    },
    {
      "epoch": 1.2460615153788448,
      "grad_norm": 0.18467660248279572,
      "learning_rate": 3.740195320805551e-05,
      "loss": 0.0375,
      "step": 1661
    },
    {
      "epoch": 1.2468117029257315,
      "grad_norm": 0.23316818475723267,
      "learning_rate": 3.733859936132158e-05,
      "loss": 0.0444,
      "step": 1662
    },
    {
      "epoch": 1.2475618904726182,
      "grad_norm": 0.21662329137325287,
      "learning_rate": 3.727526722764297e-05,
      "loss": 0.0542,
      "step": 1663
    },
    {
      "epoch": 1.248312078019505,
      "grad_norm": 0.15543553233146667,
      "learning_rate": 3.7211956915628035e-05,
      "loss": 0.0433,
      "step": 1664
    },
    {
      "epoch": 1.2490622655663917,
      "grad_norm": 0.1937057077884674,
      "learning_rate": 3.7148668533847744e-05,
      "loss": 0.05,
      "step": 1665
    },
    {
      "epoch": 1.2498124531132784,
      "grad_norm": 0.18624050915241241,
      "learning_rate": 3.7085402190835406e-05,
      "loss": 0.0468,
      "step": 1666
    },
    {
      "epoch": 1.250562640660165,
      "grad_norm": 0.2042441964149475,
      "learning_rate": 3.702215799508659e-05,
      "loss": 0.0623,
      "step": 1667
    },
    {
      "epoch": 1.2513128282070518,
      "grad_norm": 0.15845929086208344,
      "learning_rate": 3.695893605505887e-05,
      "loss": 0.0487,
      "step": 1668
    },
    {
      "epoch": 1.2520630157539385,
      "grad_norm": 0.13065217435359955,
      "learning_rate": 3.689573647917162e-05,
      "loss": 0.039,
      "step": 1669
    },
    {
      "epoch": 1.2528132033008252,
      "grad_norm": 0.2207261621952057,
      "learning_rate": 3.683255937580592e-05,
      "loss": 0.0476,
      "step": 1670
    },
    {
      "epoch": 1.253563390847712,
      "grad_norm": 0.18256017565727234,
      "learning_rate": 3.6769404853304276e-05,
      "loss": 0.0377,
      "step": 1671
    },
    {
      "epoch": 1.2543135783945987,
      "grad_norm": 0.15684552490711212,
      "learning_rate": 3.670627301997047e-05,
      "loss": 0.0527,
      "step": 1672
    },
    {
      "epoch": 1.2550637659414854,
      "grad_norm": 0.13574464619159698,
      "learning_rate": 3.664316398406939e-05,
      "loss": 0.0327,
      "step": 1673
    },
    {
      "epoch": 1.255813953488372,
      "grad_norm": 0.14744171500205994,
      "learning_rate": 3.658007785382679e-05,
      "loss": 0.0409,
      "step": 1674
    },
    {
      "epoch": 1.2565641410352588,
      "grad_norm": 0.1688900589942932,
      "learning_rate": 3.65170147374292e-05,
      "loss": 0.0461,
      "step": 1675
    },
    {
      "epoch": 1.2573143285821455,
      "grad_norm": 0.14750975370407104,
      "learning_rate": 3.645397474302363e-05,
      "loss": 0.039,
      "step": 1676
    },
    {
      "epoch": 1.2580645161290323,
      "grad_norm": 0.18498747050762177,
      "learning_rate": 3.639095797871748e-05,
      "loss": 0.0432,
      "step": 1677
    },
    {
      "epoch": 1.258814703675919,
      "grad_norm": 0.19537626206874847,
      "learning_rate": 3.63279645525783e-05,
      "loss": 0.0564,
      "step": 1678
    },
    {
      "epoch": 1.2595648912228057,
      "grad_norm": 0.11299839615821838,
      "learning_rate": 3.626499457263359e-05,
      "loss": 0.0265,
      "step": 1679
    },
    {
      "epoch": 1.2603150787696924,
      "grad_norm": 0.20593377947807312,
      "learning_rate": 3.620204814687069e-05,
      "loss": 0.0464,
      "step": 1680
    },
    {
      "epoch": 1.2610652663165791,
      "grad_norm": 0.16698966920375824,
      "learning_rate": 3.61391253832365e-05,
      "loss": 0.0484,
      "step": 1681
    },
    {
      "epoch": 1.2618154538634658,
      "grad_norm": 0.18022559583187103,
      "learning_rate": 3.607622638963739e-05,
      "loss": 0.0477,
      "step": 1682
    },
    {
      "epoch": 1.2625656414103525,
      "grad_norm": 0.18286746740341187,
      "learning_rate": 3.601335127393889e-05,
      "loss": 0.0473,
      "step": 1683
    },
    {
      "epoch": 1.2633158289572393,
      "grad_norm": 0.20090457797050476,
      "learning_rate": 3.59505001439657e-05,
      "loss": 0.0599,
      "step": 1684
    },
    {
      "epoch": 1.264066016504126,
      "grad_norm": 0.2767641544342041,
      "learning_rate": 3.588767310750127e-05,
      "loss": 0.0922,
      "step": 1685
    },
    {
      "epoch": 1.2648162040510127,
      "grad_norm": 0.2000061422586441,
      "learning_rate": 3.5824870272287815e-05,
      "loss": 0.0663,
      "step": 1686
    },
    {
      "epoch": 1.2655663915978994,
      "grad_norm": 0.2778478264808655,
      "learning_rate": 3.576209174602597e-05,
      "loss": 0.0472,
      "step": 1687
    },
    {
      "epoch": 1.2663165791447861,
      "grad_norm": 0.19717861711978912,
      "learning_rate": 3.569933763637477e-05,
      "loss": 0.0452,
      "step": 1688
    },
    {
      "epoch": 1.2670667666916728,
      "grad_norm": 0.11952456831932068,
      "learning_rate": 3.56366080509513e-05,
      "loss": 0.0273,
      "step": 1689
    },
    {
      "epoch": 1.2678169542385596,
      "grad_norm": 0.22591383755207062,
      "learning_rate": 3.557390309733065e-05,
      "loss": 0.0721,
      "step": 1690
    },
    {
      "epoch": 1.2685671417854465,
      "grad_norm": 0.1697269082069397,
      "learning_rate": 3.551122288304561e-05,
      "loss": 0.0456,
      "step": 1691
    },
    {
      "epoch": 1.269317329332333,
      "grad_norm": 0.1867348998785019,
      "learning_rate": 3.544856751558659e-05,
      "loss": 0.0837,
      "step": 1692
    },
    {
      "epoch": 1.27006751687922,
      "grad_norm": 0.16158032417297363,
      "learning_rate": 3.538593710240139e-05,
      "loss": 0.0304,
      "step": 1693
    },
    {
      "epoch": 1.2708177044261064,
      "grad_norm": 0.15402421355247498,
      "learning_rate": 3.532333175089498e-05,
      "loss": 0.0466,
      "step": 1694
    },
    {
      "epoch": 1.2715678919729934,
      "grad_norm": 0.18709032237529755,
      "learning_rate": 3.526075156842938e-05,
      "loss": 0.0432,
      "step": 1695
    },
    {
      "epoch": 1.2723180795198799,
      "grad_norm": 0.23994222283363342,
      "learning_rate": 3.519819666232345e-05,
      "loss": 0.0497,
      "step": 1696
    },
    {
      "epoch": 1.2730682670667668,
      "grad_norm": 0.12393349409103394,
      "learning_rate": 3.5135667139852654e-05,
      "loss": 0.0319,
      "step": 1697
    },
    {
      "epoch": 1.2738184546136533,
      "grad_norm": 0.17412930727005005,
      "learning_rate": 3.507316310824902e-05,
      "loss": 0.0488,
      "step": 1698
    },
    {
      "epoch": 1.2745686421605402,
      "grad_norm": 0.1547398716211319,
      "learning_rate": 3.50106846747008e-05,
      "loss": 0.045,
      "step": 1699
    },
    {
      "epoch": 1.275318829707427,
      "grad_norm": 0.18785391747951508,
      "learning_rate": 3.4948231946352314e-05,
      "loss": 0.0541,
      "step": 1700
    },
    {
      "epoch": 1.2760690172543137,
      "grad_norm": 0.11669635027647018,
      "learning_rate": 3.488580503030389e-05,
      "loss": 0.0396,
      "step": 1701
    },
    {
      "epoch": 1.2768192048012004,
      "grad_norm": 0.16712819039821625,
      "learning_rate": 3.482340403361151e-05,
      "loss": 0.0411,
      "step": 1702
    },
    {
      "epoch": 1.277569392348087,
      "grad_norm": 0.17171481251716614,
      "learning_rate": 3.4761029063286745e-05,
      "loss": 0.0426,
      "step": 1703
    },
    {
      "epoch": 1.2783195798949738,
      "grad_norm": 0.14547479152679443,
      "learning_rate": 3.4698680226296526e-05,
      "loss": 0.0384,
      "step": 1704
    },
    {
      "epoch": 1.2790697674418605,
      "grad_norm": 0.18368588387966156,
      "learning_rate": 3.4636357629562986e-05,
      "loss": 0.05,
      "step": 1705
    },
    {
      "epoch": 1.2798199549887472,
      "grad_norm": 0.17419405281543732,
      "learning_rate": 3.457406137996321e-05,
      "loss": 0.052,
      "step": 1706
    },
    {
      "epoch": 1.280570142535634,
      "grad_norm": 0.13105162978172302,
      "learning_rate": 3.4511791584329154e-05,
      "loss": 0.0368,
      "step": 1707
    },
    {
      "epoch": 1.2813203300825207,
      "grad_norm": 0.1408979743719101,
      "learning_rate": 3.4449548349447394e-05,
      "loss": 0.0459,
      "step": 1708
    },
    {
      "epoch": 1.2820705176294074,
      "grad_norm": 0.15759265422821045,
      "learning_rate": 3.438733178205892e-05,
      "loss": 0.0293,
      "step": 1709
    },
    {
      "epoch": 1.282820705176294,
      "grad_norm": 0.18341770768165588,
      "learning_rate": 3.4325141988859046e-05,
      "loss": 0.0572,
      "step": 1710
    },
    {
      "epoch": 1.2835708927231808,
      "grad_norm": 0.1435495913028717,
      "learning_rate": 3.426297907649711e-05,
      "loss": 0.028,
      "step": 1711
    },
    {
      "epoch": 1.2843210802700675,
      "grad_norm": 0.1692110002040863,
      "learning_rate": 3.4200843151576414e-05,
      "loss": 0.0424,
      "step": 1712
    },
    {
      "epoch": 1.2850712678169542,
      "grad_norm": 0.13991141319274902,
      "learning_rate": 3.413873432065394e-05,
      "loss": 0.0439,
      "step": 1713
    },
    {
      "epoch": 1.285821455363841,
      "grad_norm": 0.1332203447818756,
      "learning_rate": 3.407665269024024e-05,
      "loss": 0.0324,
      "step": 1714
    },
    {
      "epoch": 1.2865716429107277,
      "grad_norm": 0.26455023884773254,
      "learning_rate": 3.401459836679917e-05,
      "loss": 0.0664,
      "step": 1715
    },
    {
      "epoch": 1.2873218304576144,
      "grad_norm": 0.14125743508338928,
      "learning_rate": 3.39525714567478e-05,
      "loss": 0.0454,
      "step": 1716
    },
    {
      "epoch": 1.288072018004501,
      "grad_norm": 0.17687004804611206,
      "learning_rate": 3.389057206645614e-05,
      "loss": 0.0413,
      "step": 1717
    },
    {
      "epoch": 1.2888222055513878,
      "grad_norm": 0.21190932393074036,
      "learning_rate": 3.382860030224708e-05,
      "loss": 0.0421,
      "step": 1718
    },
    {
      "epoch": 1.2895723930982745,
      "grad_norm": 0.16498279571533203,
      "learning_rate": 3.3766656270396074e-05,
      "loss": 0.0589,
      "step": 1719
    },
    {
      "epoch": 1.2903225806451613,
      "grad_norm": 0.20356465876102448,
      "learning_rate": 3.3704740077131036e-05,
      "loss": 0.0408,
      "step": 1720
    },
    {
      "epoch": 1.291072768192048,
      "grad_norm": 0.12303636223077774,
      "learning_rate": 3.3642851828632155e-05,
      "loss": 0.0296,
      "step": 1721
    },
    {
      "epoch": 1.2918229557389347,
      "grad_norm": 0.1438593566417694,
      "learning_rate": 3.3580991631031656e-05,
      "loss": 0.0349,
      "step": 1722
    },
    {
      "epoch": 1.2925731432858214,
      "grad_norm": 0.13478383421897888,
      "learning_rate": 3.3519159590413715e-05,
      "loss": 0.0339,
      "step": 1723
    },
    {
      "epoch": 1.2933233308327081,
      "grad_norm": 0.20928816497325897,
      "learning_rate": 3.345735581281417e-05,
      "loss": 0.0705,
      "step": 1724
    },
    {
      "epoch": 1.2940735183795948,
      "grad_norm": 0.15648962557315826,
      "learning_rate": 3.339558040422042e-05,
      "loss": 0.0361,
      "step": 1725
    },
    {
      "epoch": 1.2948237059264815,
      "grad_norm": 0.14362314343452454,
      "learning_rate": 3.333383347057123e-05,
      "loss": 0.0453,
      "step": 1726
    },
    {
      "epoch": 1.2955738934733683,
      "grad_norm": 0.2117234170436859,
      "learning_rate": 3.3272115117756476e-05,
      "loss": 0.0385,
      "step": 1727
    },
    {
      "epoch": 1.296324081020255,
      "grad_norm": 0.1841832548379898,
      "learning_rate": 3.3210425451617074e-05,
      "loss": 0.0396,
      "step": 1728
    },
    {
      "epoch": 1.2970742685671417,
      "grad_norm": 0.13548365235328674,
      "learning_rate": 3.314876457794474e-05,
      "loss": 0.0256,
      "step": 1729
    },
    {
      "epoch": 1.2978244561140286,
      "grad_norm": 0.1532963514328003,
      "learning_rate": 3.3087132602481774e-05,
      "loss": 0.0306,
      "step": 1730
    },
    {
      "epoch": 1.2985746436609151,
      "grad_norm": 0.14116926491260529,
      "learning_rate": 3.302552963092096e-05,
      "loss": 0.0312,
      "step": 1731
    },
    {
      "epoch": 1.299324831207802,
      "grad_norm": 0.1805046647787094,
      "learning_rate": 3.296395576890532e-05,
      "loss": 0.0597,
      "step": 1732
    },
    {
      "epoch": 1.3000750187546886,
      "grad_norm": 0.20268777012825012,
      "learning_rate": 3.290241112202797e-05,
      "loss": 0.051,
      "step": 1733
    },
    {
      "epoch": 1.3008252063015755,
      "grad_norm": 0.45598918199539185,
      "learning_rate": 3.284089579583192e-05,
      "loss": 0.0613,
      "step": 1734
    },
    {
      "epoch": 1.301575393848462,
      "grad_norm": 0.15112002193927765,
      "learning_rate": 3.2779409895809886e-05,
      "loss": 0.0344,
      "step": 1735
    },
    {
      "epoch": 1.302325581395349,
      "grad_norm": 0.13411392271518707,
      "learning_rate": 3.2717953527404155e-05,
      "loss": 0.0368,
      "step": 1736
    },
    {
      "epoch": 1.3030757689422354,
      "grad_norm": 0.2052137404680252,
      "learning_rate": 3.265652679600631e-05,
      "loss": 0.0556,
      "step": 1737
    },
    {
      "epoch": 1.3038259564891224,
      "grad_norm": 0.21926912665367126,
      "learning_rate": 3.25951298069572e-05,
      "loss": 0.0439,
      "step": 1738
    },
    {
      "epoch": 1.304576144036009,
      "grad_norm": 0.18303723633289337,
      "learning_rate": 3.253376266554655e-05,
      "loss": 0.0514,
      "step": 1739
    },
    {
      "epoch": 1.3053263315828958,
      "grad_norm": 0.188846617937088,
      "learning_rate": 3.247242547701301e-05,
      "loss": 0.0406,
      "step": 1740
    },
    {
      "epoch": 1.3060765191297825,
      "grad_norm": 0.22483086585998535,
      "learning_rate": 3.241111834654382e-05,
      "loss": 0.0608,
      "step": 1741
    },
    {
      "epoch": 1.3068267066766692,
      "grad_norm": 0.19677475094795227,
      "learning_rate": 3.234984137927464e-05,
      "loss": 0.0471,
      "step": 1742
    },
    {
      "epoch": 1.307576894223556,
      "grad_norm": 0.20563624799251556,
      "learning_rate": 3.228859468028946e-05,
      "loss": 0.0508,
      "step": 1743
    },
    {
      "epoch": 1.3083270817704427,
      "grad_norm": 0.3104844391345978,
      "learning_rate": 3.222737835462034e-05,
      "loss": 0.0721,
      "step": 1744
    },
    {
      "epoch": 1.3090772693173294,
      "grad_norm": 0.12385188043117523,
      "learning_rate": 3.216619250724724e-05,
      "loss": 0.0274,
      "step": 1745
    },
    {
      "epoch": 1.309827456864216,
      "grad_norm": 0.16119906306266785,
      "learning_rate": 3.2105037243097866e-05,
      "loss": 0.0411,
      "step": 1746
    },
    {
      "epoch": 1.3105776444111028,
      "grad_norm": 0.2001246064901352,
      "learning_rate": 3.2043912667047465e-05,
      "loss": 0.0504,
      "step": 1747
    },
    {
      "epoch": 1.3113278319579895,
      "grad_norm": 0.14651797711849213,
      "learning_rate": 3.198281888391869e-05,
      "loss": 0.033,
      "step": 1748
    },
    {
      "epoch": 1.3120780195048762,
      "grad_norm": 0.18450623750686646,
      "learning_rate": 3.192175599848133e-05,
      "loss": 0.0469,
      "step": 1749
    },
    {
      "epoch": 1.312828207051763,
      "grad_norm": 0.1982034295797348,
      "learning_rate": 3.1860724115452234e-05,
      "loss": 0.0459,
      "step": 1750
    },
    {
      "epoch": 1.3135783945986497,
      "grad_norm": 0.12932223081588745,
      "learning_rate": 3.179972333949509e-05,
      "loss": 0.0352,
      "step": 1751
    },
    {
      "epoch": 1.3143285821455364,
      "grad_norm": 0.18193656206130981,
      "learning_rate": 3.173875377522019e-05,
      "loss": 0.0422,
      "step": 1752
    },
    {
      "epoch": 1.315078769692423,
      "grad_norm": 0.18322080373764038,
      "learning_rate": 3.167781552718435e-05,
      "loss": 0.0373,
      "step": 1753
    },
    {
      "epoch": 1.3158289572393098,
      "grad_norm": 0.16063223779201508,
      "learning_rate": 3.161690869989068e-05,
      "loss": 0.0458,
      "step": 1754
    },
    {
      "epoch": 1.3165791447861965,
      "grad_norm": 0.1955423504114151,
      "learning_rate": 3.155603339778837e-05,
      "loss": 0.0442,
      "step": 1755
    },
    {
      "epoch": 1.3173293323330832,
      "grad_norm": 0.18707959353923798,
      "learning_rate": 3.149518972527257e-05,
      "loss": 0.038,
      "step": 1756
    },
    {
      "epoch": 1.31807951987997,
      "grad_norm": 0.19283926486968994,
      "learning_rate": 3.1434377786684197e-05,
      "loss": 0.0521,
      "step": 1757
    },
    {
      "epoch": 1.3188297074268567,
      "grad_norm": 0.14124998450279236,
      "learning_rate": 3.137359768630972e-05,
      "loss": 0.0356,
      "step": 1758
    },
    {
      "epoch": 1.3195798949737434,
      "grad_norm": 0.17933954298496246,
      "learning_rate": 3.131284952838106e-05,
      "loss": 0.039,
      "step": 1759
    },
    {
      "epoch": 1.32033008252063,
      "grad_norm": 0.13622942566871643,
      "learning_rate": 3.125213341707528e-05,
      "loss": 0.0468,
      "step": 1760
    },
    {
      "epoch": 1.3210802700675168,
      "grad_norm": 0.19437870383262634,
      "learning_rate": 3.1191449456514575e-05,
      "loss": 0.0451,
      "step": 1761
    },
    {
      "epoch": 1.3218304576144035,
      "grad_norm": 0.14863534271717072,
      "learning_rate": 3.113079775076593e-05,
      "loss": 0.0361,
      "step": 1762
    },
    {
      "epoch": 1.3225806451612903,
      "grad_norm": 0.19207067787647247,
      "learning_rate": 3.107017840384107e-05,
      "loss": 0.053,
      "step": 1763
    },
    {
      "epoch": 1.323330832708177,
      "grad_norm": 0.1651337742805481,
      "learning_rate": 3.100959151969619e-05,
      "loss": 0.0484,
      "step": 1764
    },
    {
      "epoch": 1.3240810202550637,
      "grad_norm": 0.1622065007686615,
      "learning_rate": 3.0949037202231826e-05,
      "loss": 0.0361,
      "step": 1765
    },
    {
      "epoch": 1.3248312078019504,
      "grad_norm": 0.20351789891719818,
      "learning_rate": 3.08885155552927e-05,
      "loss": 0.0384,
      "step": 1766
    },
    {
      "epoch": 1.3255813953488373,
      "grad_norm": 0.1585976481437683,
      "learning_rate": 3.082802668266743e-05,
      "loss": 0.0434,
      "step": 1767
    },
    {
      "epoch": 1.3263315828957238,
      "grad_norm": 0.1370503306388855,
      "learning_rate": 3.076757068808852e-05,
      "loss": 0.0394,
      "step": 1768
    },
    {
      "epoch": 1.3270817704426108,
      "grad_norm": 0.2939653694629669,
      "learning_rate": 3.070714767523203e-05,
      "loss": 0.0864,
      "step": 1769
    },
    {
      "epoch": 1.3278319579894973,
      "grad_norm": 0.19903366267681122,
      "learning_rate": 3.0646757747717475e-05,
      "loss": 0.0481,
      "step": 1770
    },
    {
      "epoch": 1.3285821455363842,
      "grad_norm": 0.1912355273962021,
      "learning_rate": 3.0586401009107636e-05,
      "loss": 0.0426,
      "step": 1771
    },
    {
      "epoch": 1.3293323330832707,
      "grad_norm": 0.2867732048034668,
      "learning_rate": 3.0526077562908386e-05,
      "loss": 0.0628,
      "step": 1772
    },
    {
      "epoch": 1.3300825206301576,
      "grad_norm": 0.20162878930568695,
      "learning_rate": 3.0465787512568466e-05,
      "loss": 0.0463,
      "step": 1773
    },
    {
      "epoch": 1.3308327081770441,
      "grad_norm": 0.15980598330497742,
      "learning_rate": 3.040553096147942e-05,
      "loss": 0.0481,
      "step": 1774
    },
    {
      "epoch": 1.331582895723931,
      "grad_norm": 0.20228786766529083,
      "learning_rate": 3.0345308012975255e-05,
      "loss": 0.05,
      "step": 1775
    },
    {
      "epoch": 1.3323330832708178,
      "grad_norm": 0.1600971817970276,
      "learning_rate": 3.0285118770332428e-05,
      "loss": 0.0336,
      "step": 1776
    },
    {
      "epoch": 1.3330832708177045,
      "grad_norm": 0.18647384643554688,
      "learning_rate": 3.022496333676954e-05,
      "loss": 0.048,
      "step": 1777
    },
    {
      "epoch": 1.3338334583645912,
      "grad_norm": 0.18941383063793182,
      "learning_rate": 3.0164841815447263e-05,
      "loss": 0.0525,
      "step": 1778
    },
    {
      "epoch": 1.334583645911478,
      "grad_norm": 0.1617792248725891,
      "learning_rate": 3.0104754309468066e-05,
      "loss": 0.036,
      "step": 1779
    },
    {
      "epoch": 1.3353338334583646,
      "grad_norm": 0.26556825637817383,
      "learning_rate": 3.00447009218761e-05,
      "loss": 0.0658,
      "step": 1780
    },
    {
      "epoch": 1.3360840210052514,
      "grad_norm": 0.17816261947155,
      "learning_rate": 2.9984681755657017e-05,
      "loss": 0.0447,
      "step": 1781
    },
    {
      "epoch": 1.336834208552138,
      "grad_norm": 0.17538125813007355,
      "learning_rate": 2.9924696913737792e-05,
      "loss": 0.0626,
      "step": 1782
    },
    {
      "epoch": 1.3375843960990248,
      "grad_norm": 0.18380333483219147,
      "learning_rate": 2.986474649898651e-05,
      "loss": 0.0449,
      "step": 1783
    },
    {
      "epoch": 1.3383345836459115,
      "grad_norm": 0.17858687043190002,
      "learning_rate": 2.9804830614212242e-05,
      "loss": 0.0371,
      "step": 1784
    },
    {
      "epoch": 1.3390847711927982,
      "grad_norm": 0.19844569265842438,
      "learning_rate": 2.9744949362164798e-05,
      "loss": 0.0414,
      "step": 1785
    },
    {
      "epoch": 1.339834958739685,
      "grad_norm": 0.1734311580657959,
      "learning_rate": 2.9685102845534658e-05,
      "loss": 0.042,
      "step": 1786
    },
    {
      "epoch": 1.3405851462865717,
      "grad_norm": 0.1407657116651535,
      "learning_rate": 2.9625291166952702e-05,
      "loss": 0.0615,
      "step": 1787
    },
    {
      "epoch": 1.3413353338334584,
      "grad_norm": 0.20041944086551666,
      "learning_rate": 2.956551442899005e-05,
      "loss": 0.0642,
      "step": 1788
    },
    {
      "epoch": 1.342085521380345,
      "grad_norm": 0.18350687623023987,
      "learning_rate": 2.9505772734157948e-05,
      "loss": 0.0562,
      "step": 1789
    },
    {
      "epoch": 1.3428357089272318,
      "grad_norm": 0.20079989731311798,
      "learning_rate": 2.9446066184907495e-05,
      "loss": 0.0646,
      "step": 1790
    },
    {
      "epoch": 1.3435858964741185,
      "grad_norm": 0.17485657334327698,
      "learning_rate": 2.9386394883629565e-05,
      "loss": 0.0414,
      "step": 1791
    },
    {
      "epoch": 1.3443360840210052,
      "grad_norm": 0.2188062071800232,
      "learning_rate": 2.932675893265454e-05,
      "loss": 0.0834,
      "step": 1792
    },
    {
      "epoch": 1.345086271567892,
      "grad_norm": 0.1565270721912384,
      "learning_rate": 2.926715843425223e-05,
      "loss": 0.0452,
      "step": 1793
    },
    {
      "epoch": 1.3458364591147787,
      "grad_norm": 0.16900846362113953,
      "learning_rate": 2.9207593490631592e-05,
      "loss": 0.0454,
      "step": 1794
    },
    {
      "epoch": 1.3465866466616654,
      "grad_norm": 0.18272441625595093,
      "learning_rate": 2.914806420394064e-05,
      "loss": 0.0546,
      "step": 1795
    },
    {
      "epoch": 1.347336834208552,
      "grad_norm": 0.18571123480796814,
      "learning_rate": 2.908857067626629e-05,
      "loss": 0.0303,
      "step": 1796
    },
    {
      "epoch": 1.3480870217554388,
      "grad_norm": 0.17477184534072876,
      "learning_rate": 2.902911300963403e-05,
      "loss": 0.0569,
      "step": 1797
    },
    {
      "epoch": 1.3488372093023255,
      "grad_norm": 0.16123870015144348,
      "learning_rate": 2.8969691306007918e-05,
      "loss": 0.0316,
      "step": 1798
    },
    {
      "epoch": 1.3495873968492123,
      "grad_norm": 0.14804422855377197,
      "learning_rate": 2.891030566729032e-05,
      "loss": 0.0489,
      "step": 1799
    },
    {
      "epoch": 1.350337584396099,
      "grad_norm": 0.16255329549312592,
      "learning_rate": 2.8850956195321795e-05,
      "loss": 0.0481,
      "step": 1800
    },
    {
      "epoch": 1.350337584396099,
      "eval_loss": 0.06747418642044067,
      "eval_runtime": 2.6568,
      "eval_samples_per_second": 20.325,
      "eval_steps_per_second": 5.27,
      "step": 1800
    },
    {
      "epoch": 1.3510877719429857,
      "grad_norm": 0.2039135992527008,
      "learning_rate": 2.8791642991880784e-05,
      "loss": 0.0394,
      "step": 1801
    },
    {
      "epoch": 1.3518379594898724,
      "grad_norm": 0.1728277951478958,
      "learning_rate": 2.873236615868362e-05,
      "loss": 0.0559,
      "step": 1802
    },
    {
      "epoch": 1.3525881470367591,
      "grad_norm": 0.1896568089723587,
      "learning_rate": 2.8673125797384243e-05,
      "loss": 0.0436,
      "step": 1803
    },
    {
      "epoch": 1.3533383345836458,
      "grad_norm": 0.13357765972614288,
      "learning_rate": 2.8613922009574024e-05,
      "loss": 0.0338,
      "step": 1804
    },
    {
      "epoch": 1.3540885221305325,
      "grad_norm": 0.1681651473045349,
      "learning_rate": 2.8554754896781656e-05,
      "loss": 0.0479,
      "step": 1805
    },
    {
      "epoch": 1.3548387096774195,
      "grad_norm": 0.11592215299606323,
      "learning_rate": 2.8495624560472866e-05,
      "loss": 0.0266,
      "step": 1806
    },
    {
      "epoch": 1.355588897224306,
      "grad_norm": 0.22524559497833252,
      "learning_rate": 2.843653110205039e-05,
      "loss": 0.0572,
      "step": 1807
    },
    {
      "epoch": 1.356339084771193,
      "grad_norm": 0.21835164725780487,
      "learning_rate": 2.8377474622853683e-05,
      "loss": 0.0618,
      "step": 1808
    },
    {
      "epoch": 1.3570892723180794,
      "grad_norm": 0.16861413419246674,
      "learning_rate": 2.8318455224158786e-05,
      "loss": 0.0376,
      "step": 1809
    },
    {
      "epoch": 1.3578394598649663,
      "grad_norm": 0.1737951785326004,
      "learning_rate": 2.8259473007178163e-05,
      "loss": 0.0469,
      "step": 1810
    },
    {
      "epoch": 1.3585896474118528,
      "grad_norm": 0.1801527738571167,
      "learning_rate": 2.8200528073060507e-05,
      "loss": 0.0421,
      "step": 1811
    },
    {
      "epoch": 1.3593398349587398,
      "grad_norm": 0.203287273645401,
      "learning_rate": 2.814162052289058e-05,
      "loss": 0.0465,
      "step": 1812
    },
    {
      "epoch": 1.3600900225056263,
      "grad_norm": 0.14048676192760468,
      "learning_rate": 2.8082750457689033e-05,
      "loss": 0.0327,
      "step": 1813
    },
    {
      "epoch": 1.3608402100525132,
      "grad_norm": 0.16412879526615143,
      "learning_rate": 2.8023917978412207e-05,
      "loss": 0.0349,
      "step": 1814
    },
    {
      "epoch": 1.3615903975994,
      "grad_norm": 0.18040567636489868,
      "learning_rate": 2.7965123185952023e-05,
      "loss": 0.0466,
      "step": 1815
    },
    {
      "epoch": 1.3623405851462866,
      "grad_norm": 0.19568736851215363,
      "learning_rate": 2.7906366181135775e-05,
      "loss": 0.0449,
      "step": 1816
    },
    {
      "epoch": 1.3630907726931734,
      "grad_norm": 0.1617717146873474,
      "learning_rate": 2.7847647064725924e-05,
      "loss": 0.0372,
      "step": 1817
    },
    {
      "epoch": 1.36384096024006,
      "grad_norm": 0.14724799990653992,
      "learning_rate": 2.778896593741999e-05,
      "loss": 0.0315,
      "step": 1818
    },
    {
      "epoch": 1.3645911477869468,
      "grad_norm": 0.09599991142749786,
      "learning_rate": 2.77303228998503e-05,
      "loss": 0.0195,
      "step": 1819
    },
    {
      "epoch": 1.3653413353338335,
      "grad_norm": 0.2500494718551636,
      "learning_rate": 2.7671718052583908e-05,
      "loss": 0.0723,
      "step": 1820
    },
    {
      "epoch": 1.3660915228807202,
      "grad_norm": 0.1894386112689972,
      "learning_rate": 2.7613151496122347e-05,
      "loss": 0.051,
      "step": 1821
    },
    {
      "epoch": 1.366841710427607,
      "grad_norm": 0.14874202013015747,
      "learning_rate": 2.7554623330901524e-05,
      "loss": 0.0496,
      "step": 1822
    },
    {
      "epoch": 1.3675918979744937,
      "grad_norm": 0.24037110805511475,
      "learning_rate": 2.749613365729141e-05,
      "loss": 0.0723,
      "step": 1823
    },
    {
      "epoch": 1.3683420855213804,
      "grad_norm": 0.19027547538280487,
      "learning_rate": 2.7437682575596104e-05,
      "loss": 0.0434,
      "step": 1824
    },
    {
      "epoch": 1.369092273068267,
      "grad_norm": 0.23864826560020447,
      "learning_rate": 2.7379270186053428e-05,
      "loss": 0.0655,
      "step": 1825
    },
    {
      "epoch": 1.3698424606151538,
      "grad_norm": 0.1366167664527893,
      "learning_rate": 2.7320896588834903e-05,
      "loss": 0.0398,
      "step": 1826
    },
    {
      "epoch": 1.3705926481620405,
      "grad_norm": 0.17698831856250763,
      "learning_rate": 2.7262561884045457e-05,
      "loss": 0.0354,
      "step": 1827
    },
    {
      "epoch": 1.3713428357089272,
      "grad_norm": 0.15502697229385376,
      "learning_rate": 2.720426617172339e-05,
      "loss": 0.0398,
      "step": 1828
    },
    {
      "epoch": 1.372093023255814,
      "grad_norm": 0.2184450775384903,
      "learning_rate": 2.71460095518401e-05,
      "loss": 0.0645,
      "step": 1829
    },
    {
      "epoch": 1.3728432108027007,
      "grad_norm": 0.16089852154254913,
      "learning_rate": 2.708779212429996e-05,
      "loss": 0.0406,
      "step": 1830
    },
    {
      "epoch": 1.3735933983495874,
      "grad_norm": 0.13291966915130615,
      "learning_rate": 2.702961398894014e-05,
      "loss": 0.032,
      "step": 1831
    },
    {
      "epoch": 1.374343585896474,
      "grad_norm": 0.2512020766735077,
      "learning_rate": 2.6971475245530375e-05,
      "loss": 0.0538,
      "step": 1832
    },
    {
      "epoch": 1.3750937734433608,
      "grad_norm": 0.18554271757602692,
      "learning_rate": 2.6913375993772915e-05,
      "loss": 0.0419,
      "step": 1833
    },
    {
      "epoch": 1.3758439609902475,
      "grad_norm": 0.18187479674816132,
      "learning_rate": 2.6855316333302237e-05,
      "loss": 0.0516,
      "step": 1834
    },
    {
      "epoch": 1.3765941485371342,
      "grad_norm": 0.2124125361442566,
      "learning_rate": 2.6797296363684977e-05,
      "loss": 0.0442,
      "step": 1835
    },
    {
      "epoch": 1.377344336084021,
      "grad_norm": 0.27118852734565735,
      "learning_rate": 2.6739316184419622e-05,
      "loss": 0.0541,
      "step": 1836
    },
    {
      "epoch": 1.3780945236309077,
      "grad_norm": 0.1586301028728485,
      "learning_rate": 2.6681375894936472e-05,
      "loss": 0.0331,
      "step": 1837
    },
    {
      "epoch": 1.3788447111777944,
      "grad_norm": 0.1749737709760666,
      "learning_rate": 2.662347559459746e-05,
      "loss": 0.0369,
      "step": 1838
    },
    {
      "epoch": 1.379594898724681,
      "grad_norm": 0.10476631671190262,
      "learning_rate": 2.6565615382695896e-05,
      "loss": 0.0264,
      "step": 1839
    },
    {
      "epoch": 1.3803450862715678,
      "grad_norm": 0.1476007103919983,
      "learning_rate": 2.6507795358456307e-05,
      "loss": 0.0586,
      "step": 1840
    },
    {
      "epoch": 1.3810952738184545,
      "grad_norm": 0.11646662652492523,
      "learning_rate": 2.6450015621034362e-05,
      "loss": 0.0252,
      "step": 1841
    },
    {
      "epoch": 1.3818454613653413,
      "grad_norm": 0.15108223259449005,
      "learning_rate": 2.6392276269516613e-05,
      "loss": 0.0314,
      "step": 1842
    },
    {
      "epoch": 1.3825956489122282,
      "grad_norm": 0.1876702457666397,
      "learning_rate": 2.63345774029204e-05,
      "loss": 0.0603,
      "step": 1843
    },
    {
      "epoch": 1.3833458364591147,
      "grad_norm": 0.15735001862049103,
      "learning_rate": 2.6276919120193543e-05,
      "loss": 0.0452,
      "step": 1844
    },
    {
      "epoch": 1.3840960240060016,
      "grad_norm": 0.16730938851833344,
      "learning_rate": 2.621930152021434e-05,
      "loss": 0.042,
      "step": 1845
    },
    {
      "epoch": 1.3848462115528881,
      "grad_norm": 0.12933415174484253,
      "learning_rate": 2.6161724701791306e-05,
      "loss": 0.0246,
      "step": 1846
    },
    {
      "epoch": 1.385596399099775,
      "grad_norm": 0.14777973294258118,
      "learning_rate": 2.6104188763663018e-05,
      "loss": 0.0299,
      "step": 1847
    },
    {
      "epoch": 1.3863465866466615,
      "grad_norm": 0.14805126190185547,
      "learning_rate": 2.604669380449795e-05,
      "loss": 0.0303,
      "step": 1848
    },
    {
      "epoch": 1.3870967741935485,
      "grad_norm": 0.1918485164642334,
      "learning_rate": 2.598923992289427e-05,
      "loss": 0.0364,
      "step": 1849
    },
    {
      "epoch": 1.387846961740435,
      "grad_norm": 0.16532783210277557,
      "learning_rate": 2.5931827217379746e-05,
      "loss": 0.0371,
      "step": 1850
    },
    {
      "epoch": 1.388597149287322,
      "grad_norm": 0.14476117491722107,
      "learning_rate": 2.5874455786411505e-05,
      "loss": 0.0367,
      "step": 1851
    },
    {
      "epoch": 1.3893473368342086,
      "grad_norm": 0.12052533775568008,
      "learning_rate": 2.5817125728375912e-05,
      "loss": 0.023,
      "step": 1852
    },
    {
      "epoch": 1.3900975243810954,
      "grad_norm": 0.13191531598567963,
      "learning_rate": 2.5759837141588362e-05,
      "loss": 0.0259,
      "step": 1853
    },
    {
      "epoch": 1.390847711927982,
      "grad_norm": 0.27699872851371765,
      "learning_rate": 2.5702590124293147e-05,
      "loss": 0.045,
      "step": 1854
    },
    {
      "epoch": 1.3915978994748688,
      "grad_norm": 0.16797631978988647,
      "learning_rate": 2.5645384774663262e-05,
      "loss": 0.0402,
      "step": 1855
    },
    {
      "epoch": 1.3923480870217555,
      "grad_norm": 0.11576727032661438,
      "learning_rate": 2.5588221190800264e-05,
      "loss": 0.0383,
      "step": 1856
    },
    {
      "epoch": 1.3930982745686422,
      "grad_norm": 0.3373456597328186,
      "learning_rate": 2.5531099470734038e-05,
      "loss": 0.0621,
      "step": 1857
    },
    {
      "epoch": 1.393848462115529,
      "grad_norm": 0.17893192172050476,
      "learning_rate": 2.5474019712422724e-05,
      "loss": 0.0275,
      "step": 1858
    },
    {
      "epoch": 1.3945986496624156,
      "grad_norm": 0.2578313648700714,
      "learning_rate": 2.541698201375249e-05,
      "loss": 0.0579,
      "step": 1859
    },
    {
      "epoch": 1.3953488372093024,
      "grad_norm": 0.15810662508010864,
      "learning_rate": 2.5359986472537373e-05,
      "loss": 0.0391,
      "step": 1860
    },
    {
      "epoch": 1.396099024756189,
      "grad_norm": 0.15665064752101898,
      "learning_rate": 2.530303318651913e-05,
      "loss": 0.0357,
      "step": 1861
    },
    {
      "epoch": 1.3968492123030758,
      "grad_norm": 0.187716543674469,
      "learning_rate": 2.5246122253366998e-05,
      "loss": 0.0397,
      "step": 1862
    },
    {
      "epoch": 1.3975993998499625,
      "grad_norm": 0.1098187267780304,
      "learning_rate": 2.5189253770677644e-05,
      "loss": 0.0242,
      "step": 1863
    },
    {
      "epoch": 1.3983495873968492,
      "grad_norm": 0.13694116473197937,
      "learning_rate": 2.5132427835974926e-05,
      "loss": 0.0208,
      "step": 1864
    },
    {
      "epoch": 1.399099774943736,
      "grad_norm": 0.12738709151744843,
      "learning_rate": 2.507564454670971e-05,
      "loss": 0.0238,
      "step": 1865
    },
    {
      "epoch": 1.3998499624906227,
      "grad_norm": 0.13771750032901764,
      "learning_rate": 2.5018904000259757e-05,
      "loss": 0.0372,
      "step": 1866
    },
    {
      "epoch": 1.4006001500375094,
      "grad_norm": 0.14189371466636658,
      "learning_rate": 2.4962206293929512e-05,
      "loss": 0.0328,
      "step": 1867
    },
    {
      "epoch": 1.401350337584396,
      "grad_norm": 0.1648484170436859,
      "learning_rate": 2.490555152494996e-05,
      "loss": 0.0369,
      "step": 1868
    },
    {
      "epoch": 1.4021005251312828,
      "grad_norm": 0.1412844955921173,
      "learning_rate": 2.4848939790478463e-05,
      "loss": 0.024,
      "step": 1869
    },
    {
      "epoch": 1.4028507126781695,
      "grad_norm": 0.12301353365182877,
      "learning_rate": 2.4792371187598544e-05,
      "loss": 0.0243,
      "step": 1870
    },
    {
      "epoch": 1.4036009002250562,
      "grad_norm": 0.1523589789867401,
      "learning_rate": 2.4735845813319804e-05,
      "loss": 0.0341,
      "step": 1871
    },
    {
      "epoch": 1.404351087771943,
      "grad_norm": 0.2967745065689087,
      "learning_rate": 2.4679363764577683e-05,
      "loss": 0.0538,
      "step": 1872
    },
    {
      "epoch": 1.4051012753188297,
      "grad_norm": 0.19742630422115326,
      "learning_rate": 2.462292513823336e-05,
      "loss": 0.054,
      "step": 1873
    },
    {
      "epoch": 1.4058514628657164,
      "grad_norm": 0.19836518168449402,
      "learning_rate": 2.4566530031073486e-05,
      "loss": 0.0372,
      "step": 1874
    },
    {
      "epoch": 1.406601650412603,
      "grad_norm": 0.2233537882566452,
      "learning_rate": 2.451017853981013e-05,
      "loss": 0.0294,
      "step": 1875
    },
    {
      "epoch": 1.4073518379594898,
      "grad_norm": 0.19185113906860352,
      "learning_rate": 2.4453870761080554e-05,
      "loss": 0.0606,
      "step": 1876
    },
    {
      "epoch": 1.4081020255063765,
      "grad_norm": 0.17896531522274017,
      "learning_rate": 2.4397606791447052e-05,
      "loss": 0.0479,
      "step": 1877
    },
    {
      "epoch": 1.4088522130532632,
      "grad_norm": 0.19202962517738342,
      "learning_rate": 2.4341386727396793e-05,
      "loss": 0.0554,
      "step": 1878
    },
    {
      "epoch": 1.40960240060015,
      "grad_norm": 0.15726254880428314,
      "learning_rate": 2.4285210665341646e-05,
      "loss": 0.0357,
      "step": 1879
    },
    {
      "epoch": 1.4103525881470367,
      "grad_norm": 0.18670770525932312,
      "learning_rate": 2.422907870161803e-05,
      "loss": 0.0381,
      "step": 1880
    },
    {
      "epoch": 1.4111027756939234,
      "grad_norm": 0.3125583231449127,
      "learning_rate": 2.4172990932486733e-05,
      "loss": 0.0461,
      "step": 1881
    },
    {
      "epoch": 1.4118529632408103,
      "grad_norm": 0.19309872388839722,
      "learning_rate": 2.4116947454132782e-05,
      "loss": 0.0294,
      "step": 1882
    },
    {
      "epoch": 1.4126031507876968,
      "grad_norm": 0.17205388844013214,
      "learning_rate": 2.4060948362665176e-05,
      "loss": 0.0387,
      "step": 1883
    },
    {
      "epoch": 1.4133533383345838,
      "grad_norm": 0.17235881090164185,
      "learning_rate": 2.4004993754116867e-05,
      "loss": 0.0367,
      "step": 1884
    },
    {
      "epoch": 1.4141035258814703,
      "grad_norm": 0.180396169424057,
      "learning_rate": 2.39490837244445e-05,
      "loss": 0.0409,
      "step": 1885
    },
    {
      "epoch": 1.4148537134283572,
      "grad_norm": 0.16136233508586884,
      "learning_rate": 2.389321836952828e-05,
      "loss": 0.0439,
      "step": 1886
    },
    {
      "epoch": 1.4156039009752437,
      "grad_norm": 0.23657025396823883,
      "learning_rate": 2.383739778517176e-05,
      "loss": 0.0392,
      "step": 1887
    },
    {
      "epoch": 1.4163540885221306,
      "grad_norm": 0.25855350494384766,
      "learning_rate": 2.3781622067101767e-05,
      "loss": 0.0626,
      "step": 1888
    },
    {
      "epoch": 1.4171042760690171,
      "grad_norm": 0.19857187569141388,
      "learning_rate": 2.372589131096816e-05,
      "loss": 0.0534,
      "step": 1889
    },
    {
      "epoch": 1.417854463615904,
      "grad_norm": 0.1421574205160141,
      "learning_rate": 2.36702056123437e-05,
      "loss": 0.0361,
      "step": 1890
    },
    {
      "epoch": 1.4186046511627908,
      "grad_norm": 0.1884581595659256,
      "learning_rate": 2.3614565066723892e-05,
      "loss": 0.0413,
      "step": 1891
    },
    {
      "epoch": 1.4193548387096775,
      "grad_norm": 0.21691051125526428,
      "learning_rate": 2.355896976952674e-05,
      "loss": 0.0492,
      "step": 1892
    },
    {
      "epoch": 1.4201050262565642,
      "grad_norm": 0.2113209068775177,
      "learning_rate": 2.350341981609276e-05,
      "loss": 0.0581,
      "step": 1893
    },
    {
      "epoch": 1.420855213803451,
      "grad_norm": 0.21820048987865448,
      "learning_rate": 2.344791530168465e-05,
      "loss": 0.0495,
      "step": 1894
    },
    {
      "epoch": 1.4216054013503376,
      "grad_norm": 0.15237504243850708,
      "learning_rate": 2.339245632148715e-05,
      "loss": 0.0431,
      "step": 1895
    },
    {
      "epoch": 1.4223555888972244,
      "grad_norm": 0.18294626474380493,
      "learning_rate": 2.3337042970606965e-05,
      "loss": 0.0436,
      "step": 1896
    },
    {
      "epoch": 1.423105776444111,
      "grad_norm": 0.1634780317544937,
      "learning_rate": 2.3281675344072545e-05,
      "loss": 0.0402,
      "step": 1897
    },
    {
      "epoch": 1.4238559639909978,
      "grad_norm": 0.2057954967021942,
      "learning_rate": 2.3226353536833907e-05,
      "loss": 0.054,
      "step": 1898
    },
    {
      "epoch": 1.4246061515378845,
      "grad_norm": 0.160613551735878,
      "learning_rate": 2.317107764376253e-05,
      "loss": 0.0464,
      "step": 1899
    },
    {
      "epoch": 1.4253563390847712,
      "grad_norm": 0.18292781710624695,
      "learning_rate": 2.3115847759651082e-05,
      "loss": 0.0353,
      "step": 1900
    },
    {
      "epoch": 1.426106526631658,
      "grad_norm": 0.16484901309013367,
      "learning_rate": 2.3060663979213404e-05,
      "loss": 0.0401,
      "step": 1901
    },
    {
      "epoch": 1.4268567141785446,
      "grad_norm": 0.20094150304794312,
      "learning_rate": 2.300552639708423e-05,
      "loss": 0.0437,
      "step": 1902
    },
    {
      "epoch": 1.4276069017254314,
      "grad_norm": 0.2095361202955246,
      "learning_rate": 2.2950435107819124e-05,
      "loss": 0.0466,
      "step": 1903
    },
    {
      "epoch": 1.428357089272318,
      "grad_norm": 0.2117299735546112,
      "learning_rate": 2.2895390205894164e-05,
      "loss": 0.0448,
      "step": 1904
    },
    {
      "epoch": 1.4291072768192048,
      "grad_norm": 0.3316866159439087,
      "learning_rate": 2.2840391785705967e-05,
      "loss": 0.0653,
      "step": 1905
    },
    {
      "epoch": 1.4298574643660915,
      "grad_norm": 0.19901429116725922,
      "learning_rate": 2.278543994157139e-05,
      "loss": 0.0507,
      "step": 1906
    },
    {
      "epoch": 1.4306076519129782,
      "grad_norm": 0.22417782247066498,
      "learning_rate": 2.2730534767727483e-05,
      "loss": 0.0465,
      "step": 1907
    },
    {
      "epoch": 1.431357839459865,
      "grad_norm": 0.16476863622665405,
      "learning_rate": 2.267567635833116e-05,
      "loss": 0.0706,
      "step": 1908
    },
    {
      "epoch": 1.4321080270067517,
      "grad_norm": 0.18171042203903198,
      "learning_rate": 2.2620864807459213e-05,
      "loss": 0.0417,
      "step": 1909
    },
    {
      "epoch": 1.4328582145536384,
      "grad_norm": 0.16148392856121063,
      "learning_rate": 2.2566100209108048e-05,
      "loss": 0.0473,
      "step": 1910
    },
    {
      "epoch": 1.433608402100525,
      "grad_norm": 0.20606650412082672,
      "learning_rate": 2.2511382657193565e-05,
      "loss": 0.0524,
      "step": 1911
    },
    {
      "epoch": 1.4343585896474118,
      "grad_norm": 0.18486949801445007,
      "learning_rate": 2.2456712245550993e-05,
      "loss": 0.0378,
      "step": 1912
    },
    {
      "epoch": 1.4351087771942985,
      "grad_norm": 0.1495230346918106,
      "learning_rate": 2.2402089067934668e-05,
      "loss": 0.0458,
      "step": 1913
    },
    {
      "epoch": 1.4358589647411852,
      "grad_norm": 0.1609950214624405,
      "learning_rate": 2.2347513218017974e-05,
      "loss": 0.0404,
      "step": 1914
    },
    {
      "epoch": 1.436609152288072,
      "grad_norm": 0.20721952617168427,
      "learning_rate": 2.2292984789393122e-05,
      "loss": 0.0455,
      "step": 1915
    },
    {
      "epoch": 1.4373593398349587,
      "grad_norm": 0.10836555808782578,
      "learning_rate": 2.2238503875571028e-05,
      "loss": 0.0248,
      "step": 1916
    },
    {
      "epoch": 1.4381095273818454,
      "grad_norm": 0.2047933042049408,
      "learning_rate": 2.218407056998104e-05,
      "loss": 0.0429,
      "step": 1917
    },
    {
      "epoch": 1.438859714928732,
      "grad_norm": 0.1827002912759781,
      "learning_rate": 2.2129684965970948e-05,
      "loss": 0.0248,
      "step": 1918
    },
    {
      "epoch": 1.439609902475619,
      "grad_norm": 0.2153414636850357,
      "learning_rate": 2.2075347156806697e-05,
      "loss": 0.0475,
      "step": 1919
    },
    {
      "epoch": 1.4403600900225055,
      "grad_norm": 0.19228680431842804,
      "learning_rate": 2.2021057235672288e-05,
      "loss": 0.0578,
      "step": 1920
    },
    {
      "epoch": 1.4411102775693925,
      "grad_norm": 0.17686350643634796,
      "learning_rate": 2.1966815295669585e-05,
      "loss": 0.042,
      "step": 1921
    },
    {
      "epoch": 1.441860465116279,
      "grad_norm": 0.1612776815891266,
      "learning_rate": 2.1912621429818177e-05,
      "loss": 0.0412,
      "step": 1922
    },
    {
      "epoch": 1.442610652663166,
      "grad_norm": 0.16681015491485596,
      "learning_rate": 2.18584757310552e-05,
      "loss": 0.0335,
      "step": 1923
    },
    {
      "epoch": 1.4433608402100524,
      "grad_norm": 0.16631953418254852,
      "learning_rate": 2.1804378292235224e-05,
      "loss": 0.0422,
      "step": 1924
    },
    {
      "epoch": 1.4441110277569393,
      "grad_norm": 0.1537289172410965,
      "learning_rate": 2.1750329206129988e-05,
      "loss": 0.0484,
      "step": 1925
    },
    {
      "epoch": 1.4448612153038258,
      "grad_norm": 0.1737208366394043,
      "learning_rate": 2.1696328565428364e-05,
      "loss": 0.0377,
      "step": 1926
    },
    {
      "epoch": 1.4456114028507128,
      "grad_norm": 0.14561980962753296,
      "learning_rate": 2.1642376462736148e-05,
      "loss": 0.0326,
      "step": 1927
    },
    {
      "epoch": 1.4463615903975993,
      "grad_norm": 0.3279992938041687,
      "learning_rate": 2.158847299057587e-05,
      "loss": 0.0619,
      "step": 1928
    },
    {
      "epoch": 1.4471117779444862,
      "grad_norm": 0.14300592243671417,
      "learning_rate": 2.1534618241386705e-05,
      "loss": 0.0441,
      "step": 1929
    },
    {
      "epoch": 1.447861965491373,
      "grad_norm": 0.19687087833881378,
      "learning_rate": 2.14808123075242e-05,
      "loss": 0.0643,
      "step": 1930
    },
    {
      "epoch": 1.4486121530382596,
      "grad_norm": 0.15415306389331818,
      "learning_rate": 2.1427055281260255e-05,
      "loss": 0.0371,
      "step": 1931
    },
    {
      "epoch": 1.4493623405851463,
      "grad_norm": 0.10696108639240265,
      "learning_rate": 2.1373347254782882e-05,
      "loss": 0.0289,
      "step": 1932
    },
    {
      "epoch": 1.450112528132033,
      "grad_norm": 0.17456509172916412,
      "learning_rate": 2.1319688320196048e-05,
      "loss": 0.0342,
      "step": 1933
    },
    {
      "epoch": 1.4508627156789198,
      "grad_norm": 0.18844592571258545,
      "learning_rate": 2.1266078569519542e-05,
      "loss": 0.0404,
      "step": 1934
    },
    {
      "epoch": 1.4516129032258065,
      "grad_norm": 0.1761346310377121,
      "learning_rate": 2.121251809468882e-05,
      "loss": 0.0444,
      "step": 1935
    },
    {
      "epoch": 1.4523630907726932,
      "grad_norm": 0.14905095100402832,
      "learning_rate": 2.1159006987554807e-05,
      "loss": 0.0456,
      "step": 1936
    },
    {
      "epoch": 1.45311327831958,
      "grad_norm": 0.1570758819580078,
      "learning_rate": 2.1105545339883808e-05,
      "loss": 0.0377,
      "step": 1937
    },
    {
      "epoch": 1.4538634658664666,
      "grad_norm": 0.20766982436180115,
      "learning_rate": 2.1052133243357253e-05,
      "loss": 0.0483,
      "step": 1938
    },
    {
      "epoch": 1.4546136534133534,
      "grad_norm": 0.1821688413619995,
      "learning_rate": 2.0998770789571636e-05,
      "loss": 0.0325,
      "step": 1939
    },
    {
      "epoch": 1.45536384096024,
      "grad_norm": 0.17206883430480957,
      "learning_rate": 2.0945458070038315e-05,
      "loss": 0.0523,
      "step": 1940
    },
    {
      "epoch": 1.4561140285071268,
      "grad_norm": 0.19073475897312164,
      "learning_rate": 2.0892195176183354e-05,
      "loss": 0.0393,
      "step": 1941
    },
    {
      "epoch": 1.4568642160540135,
      "grad_norm": 0.1695142388343811,
      "learning_rate": 2.083898219934739e-05,
      "loss": 0.0387,
      "step": 1942
    },
    {
      "epoch": 1.4576144036009002,
      "grad_norm": 0.1988031268119812,
      "learning_rate": 2.0785819230785398e-05,
      "loss": 0.0461,
      "step": 1943
    },
    {
      "epoch": 1.458364591147787,
      "grad_norm": 0.14169101417064667,
      "learning_rate": 2.073270636166666e-05,
      "loss": 0.0552,
      "step": 1944
    },
    {
      "epoch": 1.4591147786946737,
      "grad_norm": 0.1895540952682495,
      "learning_rate": 2.0679643683074513e-05,
      "loss": 0.0404,
      "step": 1945
    },
    {
      "epoch": 1.4598649662415604,
      "grad_norm": 0.15737725794315338,
      "learning_rate": 2.0626631286006236e-05,
      "loss": 0.0355,
      "step": 1946
    },
    {
      "epoch": 1.460615153788447,
      "grad_norm": 0.2160894274711609,
      "learning_rate": 2.0573669261372847e-05,
      "loss": 0.041,
      "step": 1947
    },
    {
      "epoch": 1.4613653413353338,
      "grad_norm": 0.17199784517288208,
      "learning_rate": 2.052075769999899e-05,
      "loss": 0.0443,
      "step": 1948
    },
    {
      "epoch": 1.4621155288822205,
      "grad_norm": 0.1754937618970871,
      "learning_rate": 2.046789669262283e-05,
      "loss": 0.0391,
      "step": 1949
    },
    {
      "epoch": 1.4628657164291072,
      "grad_norm": 0.15710797905921936,
      "learning_rate": 2.0415086329895784e-05,
      "loss": 0.0345,
      "step": 1950
    },
    {
      "epoch": 1.463615903975994,
      "grad_norm": 0.15425771474838257,
      "learning_rate": 2.0362326702382384e-05,
      "loss": 0.0294,
      "step": 1951
    },
    {
      "epoch": 1.4643660915228807,
      "grad_norm": 0.15144193172454834,
      "learning_rate": 2.0309617900560218e-05,
      "loss": 0.0364,
      "step": 1952
    },
    {
      "epoch": 1.4651162790697674,
      "grad_norm": 0.20072755217552185,
      "learning_rate": 2.0256960014819692e-05,
      "loss": 0.05,
      "step": 1953
    },
    {
      "epoch": 1.465866466616654,
      "grad_norm": 0.17849725484848022,
      "learning_rate": 2.020435313546391e-05,
      "loss": 0.0433,
      "step": 1954
    },
    {
      "epoch": 1.4666166541635408,
      "grad_norm": 0.30021312832832336,
      "learning_rate": 2.0151797352708457e-05,
      "loss": 0.0862,
      "step": 1955
    },
    {
      "epoch": 1.4673668417104275,
      "grad_norm": 0.211557537317276,
      "learning_rate": 2.0099292756681343e-05,
      "loss": 0.0625,
      "step": 1956
    },
    {
      "epoch": 1.4681170292573142,
      "grad_norm": 0.18450140953063965,
      "learning_rate": 2.0046839437422772e-05,
      "loss": 0.0407,
      "step": 1957
    },
    {
      "epoch": 1.4688672168042012,
      "grad_norm": 0.19659367203712463,
      "learning_rate": 1.999443748488503e-05,
      "loss": 0.0356,
      "step": 1958
    },
    {
      "epoch": 1.4696174043510877,
      "grad_norm": 0.18041273951530457,
      "learning_rate": 1.9942086988932323e-05,
      "loss": 0.0465,
      "step": 1959
    },
    {
      "epoch": 1.4703675918979746,
      "grad_norm": 0.13432402908802032,
      "learning_rate": 1.9889788039340558e-05,
      "loss": 0.0424,
      "step": 1960
    },
    {
      "epoch": 1.471117779444861,
      "grad_norm": 0.16820655763149261,
      "learning_rate": 1.9837540725797305e-05,
      "loss": 0.0405,
      "step": 1961
    },
    {
      "epoch": 1.471867966991748,
      "grad_norm": 0.18088965117931366,
      "learning_rate": 1.9785345137901533e-05,
      "loss": 0.0462,
      "step": 1962
    },
    {
      "epoch": 1.4726181545386345,
      "grad_norm": 0.14395685493946075,
      "learning_rate": 1.9733201365163607e-05,
      "loss": 0.0365,
      "step": 1963
    },
    {
      "epoch": 1.4733683420855215,
      "grad_norm": 0.2003966122865677,
      "learning_rate": 1.968110949700489e-05,
      "loss": 0.0589,
      "step": 1964
    },
    {
      "epoch": 1.474118529632408,
      "grad_norm": 0.15150992572307587,
      "learning_rate": 1.962906962275784e-05,
      "loss": 0.0371,
      "step": 1965
    },
    {
      "epoch": 1.474868717179295,
      "grad_norm": 0.16001686453819275,
      "learning_rate": 1.9577081831665707e-05,
      "loss": 0.0507,
      "step": 1966
    },
    {
      "epoch": 1.4756189047261816,
      "grad_norm": 0.1944255381822586,
      "learning_rate": 1.9525146212882456e-05,
      "loss": 0.0621,
      "step": 1967
    },
    {
      "epoch": 1.4763690922730683,
      "grad_norm": 0.15253233909606934,
      "learning_rate": 1.9473262855472517e-05,
      "loss": 0.0478,
      "step": 1968
    },
    {
      "epoch": 1.477119279819955,
      "grad_norm": 0.1423439085483551,
      "learning_rate": 1.942143184841077e-05,
      "loss": 0.038,
      "step": 1969
    },
    {
      "epoch": 1.4778694673668418,
      "grad_norm": 0.150668203830719,
      "learning_rate": 1.9369653280582273e-05,
      "loss": 0.0412,
      "step": 1970
    },
    {
      "epoch": 1.4786196549137285,
      "grad_norm": 0.25101804733276367,
      "learning_rate": 1.931792724078218e-05,
      "loss": 0.0646,
      "step": 1971
    },
    {
      "epoch": 1.4793698424606152,
      "grad_norm": 0.17958179116249084,
      "learning_rate": 1.9266253817715575e-05,
      "loss": 0.0383,
      "step": 1972
    },
    {
      "epoch": 1.480120030007502,
      "grad_norm": 0.20443418622016907,
      "learning_rate": 1.921463309999724e-05,
      "loss": 0.0373,
      "step": 1973
    },
    {
      "epoch": 1.4808702175543886,
      "grad_norm": 0.1706647425889969,
      "learning_rate": 1.9163065176151662e-05,
      "loss": 0.0426,
      "step": 1974
    },
    {
      "epoch": 1.4816204051012754,
      "grad_norm": 0.17196975648403168,
      "learning_rate": 1.9111550134612738e-05,
      "loss": 0.0405,
      "step": 1975
    },
    {
      "epoch": 1.482370592648162,
      "grad_norm": 0.2115897536277771,
      "learning_rate": 1.9060088063723696e-05,
      "loss": 0.0457,
      "step": 1976
    },
    {
      "epoch": 1.4831207801950488,
      "grad_norm": 0.1429477035999298,
      "learning_rate": 1.900867905173692e-05,
      "loss": 0.0374,
      "step": 1977
    },
    {
      "epoch": 1.4838709677419355,
      "grad_norm": 0.1279119849205017,
      "learning_rate": 1.8957323186813803e-05,
      "loss": 0.0285,
      "step": 1978
    },
    {
      "epoch": 1.4846211552888222,
      "grad_norm": 0.13002172112464905,
      "learning_rate": 1.8906020557024597e-05,
      "loss": 0.0288,
      "step": 1979
    },
    {
      "epoch": 1.485371342835709,
      "grad_norm": 0.19170233607292175,
      "learning_rate": 1.885477125034827e-05,
      "loss": 0.0396,
      "step": 1980
    },
    {
      "epoch": 1.4861215303825956,
      "grad_norm": 0.278576135635376,
      "learning_rate": 1.8803575354672315e-05,
      "loss": 0.0544,
      "step": 1981
    },
    {
      "epoch": 1.4868717179294824,
      "grad_norm": 0.1751520186662674,
      "learning_rate": 1.8752432957792654e-05,
      "loss": 0.0458,
      "step": 1982
    },
    {
      "epoch": 1.487621905476369,
      "grad_norm": 0.27717238664627075,
      "learning_rate": 1.8701344147413474e-05,
      "loss": 0.107,
      "step": 1983
    },
    {
      "epoch": 1.4883720930232558,
      "grad_norm": 0.14058585464954376,
      "learning_rate": 1.8650309011147053e-05,
      "loss": 0.0354,
      "step": 1984
    },
    {
      "epoch": 1.4891222805701425,
      "grad_norm": 0.1992284208536148,
      "learning_rate": 1.8599327636513636e-05,
      "loss": 0.037,
      "step": 1985
    },
    {
      "epoch": 1.4898724681170292,
      "grad_norm": 0.1688821017742157,
      "learning_rate": 1.8548400110941228e-05,
      "loss": 0.0317,
      "step": 1986
    },
    {
      "epoch": 1.490622655663916,
      "grad_norm": 0.21011146903038025,
      "learning_rate": 1.8497526521765534e-05,
      "loss": 0.0434,
      "step": 1987
    },
    {
      "epoch": 1.4913728432108027,
      "grad_norm": 0.13675890862941742,
      "learning_rate": 1.844670695622976e-05,
      "loss": 0.0268,
      "step": 1988
    },
    {
      "epoch": 1.4921230307576894,
      "grad_norm": 0.15465782582759857,
      "learning_rate": 1.8395941501484464e-05,
      "loss": 0.0332,
      "step": 1989
    },
    {
      "epoch": 1.492873218304576,
      "grad_norm": 0.1339779645204544,
      "learning_rate": 1.8345230244587354e-05,
      "loss": 0.0336,
      "step": 1990
    },
    {
      "epoch": 1.4936234058514628,
      "grad_norm": 0.16057631373405457,
      "learning_rate": 1.829457327250329e-05,
      "loss": 0.0392,
      "step": 1991
    },
    {
      "epoch": 1.4943735933983495,
      "grad_norm": 0.240296870470047,
      "learning_rate": 1.8243970672103982e-05,
      "loss": 0.0459,
      "step": 1992
    },
    {
      "epoch": 1.4951237809452362,
      "grad_norm": 0.2193770408630371,
      "learning_rate": 1.8193422530167914e-05,
      "loss": 0.0572,
      "step": 1993
    },
    {
      "epoch": 1.495873968492123,
      "grad_norm": 0.16321928799152374,
      "learning_rate": 1.8142928933380142e-05,
      "loss": 0.0391,
      "step": 1994
    },
    {
      "epoch": 1.49662415603901,
      "grad_norm": 0.14024445414543152,
      "learning_rate": 1.8092489968332233e-05,
      "loss": 0.0336,
      "step": 1995
    },
    {
      "epoch": 1.4973743435858964,
      "grad_norm": 0.15499421954154968,
      "learning_rate": 1.804210572152204e-05,
      "loss": 0.0409,
      "step": 1996
    },
    {
      "epoch": 1.4981245311327833,
      "grad_norm": 0.1850685328245163,
      "learning_rate": 1.7991776279353604e-05,
      "loss": 0.0511,
      "step": 1997
    },
    {
      "epoch": 1.4988747186796698,
      "grad_norm": 0.17802977561950684,
      "learning_rate": 1.794150172813693e-05,
      "loss": 0.0332,
      "step": 1998
    },
    {
      "epoch": 1.4996249062265568,
      "grad_norm": 0.2455652356147766,
      "learning_rate": 1.7891282154087934e-05,
      "loss": 0.0458,
      "step": 1999
    },
    {
      "epoch": 1.5003750937734432,
      "grad_norm": 0.16653724014759064,
      "learning_rate": 1.7841117643328246e-05,
      "loss": 0.045,
      "step": 2000
    },
    {
      "epoch": 1.5003750937734432,
      "eval_loss": 0.06891622394323349,
      "eval_runtime": 2.6405,
      "eval_samples_per_second": 20.451,
      "eval_steps_per_second": 5.302,
      "step": 2000
    },
    {
      "epoch": 1.5011252813203302,
      "grad_norm": 0.30463671684265137,
      "learning_rate": 1.779100828188506e-05,
      "loss": 0.0654,
      "step": 2001
    },
    {
      "epoch": 1.5018754688672167,
      "grad_norm": 0.18647217750549316,
      "learning_rate": 1.774095415569102e-05,
      "loss": 0.0356,
      "step": 2002
    },
    {
      "epoch": 1.5026256564141036,
      "grad_norm": 0.17118555307388306,
      "learning_rate": 1.7690955350583976e-05,
      "loss": 0.037,
      "step": 2003
    },
    {
      "epoch": 1.50337584396099,
      "grad_norm": 0.26206356287002563,
      "learning_rate": 1.764101195230696e-05,
      "loss": 0.0774,
      "step": 2004
    },
    {
      "epoch": 1.504126031507877,
      "grad_norm": 0.5694636702537537,
      "learning_rate": 1.7591124046508045e-05,
      "loss": 0.1328,
      "step": 2005
    },
    {
      "epoch": 1.5048762190547635,
      "grad_norm": 0.16807466745376587,
      "learning_rate": 1.7541291718740012e-05,
      "loss": 0.0425,
      "step": 2006
    },
    {
      "epoch": 1.5056264066016505,
      "grad_norm": 0.1880795806646347,
      "learning_rate": 1.7491515054460418e-05,
      "loss": 0.0416,
      "step": 2007
    },
    {
      "epoch": 1.506376594148537,
      "grad_norm": 0.15772967040538788,
      "learning_rate": 1.7441794139031337e-05,
      "loss": 0.0561,
      "step": 2008
    },
    {
      "epoch": 1.507126781695424,
      "grad_norm": 0.26514455676078796,
      "learning_rate": 1.7392129057719246e-05,
      "loss": 0.0603,
      "step": 2009
    },
    {
      "epoch": 1.5078769692423106,
      "grad_norm": 0.17776095867156982,
      "learning_rate": 1.7342519895694886e-05,
      "loss": 0.0555,
      "step": 2010
    },
    {
      "epoch": 1.5086271567891973,
      "grad_norm": 0.24651385843753815,
      "learning_rate": 1.7292966738033057e-05,
      "loss": 0.0476,
      "step": 2011
    },
    {
      "epoch": 1.509377344336084,
      "grad_norm": 0.19671979546546936,
      "learning_rate": 1.7243469669712546e-05,
      "loss": 0.0439,
      "step": 2012
    },
    {
      "epoch": 1.5101275318829708,
      "grad_norm": 0.180504709482193,
      "learning_rate": 1.7194028775615966e-05,
      "loss": 0.0494,
      "step": 2013
    },
    {
      "epoch": 1.5108777194298575,
      "grad_norm": 0.20517569780349731,
      "learning_rate": 1.714464414052958e-05,
      "loss": 0.0524,
      "step": 2014
    },
    {
      "epoch": 1.5116279069767442,
      "grad_norm": 0.1717517226934433,
      "learning_rate": 1.7095315849143184e-05,
      "loss": 0.0551,
      "step": 2015
    },
    {
      "epoch": 1.512378094523631,
      "grad_norm": 0.1254064440727234,
      "learning_rate": 1.704604398604991e-05,
      "loss": 0.0306,
      "step": 2016
    },
    {
      "epoch": 1.5131282820705176,
      "grad_norm": 0.19165663421154022,
      "learning_rate": 1.6996828635746165e-05,
      "loss": 0.0387,
      "step": 2017
    },
    {
      "epoch": 1.5138784696174044,
      "grad_norm": 0.16394397616386414,
      "learning_rate": 1.6947669882631434e-05,
      "loss": 0.0382,
      "step": 2018
    },
    {
      "epoch": 1.514628657164291,
      "grad_norm": 0.17494425177574158,
      "learning_rate": 1.6898567811008135e-05,
      "loss": 0.0379,
      "step": 2019
    },
    {
      "epoch": 1.5153788447111778,
      "grad_norm": 0.15336370468139648,
      "learning_rate": 1.684952250508149e-05,
      "loss": 0.0355,
      "step": 2020
    },
    {
      "epoch": 1.5161290322580645,
      "grad_norm": 0.12618416547775269,
      "learning_rate": 1.6800534048959364e-05,
      "loss": 0.026,
      "step": 2021
    },
    {
      "epoch": 1.5168792198049512,
      "grad_norm": 0.15845142304897308,
      "learning_rate": 1.6751602526652133e-05,
      "loss": 0.0362,
      "step": 2022
    },
    {
      "epoch": 1.517629407351838,
      "grad_norm": 0.2082812488079071,
      "learning_rate": 1.6702728022072562e-05,
      "loss": 0.0461,
      "step": 2023
    },
    {
      "epoch": 1.5183795948987246,
      "grad_norm": 0.12379974126815796,
      "learning_rate": 1.665391061903558e-05,
      "loss": 0.0312,
      "step": 2024
    },
    {
      "epoch": 1.5191297824456114,
      "grad_norm": 0.21030539274215698,
      "learning_rate": 1.660515040125824e-05,
      "loss": 0.059,
      "step": 2025
    },
    {
      "epoch": 1.519879969992498,
      "grad_norm": 0.13408705592155457,
      "learning_rate": 1.6556447452359512e-05,
      "loss": 0.0384,
      "step": 2026
    },
    {
      "epoch": 1.5206301575393848,
      "grad_norm": 0.25092118978500366,
      "learning_rate": 1.6507801855860177e-05,
      "loss": 0.0496,
      "step": 2027
    },
    {
      "epoch": 1.5213803450862715,
      "grad_norm": 0.15542156994342804,
      "learning_rate": 1.645921369518261e-05,
      "loss": 0.0322,
      "step": 2028
    },
    {
      "epoch": 1.5221305326331582,
      "grad_norm": 0.18500564992427826,
      "learning_rate": 1.6410683053650737e-05,
      "loss": 0.0442,
      "step": 2029
    },
    {
      "epoch": 1.5228807201800452,
      "grad_norm": 0.21039170026779175,
      "learning_rate": 1.636221001448983e-05,
      "loss": 0.042,
      "step": 2030
    },
    {
      "epoch": 1.5236309077269317,
      "grad_norm": 0.21939733624458313,
      "learning_rate": 1.631379466082638e-05,
      "loss": 0.0555,
      "step": 2031
    },
    {
      "epoch": 1.5243810952738186,
      "grad_norm": 0.19629713892936707,
      "learning_rate": 1.626543707568795e-05,
      "loss": 0.0428,
      "step": 2032
    },
    {
      "epoch": 1.525131282820705,
      "grad_norm": 0.31679531931877136,
      "learning_rate": 1.6217137342003036e-05,
      "loss": 0.0398,
      "step": 2033
    },
    {
      "epoch": 1.525881470367592,
      "grad_norm": 0.23860694468021393,
      "learning_rate": 1.616889554260092e-05,
      "loss": 0.0612,
      "step": 2034
    },
    {
      "epoch": 1.5266316579144785,
      "grad_norm": 0.16307824850082397,
      "learning_rate": 1.6120711760211548e-05,
      "loss": 0.0429,
      "step": 2035
    },
    {
      "epoch": 1.5273818454613655,
      "grad_norm": 0.13818341493606567,
      "learning_rate": 1.607258607746537e-05,
      "loss": 0.0341,
      "step": 2036
    },
    {
      "epoch": 1.528132033008252,
      "grad_norm": 0.16487646102905273,
      "learning_rate": 1.602451857689316e-05,
      "loss": 0.0391,
      "step": 2037
    },
    {
      "epoch": 1.528882220555139,
      "grad_norm": 0.2172565758228302,
      "learning_rate": 1.5976509340925977e-05,
      "loss": 0.0566,
      "step": 2038
    },
    {
      "epoch": 1.5296324081020254,
      "grad_norm": 0.1681651920080185,
      "learning_rate": 1.5928558451894914e-05,
      "loss": 0.0375,
      "step": 2039
    },
    {
      "epoch": 1.5303825956489123,
      "grad_norm": 0.23445692658424377,
      "learning_rate": 1.588066599203106e-05,
      "loss": 0.0397,
      "step": 2040
    },
    {
      "epoch": 1.5311327831957988,
      "grad_norm": 0.1395978331565857,
      "learning_rate": 1.583283204346521e-05,
      "loss": 0.0345,
      "step": 2041
    },
    {
      "epoch": 1.5318829707426858,
      "grad_norm": 0.20669519901275635,
      "learning_rate": 1.5785056688227916e-05,
      "loss": 0.033,
      "step": 2042
    },
    {
      "epoch": 1.5326331582895723,
      "grad_norm": 0.19717498123645782,
      "learning_rate": 1.5737340008249202e-05,
      "loss": 0.0428,
      "step": 2043
    },
    {
      "epoch": 1.5333833458364592,
      "grad_norm": 0.15539968013763428,
      "learning_rate": 1.5689682085358465e-05,
      "loss": 0.0383,
      "step": 2044
    },
    {
      "epoch": 1.5341335333833457,
      "grad_norm": 0.1587790548801422,
      "learning_rate": 1.564208300128438e-05,
      "loss": 0.0326,
      "step": 2045
    },
    {
      "epoch": 1.5348837209302326,
      "grad_norm": 0.172133207321167,
      "learning_rate": 1.5594542837654625e-05,
      "loss": 0.0285,
      "step": 2046
    },
    {
      "epoch": 1.5356339084771191,
      "grad_norm": 0.17575670778751373,
      "learning_rate": 1.554706167599596e-05,
      "loss": 0.0442,
      "step": 2047
    },
    {
      "epoch": 1.536384096024006,
      "grad_norm": 0.20420925319194794,
      "learning_rate": 1.5499639597733902e-05,
      "loss": 0.0458,
      "step": 2048
    },
    {
      "epoch": 1.5371342835708928,
      "grad_norm": 0.20247098803520203,
      "learning_rate": 1.54522766841926e-05,
      "loss": 0.0538,
      "step": 2049
    },
    {
      "epoch": 1.5378844711177795,
      "grad_norm": 0.21190738677978516,
      "learning_rate": 1.540497301659482e-05,
      "loss": 0.0468,
      "step": 2050
    },
    {
      "epoch": 1.5386346586646662,
      "grad_norm": 0.170917809009552,
      "learning_rate": 1.5357728676061685e-05,
      "loss": 0.0456,
      "step": 2051
    },
    {
      "epoch": 1.539384846211553,
      "grad_norm": 0.30017176270484924,
      "learning_rate": 1.5310543743612582e-05,
      "loss": 0.0822,
      "step": 2052
    },
    {
      "epoch": 1.5401350337584396,
      "grad_norm": 0.17813549935817719,
      "learning_rate": 1.526341830016505e-05,
      "loss": 0.0382,
      "step": 2053
    },
    {
      "epoch": 1.5408852213053263,
      "grad_norm": 0.2268344908952713,
      "learning_rate": 1.5216352426534548e-05,
      "loss": 0.0514,
      "step": 2054
    },
    {
      "epoch": 1.541635408852213,
      "grad_norm": 0.19404159486293793,
      "learning_rate": 1.5169346203434425e-05,
      "loss": 0.0608,
      "step": 2055
    },
    {
      "epoch": 1.5423855963990998,
      "grad_norm": 0.21496349573135376,
      "learning_rate": 1.5122399711475732e-05,
      "loss": 0.0615,
      "step": 2056
    },
    {
      "epoch": 1.5431357839459865,
      "grad_norm": 0.1785394698381424,
      "learning_rate": 1.50755130311671e-05,
      "loss": 0.0438,
      "step": 2057
    },
    {
      "epoch": 1.5438859714928732,
      "grad_norm": 0.21393193304538727,
      "learning_rate": 1.502868624291452e-05,
      "loss": 0.0478,
      "step": 2058
    },
    {
      "epoch": 1.54463615903976,
      "grad_norm": 0.19037163257598877,
      "learning_rate": 1.4981919427021357e-05,
      "loss": 0.0493,
      "step": 2059
    },
    {
      "epoch": 1.5453863465866466,
      "grad_norm": 0.15680454671382904,
      "learning_rate": 1.493521266368807e-05,
      "loss": 0.0307,
      "step": 2060
    },
    {
      "epoch": 1.5461365341335334,
      "grad_norm": 0.19426371157169342,
      "learning_rate": 1.4888566033012201e-05,
      "loss": 0.0612,
      "step": 2061
    },
    {
      "epoch": 1.54688672168042,
      "grad_norm": 0.14693467319011688,
      "learning_rate": 1.4841979614988094e-05,
      "loss": 0.0273,
      "step": 2062
    },
    {
      "epoch": 1.5476369092273068,
      "grad_norm": 0.18175330758094788,
      "learning_rate": 1.4795453489506878e-05,
      "loss": 0.0683,
      "step": 2063
    },
    {
      "epoch": 1.5483870967741935,
      "grad_norm": 0.12993456423282623,
      "learning_rate": 1.4748987736356273e-05,
      "loss": 0.0367,
      "step": 2064
    },
    {
      "epoch": 1.5491372843210802,
      "grad_norm": 0.17994779348373413,
      "learning_rate": 1.4702582435220475e-05,
      "loss": 0.0612,
      "step": 2065
    },
    {
      "epoch": 1.549887471867967,
      "grad_norm": 0.20236118137836456,
      "learning_rate": 1.4656237665680017e-05,
      "loss": 0.0523,
      "step": 2066
    },
    {
      "epoch": 1.5506376594148539,
      "grad_norm": 0.11795151978731155,
      "learning_rate": 1.4609953507211593e-05,
      "loss": 0.0331,
      "step": 2067
    },
    {
      "epoch": 1.5513878469617404,
      "grad_norm": 0.1285896897315979,
      "learning_rate": 1.4563730039187984e-05,
      "loss": 0.0342,
      "step": 2068
    },
    {
      "epoch": 1.5521380345086273,
      "grad_norm": 0.11817178130149841,
      "learning_rate": 1.4517567340877886e-05,
      "loss": 0.0279,
      "step": 2069
    },
    {
      "epoch": 1.5528882220555138,
      "grad_norm": 0.15962854027748108,
      "learning_rate": 1.4471465491445802e-05,
      "loss": 0.035,
      "step": 2070
    },
    {
      "epoch": 1.5536384096024007,
      "grad_norm": 0.19675575196743011,
      "learning_rate": 1.4425424569951822e-05,
      "loss": 0.0373,
      "step": 2071
    },
    {
      "epoch": 1.5543885971492872,
      "grad_norm": 0.23658345639705658,
      "learning_rate": 1.4379444655351626e-05,
      "loss": 0.0793,
      "step": 2072
    },
    {
      "epoch": 1.5551387846961742,
      "grad_norm": 0.15732000768184662,
      "learning_rate": 1.4333525826496224e-05,
      "loss": 0.0233,
      "step": 2073
    },
    {
      "epoch": 1.5558889722430607,
      "grad_norm": 0.21282753348350525,
      "learning_rate": 1.4287668162131896e-05,
      "loss": 0.0594,
      "step": 2074
    },
    {
      "epoch": 1.5566391597899476,
      "grad_norm": 0.19476816058158875,
      "learning_rate": 1.4241871740900014e-05,
      "loss": 0.045,
      "step": 2075
    },
    {
      "epoch": 1.557389347336834,
      "grad_norm": 0.1995515525341034,
      "learning_rate": 1.4196136641336932e-05,
      "loss": 0.0501,
      "step": 2076
    },
    {
      "epoch": 1.558139534883721,
      "grad_norm": 0.18209566175937653,
      "learning_rate": 1.4150462941873843e-05,
      "loss": 0.0435,
      "step": 2077
    },
    {
      "epoch": 1.5588897224306075,
      "grad_norm": 0.16033656895160675,
      "learning_rate": 1.410485072083666e-05,
      "loss": 0.0378,
      "step": 2078
    },
    {
      "epoch": 1.5596399099774945,
      "grad_norm": 0.1421896368265152,
      "learning_rate": 1.4059300056445823e-05,
      "loss": 0.0489,
      "step": 2079
    },
    {
      "epoch": 1.560390097524381,
      "grad_norm": 0.16776643693447113,
      "learning_rate": 1.4013811026816243e-05,
      "loss": 0.0466,
      "step": 2080
    },
    {
      "epoch": 1.561140285071268,
      "grad_norm": 0.20549526810646057,
      "learning_rate": 1.3968383709957133e-05,
      "loss": 0.032,
      "step": 2081
    },
    {
      "epoch": 1.5618904726181544,
      "grad_norm": 0.17907895147800446,
      "learning_rate": 1.3923018183771868e-05,
      "loss": 0.0481,
      "step": 2082
    },
    {
      "epoch": 1.5626406601650413,
      "grad_norm": 0.2195027768611908,
      "learning_rate": 1.3877714526057872e-05,
      "loss": 0.0551,
      "step": 2083
    },
    {
      "epoch": 1.5633908477119278,
      "grad_norm": 0.18014267086982727,
      "learning_rate": 1.3832472814506425e-05,
      "loss": 0.0357,
      "step": 2084
    },
    {
      "epoch": 1.5641410352588148,
      "grad_norm": 0.24690528213977814,
      "learning_rate": 1.3787293126702622e-05,
      "loss": 0.0514,
      "step": 2085
    },
    {
      "epoch": 1.5648912228057015,
      "grad_norm": 0.2089156061410904,
      "learning_rate": 1.3742175540125179e-05,
      "loss": 0.0528,
      "step": 2086
    },
    {
      "epoch": 1.5656414103525882,
      "grad_norm": 0.17088919878005981,
      "learning_rate": 1.3697120132146318e-05,
      "loss": 0.0387,
      "step": 2087
    },
    {
      "epoch": 1.566391597899475,
      "grad_norm": 0.13750265538692474,
      "learning_rate": 1.3652126980031627e-05,
      "loss": 0.0311,
      "step": 2088
    },
    {
      "epoch": 1.5671417854463616,
      "grad_norm": 0.24524101614952087,
      "learning_rate": 1.3607196160939927e-05,
      "loss": 0.0619,
      "step": 2089
    },
    {
      "epoch": 1.5678919729932483,
      "grad_norm": 0.14645838737487793,
      "learning_rate": 1.3562327751923149e-05,
      "loss": 0.0429,
      "step": 2090
    },
    {
      "epoch": 1.568642160540135,
      "grad_norm": 0.21580055356025696,
      "learning_rate": 1.351752182992621e-05,
      "loss": 0.0603,
      "step": 2091
    },
    {
      "epoch": 1.5693923480870218,
      "grad_norm": 0.14457224309444427,
      "learning_rate": 1.3472778471786829e-05,
      "loss": 0.0258,
      "step": 2092
    },
    {
      "epoch": 1.5701425356339085,
      "grad_norm": 0.1384057253599167,
      "learning_rate": 1.3428097754235475e-05,
      "loss": 0.0366,
      "step": 2093
    },
    {
      "epoch": 1.5708927231807952,
      "grad_norm": 0.1908542960882187,
      "learning_rate": 1.3383479753895174e-05,
      "loss": 0.0517,
      "step": 2094
    },
    {
      "epoch": 1.571642910727682,
      "grad_norm": 0.21990883350372314,
      "learning_rate": 1.33389245472814e-05,
      "loss": 0.0398,
      "step": 2095
    },
    {
      "epoch": 1.5723930982745686,
      "grad_norm": 0.17277023196220398,
      "learning_rate": 1.3294432210801966e-05,
      "loss": 0.0362,
      "step": 2096
    },
    {
      "epoch": 1.5731432858214554,
      "grad_norm": 0.17331083118915558,
      "learning_rate": 1.3250002820756819e-05,
      "loss": 0.0446,
      "step": 2097
    },
    {
      "epoch": 1.573893473368342,
      "grad_norm": 0.14335882663726807,
      "learning_rate": 1.3205636453338e-05,
      "loss": 0.0268,
      "step": 2098
    },
    {
      "epoch": 1.5746436609152288,
      "grad_norm": 0.1492685228586197,
      "learning_rate": 1.316133318462946e-05,
      "loss": 0.038,
      "step": 2099
    },
    {
      "epoch": 1.5753938484621155,
      "grad_norm": 0.20548713207244873,
      "learning_rate": 1.3117093090606958e-05,
      "loss": 0.047,
      "step": 2100
    },
    {
      "epoch": 1.5761440360090022,
      "grad_norm": 0.16951590776443481,
      "learning_rate": 1.3072916247137861e-05,
      "loss": 0.0266,
      "step": 2101
    },
    {
      "epoch": 1.576894223555889,
      "grad_norm": 0.1733119636774063,
      "learning_rate": 1.302880272998112e-05,
      "loss": 0.0343,
      "step": 2102
    },
    {
      "epoch": 1.5776444111027756,
      "grad_norm": 0.1762966811656952,
      "learning_rate": 1.29847526147871e-05,
      "loss": 0.0336,
      "step": 2103
    },
    {
      "epoch": 1.5783945986496624,
      "grad_norm": 0.21547256410121918,
      "learning_rate": 1.2940765977097402e-05,
      "loss": 0.0429,
      "step": 2104
    },
    {
      "epoch": 1.579144786196549,
      "grad_norm": 0.14866936206817627,
      "learning_rate": 1.2896842892344751e-05,
      "loss": 0.0278,
      "step": 2105
    },
    {
      "epoch": 1.579894973743436,
      "grad_norm": 0.19238527119159698,
      "learning_rate": 1.2852983435852928e-05,
      "loss": 0.0467,
      "step": 2106
    },
    {
      "epoch": 1.5806451612903225,
      "grad_norm": 0.1957983523607254,
      "learning_rate": 1.2809187682836588e-05,
      "loss": 0.0352,
      "step": 2107
    },
    {
      "epoch": 1.5813953488372094,
      "grad_norm": 0.16528469324111938,
      "learning_rate": 1.2765455708401142e-05,
      "loss": 0.0461,
      "step": 2108
    },
    {
      "epoch": 1.582145536384096,
      "grad_norm": 0.23004306852817535,
      "learning_rate": 1.2721787587542595e-05,
      "loss": 0.0634,
      "step": 2109
    },
    {
      "epoch": 1.5828957239309829,
      "grad_norm": 0.1990731805562973,
      "learning_rate": 1.2678183395147485e-05,
      "loss": 0.0564,
      "step": 2110
    },
    {
      "epoch": 1.5836459114778694,
      "grad_norm": 0.188127338886261,
      "learning_rate": 1.2634643205992707e-05,
      "loss": 0.0334,
      "step": 2111
    },
    {
      "epoch": 1.5843960990247563,
      "grad_norm": 0.2097140997648239,
      "learning_rate": 1.2591167094745404e-05,
      "loss": 0.0612,
      "step": 2112
    },
    {
      "epoch": 1.5851462865716428,
      "grad_norm": 0.12479346990585327,
      "learning_rate": 1.2547755135962841e-05,
      "loss": 0.0348,
      "step": 2113
    },
    {
      "epoch": 1.5858964741185297,
      "grad_norm": 0.17138241231441498,
      "learning_rate": 1.2504407404092217e-05,
      "loss": 0.0466,
      "step": 2114
    },
    {
      "epoch": 1.5866466616654162,
      "grad_norm": 0.17071551084518433,
      "learning_rate": 1.2461123973470634e-05,
      "loss": 0.039,
      "step": 2115
    },
    {
      "epoch": 1.5873968492123032,
      "grad_norm": 0.21277554333209991,
      "learning_rate": 1.2417904918324913e-05,
      "loss": 0.0347,
      "step": 2116
    },
    {
      "epoch": 1.5881470367591897,
      "grad_norm": 0.18354669213294983,
      "learning_rate": 1.237475031277151e-05,
      "loss": 0.0344,
      "step": 2117
    },
    {
      "epoch": 1.5888972243060766,
      "grad_norm": 0.2000560760498047,
      "learning_rate": 1.2331660230816288e-05,
      "loss": 0.0502,
      "step": 2118
    },
    {
      "epoch": 1.589647411852963,
      "grad_norm": 0.18581894040107727,
      "learning_rate": 1.2288634746354505e-05,
      "loss": 0.0331,
      "step": 2119
    },
    {
      "epoch": 1.59039759939985,
      "grad_norm": 0.16207176446914673,
      "learning_rate": 1.2245673933170626e-05,
      "loss": 0.0321,
      "step": 2120
    },
    {
      "epoch": 1.5911477869467365,
      "grad_norm": 0.11233672499656677,
      "learning_rate": 1.2202777864938236e-05,
      "loss": 0.0323,
      "step": 2121
    },
    {
      "epoch": 1.5918979744936235,
      "grad_norm": 0.1832706332206726,
      "learning_rate": 1.2159946615219836e-05,
      "loss": 0.028,
      "step": 2122
    },
    {
      "epoch": 1.59264816204051,
      "grad_norm": 0.20577500760555267,
      "learning_rate": 1.211718025746682e-05,
      "loss": 0.0278,
      "step": 2123
    },
    {
      "epoch": 1.593398349587397,
      "grad_norm": 0.22615359723567963,
      "learning_rate": 1.2074478865019273e-05,
      "loss": 0.0455,
      "step": 2124
    },
    {
      "epoch": 1.5941485371342836,
      "grad_norm": 0.15508152544498444,
      "learning_rate": 1.2031842511105885e-05,
      "loss": 0.039,
      "step": 2125
    },
    {
      "epoch": 1.5948987246811703,
      "grad_norm": 0.1680385321378708,
      "learning_rate": 1.1989271268843815e-05,
      "loss": 0.0363,
      "step": 2126
    },
    {
      "epoch": 1.595648912228057,
      "grad_norm": 0.21748638153076172,
      "learning_rate": 1.1946765211238526e-05,
      "loss": 0.05,
      "step": 2127
    },
    {
      "epoch": 1.5963990997749438,
      "grad_norm": 0.2185010313987732,
      "learning_rate": 1.1904324411183731e-05,
      "loss": 0.0454,
      "step": 2128
    },
    {
      "epoch": 1.5971492873218305,
      "grad_norm": 0.2044178694486618,
      "learning_rate": 1.1861948941461226e-05,
      "loss": 0.0399,
      "step": 2129
    },
    {
      "epoch": 1.5978994748687172,
      "grad_norm": 0.27306026220321655,
      "learning_rate": 1.1819638874740769e-05,
      "loss": 0.0459,
      "step": 2130
    },
    {
      "epoch": 1.598649662415604,
      "grad_norm": 0.17152051627635956,
      "learning_rate": 1.1777394283579956e-05,
      "loss": 0.031,
      "step": 2131
    },
    {
      "epoch": 1.5993998499624906,
      "grad_norm": 0.1820501983165741,
      "learning_rate": 1.1735215240424102e-05,
      "loss": 0.0395,
      "step": 2132
    },
    {
      "epoch": 1.6001500375093773,
      "grad_norm": 0.1924869865179062,
      "learning_rate": 1.1693101817606117e-05,
      "loss": 0.064,
      "step": 2133
    },
    {
      "epoch": 1.600900225056264,
      "grad_norm": 0.1441008746623993,
      "learning_rate": 1.165105408734638e-05,
      "loss": 0.0269,
      "step": 2134
    },
    {
      "epoch": 1.6016504126031508,
      "grad_norm": 0.16650429368019104,
      "learning_rate": 1.1609072121752584e-05,
      "loss": 0.0536,
      "step": 2135
    },
    {
      "epoch": 1.6024006001500375,
      "grad_norm": 0.16469474136829376,
      "learning_rate": 1.1567155992819678e-05,
      "loss": 0.0431,
      "step": 2136
    },
    {
      "epoch": 1.6031507876969242,
      "grad_norm": 0.22901701927185059,
      "learning_rate": 1.15253057724297e-05,
      "loss": 0.0562,
      "step": 2137
    },
    {
      "epoch": 1.603900975243811,
      "grad_norm": 0.26473501324653625,
      "learning_rate": 1.1483521532351654e-05,
      "loss": 0.0505,
      "step": 2138
    },
    {
      "epoch": 1.6046511627906976,
      "grad_norm": 0.14857040345668793,
      "learning_rate": 1.144180334424141e-05,
      "loss": 0.0343,
      "step": 2139
    },
    {
      "epoch": 1.6054013503375844,
      "grad_norm": 0.1499057114124298,
      "learning_rate": 1.1400151279641525e-05,
      "loss": 0.0259,
      "step": 2140
    },
    {
      "epoch": 1.606151537884471,
      "grad_norm": 0.1453072428703308,
      "learning_rate": 1.1358565409981203e-05,
      "loss": 0.0292,
      "step": 2141
    },
    {
      "epoch": 1.6069017254313578,
      "grad_norm": 0.18547675013542175,
      "learning_rate": 1.1317045806576121e-05,
      "loss": 0.039,
      "step": 2142
    },
    {
      "epoch": 1.6076519129782447,
      "grad_norm": 0.2546239197254181,
      "learning_rate": 1.12755925406283e-05,
      "loss": 0.0736,
      "step": 2143
    },
    {
      "epoch": 1.6084021005251312,
      "grad_norm": 0.1635519117116928,
      "learning_rate": 1.1234205683226012e-05,
      "loss": 0.0301,
      "step": 2144
    },
    {
      "epoch": 1.6091522880720182,
      "grad_norm": 0.14328426122665405,
      "learning_rate": 1.1192885305343648e-05,
      "loss": 0.0429,
      "step": 2145
    },
    {
      "epoch": 1.6099024756189046,
      "grad_norm": 0.21794961392879486,
      "learning_rate": 1.1151631477841584e-05,
      "loss": 0.0577,
      "step": 2146
    },
    {
      "epoch": 1.6106526631657916,
      "grad_norm": 0.19154462218284607,
      "learning_rate": 1.1110444271466086e-05,
      "loss": 0.046,
      "step": 2147
    },
    {
      "epoch": 1.611402850712678,
      "grad_norm": 0.16161414980888367,
      "learning_rate": 1.1069323756849126e-05,
      "loss": 0.0561,
      "step": 2148
    },
    {
      "epoch": 1.612153038259565,
      "grad_norm": 0.21912692487239838,
      "learning_rate": 1.102827000450835e-05,
      "loss": 0.0428,
      "step": 2149
    },
    {
      "epoch": 1.6129032258064515,
      "grad_norm": 0.20931370556354523,
      "learning_rate": 1.0987283084846905e-05,
      "loss": 0.0545,
      "step": 2150
    },
    {
      "epoch": 1.6136534133533385,
      "grad_norm": 0.16395579278469086,
      "learning_rate": 1.0946363068153343e-05,
      "loss": 0.0344,
      "step": 2151
    },
    {
      "epoch": 1.614403600900225,
      "grad_norm": 0.24834044277668,
      "learning_rate": 1.0905510024601423e-05,
      "loss": 0.0487,
      "step": 2152
    },
    {
      "epoch": 1.6151537884471119,
      "grad_norm": 0.15264178812503815,
      "learning_rate": 1.0864724024250106e-05,
      "loss": 0.0316,
      "step": 2153
    },
    {
      "epoch": 1.6159039759939984,
      "grad_norm": 0.10665671527385712,
      "learning_rate": 1.0824005137043375e-05,
      "loss": 0.0192,
      "step": 2154
    },
    {
      "epoch": 1.6166541635408853,
      "grad_norm": 0.17089134454727173,
      "learning_rate": 1.0783353432810106e-05,
      "loss": 0.0426,
      "step": 2155
    },
    {
      "epoch": 1.6174043510877718,
      "grad_norm": 0.22492165863513947,
      "learning_rate": 1.0742768981263984e-05,
      "loss": 0.0708,
      "step": 2156
    },
    {
      "epoch": 1.6181545386346587,
      "grad_norm": 0.11420857906341553,
      "learning_rate": 1.070225185200331e-05,
      "loss": 0.0254,
      "step": 2157
    },
    {
      "epoch": 1.6189047261815452,
      "grad_norm": 0.2135169953107834,
      "learning_rate": 1.0661802114511005e-05,
      "loss": 0.0719,
      "step": 2158
    },
    {
      "epoch": 1.6196549137284322,
      "grad_norm": 0.18792471289634705,
      "learning_rate": 1.062141983815439e-05,
      "loss": 0.0557,
      "step": 2159
    },
    {
      "epoch": 1.6204051012753187,
      "grad_norm": 0.17168648540973663,
      "learning_rate": 1.0581105092185062e-05,
      "loss": 0.0556,
      "step": 2160
    },
    {
      "epoch": 1.6211552888222056,
      "grad_norm": 0.15384042263031006,
      "learning_rate": 1.0540857945738852e-05,
      "loss": 0.0441,
      "step": 2161
    },
    {
      "epoch": 1.6219054763690923,
      "grad_norm": 0.11775700747966766,
      "learning_rate": 1.0500678467835662e-05,
      "loss": 0.0239,
      "step": 2162
    },
    {
      "epoch": 1.622655663915979,
      "grad_norm": 0.1715250313282013,
      "learning_rate": 1.0460566727379335e-05,
      "loss": 0.0487,
      "step": 2163
    },
    {
      "epoch": 1.6234058514628658,
      "grad_norm": 0.18916262686252594,
      "learning_rate": 1.0420522793157567e-05,
      "loss": 0.0467,
      "step": 2164
    },
    {
      "epoch": 1.6241560390097525,
      "grad_norm": 0.29144486784935,
      "learning_rate": 1.038054673384174e-05,
      "loss": 0.0472,
      "step": 2165
    },
    {
      "epoch": 1.6249062265566392,
      "grad_norm": 0.16755078732967377,
      "learning_rate": 1.0340638617986864e-05,
      "loss": 0.035,
      "step": 2166
    },
    {
      "epoch": 1.625656414103526,
      "grad_norm": 0.1979805827140808,
      "learning_rate": 1.030079851403144e-05,
      "loss": 0.0474,
      "step": 2167
    },
    {
      "epoch": 1.6264066016504126,
      "grad_norm": 0.21421858668327332,
      "learning_rate": 1.0261026490297315e-05,
      "loss": 0.0465,
      "step": 2168
    },
    {
      "epoch": 1.6271567891972993,
      "grad_norm": 0.19427688419818878,
      "learning_rate": 1.022132261498961e-05,
      "loss": 0.061,
      "step": 2169
    },
    {
      "epoch": 1.627906976744186,
      "grad_norm": 0.1771208941936493,
      "learning_rate": 1.0181686956196529e-05,
      "loss": 0.0451,
      "step": 2170
    },
    {
      "epoch": 1.6286571642910728,
      "grad_norm": 0.19580066204071045,
      "learning_rate": 1.0142119581889332e-05,
      "loss": 0.0619,
      "step": 2171
    },
    {
      "epoch": 1.6294073518379595,
      "grad_norm": 0.20414824783802032,
      "learning_rate": 1.0102620559922204e-05,
      "loss": 0.0409,
      "step": 2172
    },
    {
      "epoch": 1.6301575393848462,
      "grad_norm": 0.19486220180988312,
      "learning_rate": 1.0063189958032043e-05,
      "loss": 0.0449,
      "step": 2173
    },
    {
      "epoch": 1.630907726931733,
      "grad_norm": 0.27886173129081726,
      "learning_rate": 1.0023827843838457e-05,
      "loss": 0.0652,
      "step": 2174
    },
    {
      "epoch": 1.6316579144786196,
      "grad_norm": 0.17671063542366028,
      "learning_rate": 9.984534284843594e-06,
      "loss": 0.0487,
      "step": 2175
    },
    {
      "epoch": 1.6324081020255063,
      "grad_norm": 0.24155737459659576,
      "learning_rate": 9.945309348432047e-06,
      "loss": 0.057,
      "step": 2176
    },
    {
      "epoch": 1.633158289572393,
      "grad_norm": 0.14792385697364807,
      "learning_rate": 9.906153101870725e-06,
      "loss": 0.0376,
      "step": 2177
    },
    {
      "epoch": 1.6339084771192798,
      "grad_norm": 0.2489890605211258,
      "learning_rate": 9.867065612308713e-06,
      "loss": 0.0447,
      "step": 2178
    },
    {
      "epoch": 1.6346586646661665,
      "grad_norm": 0.16253811120986938,
      "learning_rate": 9.82804694677722e-06,
      "loss": 0.0548,
      "step": 2179
    },
    {
      "epoch": 1.6354088522130532,
      "grad_norm": 0.20979243516921997,
      "learning_rate": 9.78909717218941e-06,
      "loss": 0.0375,
      "step": 2180
    },
    {
      "epoch": 1.63615903975994,
      "grad_norm": 0.16433793306350708,
      "learning_rate": 9.75021635534033e-06,
      "loss": 0.0362,
      "step": 2181
    },
    {
      "epoch": 1.6369092273068269,
      "grad_norm": 0.13124553859233856,
      "learning_rate": 9.711404562906717e-06,
      "loss": 0.0303,
      "step": 2182
    },
    {
      "epoch": 1.6376594148537134,
      "grad_norm": 0.2880100905895233,
      "learning_rate": 9.672661861447002e-06,
      "loss": 0.0638,
      "step": 2183
    },
    {
      "epoch": 1.6384096024006003,
      "grad_norm": 0.2295902669429779,
      "learning_rate": 9.633988317401087e-06,
      "loss": 0.0447,
      "step": 2184
    },
    {
      "epoch": 1.6391597899474868,
      "grad_norm": 0.22860339283943176,
      "learning_rate": 9.595383997090302e-06,
      "loss": 0.0571,
      "step": 2185
    },
    {
      "epoch": 1.6399099774943737,
      "grad_norm": 0.2065700888633728,
      "learning_rate": 9.556848966717247e-06,
      "loss": 0.0612,
      "step": 2186
    },
    {
      "epoch": 1.6406601650412602,
      "grad_norm": 0.18000207841396332,
      "learning_rate": 9.518383292365713e-06,
      "loss": 0.0428,
      "step": 2187
    },
    {
      "epoch": 1.6414103525881472,
      "grad_norm": 0.18875566124916077,
      "learning_rate": 9.479987040000538e-06,
      "loss": 0.0407,
      "step": 2188
    },
    {
      "epoch": 1.6421605401350337,
      "grad_norm": 0.18922288715839386,
      "learning_rate": 9.441660275467512e-06,
      "loss": 0.0433,
      "step": 2189
    },
    {
      "epoch": 1.6429107276819206,
      "grad_norm": 0.19266372919082642,
      "learning_rate": 9.403403064493282e-06,
      "loss": 0.0424,
      "step": 2190
    },
    {
      "epoch": 1.643660915228807,
      "grad_norm": 0.18794898688793182,
      "learning_rate": 9.365215472685163e-06,
      "loss": 0.0674,
      "step": 2191
    },
    {
      "epoch": 1.644411102775694,
      "grad_norm": 0.25597575306892395,
      "learning_rate": 9.32709756553114e-06,
      "loss": 0.0435,
      "step": 2192
    },
    {
      "epoch": 1.6451612903225805,
      "grad_norm": 0.2104489505290985,
      "learning_rate": 9.289049408399659e-06,
      "loss": 0.0417,
      "step": 2193
    },
    {
      "epoch": 1.6459114778694675,
      "grad_norm": 0.17522476613521576,
      "learning_rate": 9.251071066539579e-06,
      "loss": 0.0411,
      "step": 2194
    },
    {
      "epoch": 1.646661665416354,
      "grad_norm": 0.21473297476768494,
      "learning_rate": 9.21316260507999e-06,
      "loss": 0.0542,
      "step": 2195
    },
    {
      "epoch": 1.6474118529632409,
      "grad_norm": 0.18432407081127167,
      "learning_rate": 9.175324089030185e-06,
      "loss": 0.0482,
      "step": 2196
    },
    {
      "epoch": 1.6481620405101274,
      "grad_norm": 0.18733084201812744,
      "learning_rate": 9.137555583279495e-06,
      "loss": 0.045,
      "step": 2197
    },
    {
      "epoch": 1.6489122280570143,
      "grad_norm": 0.23316556215286255,
      "learning_rate": 9.099857152597185e-06,
      "loss": 0.0437,
      "step": 2198
    },
    {
      "epoch": 1.6496624156039008,
      "grad_norm": 0.1609441488981247,
      "learning_rate": 9.062228861632354e-06,
      "loss": 0.0389,
      "step": 2199
    },
    {
      "epoch": 1.6504126031507877,
      "grad_norm": 0.1733076125383377,
      "learning_rate": 9.024670774913812e-06,
      "loss": 0.0471,
      "step": 2200
    },
    {
      "epoch": 1.6504126031507877,
      "eval_loss": 0.06824243068695068,
      "eval_runtime": 2.661,
      "eval_samples_per_second": 20.293,
      "eval_steps_per_second": 5.261,
      "step": 2200
    },
    {
      "epoch": 1.6511627906976745,
      "grad_norm": 0.2067050188779831,
      "learning_rate": 8.987182956849983e-06,
      "loss": 0.0539,
      "step": 2201
    },
    {
      "epoch": 1.6519129782445612,
      "grad_norm": 0.177916020154953,
      "learning_rate": 8.949765471728789e-06,
      "loss": 0.0442,
      "step": 2202
    },
    {
      "epoch": 1.652663165791448,
      "grad_norm": 0.25947102904319763,
      "learning_rate": 8.912418383717513e-06,
      "loss": 0.0599,
      "step": 2203
    },
    {
      "epoch": 1.6534133533383346,
      "grad_norm": 0.1599457561969757,
      "learning_rate": 8.875141756862749e-06,
      "loss": 0.0352,
      "step": 2204
    },
    {
      "epoch": 1.6541635408852213,
      "grad_norm": 0.24875929951667786,
      "learning_rate": 8.837935655090241e-06,
      "loss": 0.0419,
      "step": 2205
    },
    {
      "epoch": 1.654913728432108,
      "grad_norm": 0.14069050550460815,
      "learning_rate": 8.800800142204779e-06,
      "loss": 0.0283,
      "step": 2206
    },
    {
      "epoch": 1.6556639159789948,
      "grad_norm": 0.17972294986248016,
      "learning_rate": 8.763735281890133e-06,
      "loss": 0.0311,
      "step": 2207
    },
    {
      "epoch": 1.6564141035258815,
      "grad_norm": 0.13864688575267792,
      "learning_rate": 8.726741137708866e-06,
      "loss": 0.0264,
      "step": 2208
    },
    {
      "epoch": 1.6571642910727682,
      "grad_norm": 0.1534036248922348,
      "learning_rate": 8.689817773102293e-06,
      "loss": 0.0284,
      "step": 2209
    },
    {
      "epoch": 1.657914478619655,
      "grad_norm": 0.231153205037117,
      "learning_rate": 8.65296525139036e-06,
      "loss": 0.0596,
      "step": 2210
    },
    {
      "epoch": 1.6586646661665416,
      "grad_norm": 0.21296992897987366,
      "learning_rate": 8.616183635771525e-06,
      "loss": 0.0418,
      "step": 2211
    },
    {
      "epoch": 1.6594148537134283,
      "grad_norm": 0.1815291792154312,
      "learning_rate": 8.579472989322602e-06,
      "loss": 0.0419,
      "step": 2212
    },
    {
      "epoch": 1.660165041260315,
      "grad_norm": 0.16089360415935516,
      "learning_rate": 8.542833374998744e-06,
      "loss": 0.0344,
      "step": 2213
    },
    {
      "epoch": 1.6609152288072018,
      "grad_norm": 0.35123497247695923,
      "learning_rate": 8.5062648556333e-06,
      "loss": 0.0642,
      "step": 2214
    },
    {
      "epoch": 1.6616654163540885,
      "grad_norm": 0.15117251873016357,
      "learning_rate": 8.469767493937681e-06,
      "loss": 0.0468,
      "step": 2215
    },
    {
      "epoch": 1.6624156039009752,
      "grad_norm": 0.1996140033006668,
      "learning_rate": 8.43334135250125e-06,
      "loss": 0.0451,
      "step": 2216
    },
    {
      "epoch": 1.663165791447862,
      "grad_norm": 0.1861112117767334,
      "learning_rate": 8.39698649379126e-06,
      "loss": 0.0437,
      "step": 2217
    },
    {
      "epoch": 1.6639159789947486,
      "grad_norm": 0.21752230823040009,
      "learning_rate": 8.360702980152713e-06,
      "loss": 0.0522,
      "step": 2218
    },
    {
      "epoch": 1.6646661665416356,
      "grad_norm": 0.19835425913333893,
      "learning_rate": 8.32449087380826e-06,
      "loss": 0.0337,
      "step": 2219
    },
    {
      "epoch": 1.665416354088522,
      "grad_norm": 0.18913261592388153,
      "learning_rate": 8.288350236858117e-06,
      "loss": 0.0622,
      "step": 2220
    },
    {
      "epoch": 1.666166541635409,
      "grad_norm": 0.22755774855613708,
      "learning_rate": 8.252281131279887e-06,
      "loss": 0.0503,
      "step": 2221
    },
    {
      "epoch": 1.6669167291822955,
      "grad_norm": 0.14972837269306183,
      "learning_rate": 8.21628361892855e-06,
      "loss": 0.0353,
      "step": 2222
    },
    {
      "epoch": 1.6676669167291824,
      "grad_norm": 0.16990503668785095,
      "learning_rate": 8.180357761536296e-06,
      "loss": 0.0348,
      "step": 2223
    },
    {
      "epoch": 1.668417104276069,
      "grad_norm": 0.21821126341819763,
      "learning_rate": 8.14450362071244e-06,
      "loss": 0.0561,
      "step": 2224
    },
    {
      "epoch": 1.6691672918229559,
      "grad_norm": 0.2086024284362793,
      "learning_rate": 8.10872125794328e-06,
      "loss": 0.0665,
      "step": 2225
    },
    {
      "epoch": 1.6699174793698424,
      "grad_norm": 0.18031302094459534,
      "learning_rate": 8.073010734592057e-06,
      "loss": 0.0305,
      "step": 2226
    },
    {
      "epoch": 1.6706676669167293,
      "grad_norm": 0.14847469329833984,
      "learning_rate": 8.037372111898789e-06,
      "loss": 0.0218,
      "step": 2227
    },
    {
      "epoch": 1.6714178544636158,
      "grad_norm": 0.2060483992099762,
      "learning_rate": 8.001805450980249e-06,
      "loss": 0.0472,
      "step": 2228
    },
    {
      "epoch": 1.6721680420105027,
      "grad_norm": 0.11653721332550049,
      "learning_rate": 7.966310812829709e-06,
      "loss": 0.0248,
      "step": 2229
    },
    {
      "epoch": 1.6729182295573892,
      "grad_norm": 0.2690200209617615,
      "learning_rate": 7.930888258316998e-06,
      "loss": 0.0594,
      "step": 2230
    },
    {
      "epoch": 1.6736684171042762,
      "grad_norm": 0.2156667709350586,
      "learning_rate": 7.89553784818831e-06,
      "loss": 0.0542,
      "step": 2231
    },
    {
      "epoch": 1.6744186046511627,
      "grad_norm": 0.1282075047492981,
      "learning_rate": 7.860259643066126e-06,
      "loss": 0.0209,
      "step": 2232
    },
    {
      "epoch": 1.6751687921980496,
      "grad_norm": 0.186884343624115,
      "learning_rate": 7.82505370344907e-06,
      "loss": 0.0573,
      "step": 2233
    },
    {
      "epoch": 1.675918979744936,
      "grad_norm": 0.13679596781730652,
      "learning_rate": 7.789920089711871e-06,
      "loss": 0.0264,
      "step": 2234
    },
    {
      "epoch": 1.676669167291823,
      "grad_norm": 0.14465893805027008,
      "learning_rate": 7.754858862105224e-06,
      "loss": 0.0226,
      "step": 2235
    },
    {
      "epoch": 1.6774193548387095,
      "grad_norm": 0.16200248897075653,
      "learning_rate": 7.71987008075568e-06,
      "loss": 0.0292,
      "step": 2236
    },
    {
      "epoch": 1.6781695423855965,
      "grad_norm": 0.20817679166793823,
      "learning_rate": 7.684953805665562e-06,
      "loss": 0.0463,
      "step": 2237
    },
    {
      "epoch": 1.6789197299324832,
      "grad_norm": 0.14461885392665863,
      "learning_rate": 7.65011009671282e-06,
      "loss": 0.0439,
      "step": 2238
    },
    {
      "epoch": 1.67966991747937,
      "grad_norm": 0.17346525192260742,
      "learning_rate": 7.615339013651001e-06,
      "loss": 0.0569,
      "step": 2239
    },
    {
      "epoch": 1.6804201050262566,
      "grad_norm": 0.11375146359205246,
      "learning_rate": 7.580640616109081e-06,
      "loss": 0.0219,
      "step": 2240
    },
    {
      "epoch": 1.6811702925731433,
      "grad_norm": 0.1951054185628891,
      "learning_rate": 7.546014963591397e-06,
      "loss": 0.0417,
      "step": 2241
    },
    {
      "epoch": 1.68192048012003,
      "grad_norm": 0.18035517632961273,
      "learning_rate": 7.511462115477536e-06,
      "loss": 0.0473,
      "step": 2242
    },
    {
      "epoch": 1.6826706676669168,
      "grad_norm": 0.21403639018535614,
      "learning_rate": 7.476982131022231e-06,
      "loss": 0.0366,
      "step": 2243
    },
    {
      "epoch": 1.6834208552138035,
      "grad_norm": 0.23407121002674103,
      "learning_rate": 7.442575069355256e-06,
      "loss": 0.0387,
      "step": 2244
    },
    {
      "epoch": 1.6841710427606902,
      "grad_norm": 0.17773695290088654,
      "learning_rate": 7.408240989481347e-06,
      "loss": 0.0399,
      "step": 2245
    },
    {
      "epoch": 1.684921230307577,
      "grad_norm": 0.25798365473747253,
      "learning_rate": 7.373979950280046e-06,
      "loss": 0.0537,
      "step": 2246
    },
    {
      "epoch": 1.6856714178544636,
      "grad_norm": 0.16044406592845917,
      "learning_rate": 7.33979201050568e-06,
      "loss": 0.0296,
      "step": 2247
    },
    {
      "epoch": 1.6864216054013503,
      "grad_norm": 0.182396799325943,
      "learning_rate": 7.3056772287871886e-06,
      "loss": 0.0296,
      "step": 2248
    },
    {
      "epoch": 1.687171792948237,
      "grad_norm": 0.21489420533180237,
      "learning_rate": 7.2716356636280684e-06,
      "loss": 0.0528,
      "step": 2249
    },
    {
      "epoch": 1.6879219804951238,
      "grad_norm": 0.18872296810150146,
      "learning_rate": 7.237667373406259e-06,
      "loss": 0.0326,
      "step": 2250
    },
    {
      "epoch": 1.6886721680420105,
      "grad_norm": 0.16861742734909058,
      "learning_rate": 7.203772416374016e-06,
      "loss": 0.0558,
      "step": 2251
    },
    {
      "epoch": 1.6894223555888972,
      "grad_norm": 0.22231639921665192,
      "learning_rate": 7.1699508506578636e-06,
      "loss": 0.0459,
      "step": 2252
    },
    {
      "epoch": 1.690172543135784,
      "grad_norm": 0.1857728660106659,
      "learning_rate": 7.136202734258457e-06,
      "loss": 0.0422,
      "step": 2253
    },
    {
      "epoch": 1.6909227306826706,
      "grad_norm": 0.20071634650230408,
      "learning_rate": 7.1025281250505006e-06,
      "loss": 0.0436,
      "step": 2254
    },
    {
      "epoch": 1.6916729182295573,
      "grad_norm": 0.20031875371932983,
      "learning_rate": 7.0689270807826e-06,
      "loss": 0.0341,
      "step": 2255
    },
    {
      "epoch": 1.692423105776444,
      "grad_norm": 0.16592131555080414,
      "learning_rate": 7.035399659077268e-06,
      "loss": 0.0429,
      "step": 2256
    },
    {
      "epoch": 1.6931732933233308,
      "grad_norm": 0.2352314591407776,
      "learning_rate": 7.00194591743073e-06,
      "loss": 0.0437,
      "step": 2257
    },
    {
      "epoch": 1.6939234808702177,
      "grad_norm": 0.17028115689754486,
      "learning_rate": 6.96856591321286e-06,
      "loss": 0.0333,
      "step": 2258
    },
    {
      "epoch": 1.6946736684171042,
      "grad_norm": 0.19160404801368713,
      "learning_rate": 6.9352597036670575e-06,
      "loss": 0.0613,
      "step": 2259
    },
    {
      "epoch": 1.6954238559639911,
      "grad_norm": 0.17852510511875153,
      "learning_rate": 6.902027345910211e-06,
      "loss": 0.038,
      "step": 2260
    },
    {
      "epoch": 1.6961740435108776,
      "grad_norm": 0.1913464069366455,
      "learning_rate": 6.868868896932534e-06,
      "loss": 0.0611,
      "step": 2261
    },
    {
      "epoch": 1.6969242310577646,
      "grad_norm": 0.2039322406053543,
      "learning_rate": 6.835784413597512e-06,
      "loss": 0.0546,
      "step": 2262
    },
    {
      "epoch": 1.697674418604651,
      "grad_norm": 0.1932404786348343,
      "learning_rate": 6.802773952641761e-06,
      "loss": 0.0452,
      "step": 2263
    },
    {
      "epoch": 1.698424606151538,
      "grad_norm": 0.17494481801986694,
      "learning_rate": 6.769837570674975e-06,
      "loss": 0.0472,
      "step": 2264
    },
    {
      "epoch": 1.6991747936984245,
      "grad_norm": 0.1976795345544815,
      "learning_rate": 6.7369753241798114e-06,
      "loss": 0.0379,
      "step": 2265
    },
    {
      "epoch": 1.6999249812453114,
      "grad_norm": 0.18811482191085815,
      "learning_rate": 6.70418726951178e-06,
      "loss": 0.0518,
      "step": 2266
    },
    {
      "epoch": 1.700675168792198,
      "grad_norm": 0.30055856704711914,
      "learning_rate": 6.671473462899181e-06,
      "loss": 0.061,
      "step": 2267
    },
    {
      "epoch": 1.7014253563390849,
      "grad_norm": 0.23877814412117004,
      "learning_rate": 6.638833960442948e-06,
      "loss": 0.0423,
      "step": 2268
    },
    {
      "epoch": 1.7021755438859714,
      "grad_norm": 0.16119158267974854,
      "learning_rate": 6.606268818116618e-06,
      "loss": 0.0397,
      "step": 2269
    },
    {
      "epoch": 1.7029257314328583,
      "grad_norm": 0.2363862246274948,
      "learning_rate": 6.573778091766219e-06,
      "loss": 0.0566,
      "step": 2270
    },
    {
      "epoch": 1.7036759189797448,
      "grad_norm": 0.21147215366363525,
      "learning_rate": 6.541361837110149e-06,
      "loss": 0.0564,
      "step": 2271
    },
    {
      "epoch": 1.7044261065266317,
      "grad_norm": 0.12778374552726746,
      "learning_rate": 6.509020109739078e-06,
      "loss": 0.0354,
      "step": 2272
    },
    {
      "epoch": 1.7051762940735182,
      "grad_norm": 0.2169772982597351,
      "learning_rate": 6.476752965115884e-06,
      "loss": 0.0418,
      "step": 2273
    },
    {
      "epoch": 1.7059264816204052,
      "grad_norm": 0.14315977692604065,
      "learning_rate": 6.444560458575544e-06,
      "loss": 0.0353,
      "step": 2274
    },
    {
      "epoch": 1.7066766691672917,
      "grad_norm": 0.1578858345746994,
      "learning_rate": 6.412442645325057e-06,
      "loss": 0.0338,
      "step": 2275
    },
    {
      "epoch": 1.7074268567141786,
      "grad_norm": 0.13712339103221893,
      "learning_rate": 6.38039958044328e-06,
      "loss": 0.0278,
      "step": 2276
    },
    {
      "epoch": 1.7081770442610653,
      "grad_norm": 0.19122560322284698,
      "learning_rate": 6.3484313188809265e-06,
      "loss": 0.0372,
      "step": 2277
    },
    {
      "epoch": 1.708927231807952,
      "grad_norm": 0.22988127171993256,
      "learning_rate": 6.316537915460418e-06,
      "loss": 0.0561,
      "step": 2278
    },
    {
      "epoch": 1.7096774193548387,
      "grad_norm": 0.13822755217552185,
      "learning_rate": 6.284719424875796e-06,
      "loss": 0.0279,
      "step": 2279
    },
    {
      "epoch": 1.7104276069017255,
      "grad_norm": 0.1539996862411499,
      "learning_rate": 6.252975901692659e-06,
      "loss": 0.0439,
      "step": 2280
    },
    {
      "epoch": 1.7111777944486122,
      "grad_norm": 0.15642352402210236,
      "learning_rate": 6.221307400347992e-06,
      "loss": 0.0254,
      "step": 2281
    },
    {
      "epoch": 1.711927981995499,
      "grad_norm": 0.1779630482196808,
      "learning_rate": 6.1897139751501796e-06,
      "loss": 0.0535,
      "step": 2282
    },
    {
      "epoch": 1.7126781695423856,
      "grad_norm": 0.18841272592544556,
      "learning_rate": 6.158195680278816e-06,
      "loss": 0.049,
      "step": 2283
    },
    {
      "epoch": 1.7134283570892723,
      "grad_norm": 0.2371242642402649,
      "learning_rate": 6.126752569784694e-06,
      "loss": 0.0438,
      "step": 2284
    },
    {
      "epoch": 1.714178544636159,
      "grad_norm": 0.15389996767044067,
      "learning_rate": 6.095384697589635e-06,
      "loss": 0.033,
      "step": 2285
    },
    {
      "epoch": 1.7149287321830458,
      "grad_norm": 0.14900945127010345,
      "learning_rate": 6.064092117486464e-06,
      "loss": 0.0337,
      "step": 2286
    },
    {
      "epoch": 1.7156789197299325,
      "grad_norm": 0.13682608306407928,
      "learning_rate": 6.032874883138867e-06,
      "loss": 0.0327,
      "step": 2287
    },
    {
      "epoch": 1.7164291072768192,
      "grad_norm": 0.1934865266084671,
      "learning_rate": 6.001733048081337e-06,
      "loss": 0.0405,
      "step": 2288
    },
    {
      "epoch": 1.717179294823706,
      "grad_norm": 0.2710997462272644,
      "learning_rate": 5.970666665719033e-06,
      "loss": 0.0511,
      "step": 2289
    },
    {
      "epoch": 1.7179294823705926,
      "grad_norm": 0.2753269374370575,
      "learning_rate": 5.939675789327759e-06,
      "loss": 0.0586,
      "step": 2290
    },
    {
      "epoch": 1.7186796699174793,
      "grad_norm": 0.19517338275909424,
      "learning_rate": 5.908760472053809e-06,
      "loss": 0.0648,
      "step": 2291
    },
    {
      "epoch": 1.719429857464366,
      "grad_norm": 0.21337710320949554,
      "learning_rate": 5.877920766913919e-06,
      "loss": 0.0405,
      "step": 2292
    },
    {
      "epoch": 1.7201800450112528,
      "grad_norm": 0.20285245776176453,
      "learning_rate": 5.847156726795133e-06,
      "loss": 0.0519,
      "step": 2293
    },
    {
      "epoch": 1.7209302325581395,
      "grad_norm": 0.19896537065505981,
      "learning_rate": 5.816468404454755e-06,
      "loss": 0.064,
      "step": 2294
    },
    {
      "epoch": 1.7216804201050264,
      "grad_norm": 0.17619459331035614,
      "learning_rate": 5.7858558525202336e-06,
      "loss": 0.0305,
      "step": 2295
    },
    {
      "epoch": 1.722430607651913,
      "grad_norm": 0.1564476639032364,
      "learning_rate": 5.755319123489083e-06,
      "loss": 0.0303,
      "step": 2296
    },
    {
      "epoch": 1.7231807951987999,
      "grad_norm": 0.13188333809375763,
      "learning_rate": 5.724858269728789e-06,
      "loss": 0.0277,
      "step": 2297
    },
    {
      "epoch": 1.7239309827456863,
      "grad_norm": 0.24183008074760437,
      "learning_rate": 5.694473343476714e-06,
      "loss": 0.0393,
      "step": 2298
    },
    {
      "epoch": 1.7246811702925733,
      "grad_norm": 0.2262912541627884,
      "learning_rate": 5.664164396840016e-06,
      "loss": 0.0579,
      "step": 2299
    },
    {
      "epoch": 1.7254313578394598,
      "grad_norm": 0.19143463671207428,
      "learning_rate": 5.633931481795552e-06,
      "loss": 0.037,
      "step": 2300
    },
    {
      "epoch": 1.7261815453863467,
      "grad_norm": 0.1472475230693817,
      "learning_rate": 5.603774650189808e-06,
      "loss": 0.042,
      "step": 2301
    },
    {
      "epoch": 1.7269317329332332,
      "grad_norm": 0.20033307373523712,
      "learning_rate": 5.573693953738751e-06,
      "loss": 0.0484,
      "step": 2302
    },
    {
      "epoch": 1.7276819204801201,
      "grad_norm": 0.15765565633773804,
      "learning_rate": 5.543689444027839e-06,
      "loss": 0.0344,
      "step": 2303
    },
    {
      "epoch": 1.7284321080270066,
      "grad_norm": 0.13256795704364777,
      "learning_rate": 5.513761172511833e-06,
      "loss": 0.0288,
      "step": 2304
    },
    {
      "epoch": 1.7291822955738936,
      "grad_norm": 0.24713698029518127,
      "learning_rate": 5.483909190514797e-06,
      "loss": 0.0482,
      "step": 2305
    },
    {
      "epoch": 1.72993248312078,
      "grad_norm": 0.19771939516067505,
      "learning_rate": 5.4541335492299115e-06,
      "loss": 0.0409,
      "step": 2306
    },
    {
      "epoch": 1.730682670667667,
      "grad_norm": 0.1739976406097412,
      "learning_rate": 5.424434299719483e-06,
      "loss": 0.034,
      "step": 2307
    },
    {
      "epoch": 1.7314328582145535,
      "grad_norm": 0.13870416581630707,
      "learning_rate": 5.394811492914803e-06,
      "loss": 0.0357,
      "step": 2308
    },
    {
      "epoch": 1.7321830457614404,
      "grad_norm": 0.15554337203502655,
      "learning_rate": 5.365265179616063e-06,
      "loss": 0.0295,
      "step": 2309
    },
    {
      "epoch": 1.732933233308327,
      "grad_norm": 0.21074527502059937,
      "learning_rate": 5.3357954104922895e-06,
      "loss": 0.0365,
      "step": 2310
    },
    {
      "epoch": 1.7336834208552139,
      "grad_norm": 0.22298146784305573,
      "learning_rate": 5.306402236081209e-06,
      "loss": 0.0796,
      "step": 2311
    },
    {
      "epoch": 1.7344336084021004,
      "grad_norm": 0.12427302449941635,
      "learning_rate": 5.277085706789248e-06,
      "loss": 0.026,
      "step": 2312
    },
    {
      "epoch": 1.7351837959489873,
      "grad_norm": 0.20607198774814606,
      "learning_rate": 5.247845872891371e-06,
      "loss": 0.0478,
      "step": 2313
    },
    {
      "epoch": 1.735933983495874,
      "grad_norm": 0.22108227014541626,
      "learning_rate": 5.218682784530993e-06,
      "loss": 0.05,
      "step": 2314
    },
    {
      "epoch": 1.7366841710427607,
      "grad_norm": 0.17740093171596527,
      "learning_rate": 5.1895964917199445e-06,
      "loss": 0.0367,
      "step": 2315
    },
    {
      "epoch": 1.7374343585896475,
      "grad_norm": 0.2420615255832672,
      "learning_rate": 5.160587044338355e-06,
      "loss": 0.0568,
      "step": 2316
    },
    {
      "epoch": 1.7381845461365342,
      "grad_norm": 0.14381897449493408,
      "learning_rate": 5.131654492134574e-06,
      "loss": 0.0402,
      "step": 2317
    },
    {
      "epoch": 1.7389347336834209,
      "grad_norm": 0.1720701903104782,
      "learning_rate": 5.102798884725091e-06,
      "loss": 0.037,
      "step": 2318
    },
    {
      "epoch": 1.7396849212303076,
      "grad_norm": 0.14338922500610352,
      "learning_rate": 5.074020271594404e-06,
      "loss": 0.0385,
      "step": 2319
    },
    {
      "epoch": 1.7404351087771943,
      "grad_norm": 0.19156095385551453,
      "learning_rate": 5.045318702095014e-06,
      "loss": 0.0549,
      "step": 2320
    },
    {
      "epoch": 1.741185296324081,
      "grad_norm": 0.18174941837787628,
      "learning_rate": 5.016694225447288e-06,
      "loss": 0.04,
      "step": 2321
    },
    {
      "epoch": 1.7419354838709677,
      "grad_norm": 0.21480074524879456,
      "learning_rate": 4.988146890739381e-06,
      "loss": 0.0524,
      "step": 2322
    },
    {
      "epoch": 1.7426856714178545,
      "grad_norm": 0.14862346649169922,
      "learning_rate": 4.959676746927172e-06,
      "loss": 0.0345,
      "step": 2323
    },
    {
      "epoch": 1.7434358589647412,
      "grad_norm": 0.3776375651359558,
      "learning_rate": 4.931283842834139e-06,
      "loss": 0.033,
      "step": 2324
    },
    {
      "epoch": 1.744186046511628,
      "grad_norm": 0.21091584861278534,
      "learning_rate": 4.902968227151311e-06,
      "loss": 0.0448,
      "step": 2325
    },
    {
      "epoch": 1.7449362340585146,
      "grad_norm": 0.13311506807804108,
      "learning_rate": 4.874729948437218e-06,
      "loss": 0.0267,
      "step": 2326
    },
    {
      "epoch": 1.7456864216054013,
      "grad_norm": 0.1780560463666916,
      "learning_rate": 4.846569055117684e-06,
      "loss": 0.0307,
      "step": 2327
    },
    {
      "epoch": 1.746436609152288,
      "grad_norm": 0.19919991493225098,
      "learning_rate": 4.818485595485889e-06,
      "loss": 0.0704,
      "step": 2328
    },
    {
      "epoch": 1.7471867966991748,
      "grad_norm": 0.19128254055976868,
      "learning_rate": 4.790479617702198e-06,
      "loss": 0.051,
      "step": 2329
    },
    {
      "epoch": 1.7479369842460615,
      "grad_norm": 0.16919958591461182,
      "learning_rate": 4.762551169794105e-06,
      "loss": 0.0369,
      "step": 2330
    },
    {
      "epoch": 1.7486871717929482,
      "grad_norm": 0.19580717384815216,
      "learning_rate": 4.734700299656158e-06,
      "loss": 0.041,
      "step": 2331
    },
    {
      "epoch": 1.749437359339835,
      "grad_norm": 0.20923122763633728,
      "learning_rate": 4.706927055049837e-06,
      "loss": 0.0601,
      "step": 2332
    },
    {
      "epoch": 1.7501875468867216,
      "grad_norm": 0.157396599650383,
      "learning_rate": 4.6792314836035304e-06,
      "loss": 0.0315,
      "step": 2333
    },
    {
      "epoch": 1.7509377344336086,
      "grad_norm": 0.16064214706420898,
      "learning_rate": 4.651613632812413e-06,
      "loss": 0.0391,
      "step": 2334
    },
    {
      "epoch": 1.751687921980495,
      "grad_norm": 0.20322883129119873,
      "learning_rate": 4.624073550038399e-06,
      "loss": 0.0431,
      "step": 2335
    },
    {
      "epoch": 1.752438109527382,
      "grad_norm": 0.21442653238773346,
      "learning_rate": 4.596611282509989e-06,
      "loss": 0.0483,
      "step": 2336
    },
    {
      "epoch": 1.7531882970742685,
      "grad_norm": 0.16077186167240143,
      "learning_rate": 4.56922687732228e-06,
      "loss": 0.0501,
      "step": 2337
    },
    {
      "epoch": 1.7539384846211554,
      "grad_norm": 0.15106667578220367,
      "learning_rate": 4.5419203814368376e-06,
      "loss": 0.0357,
      "step": 2338
    },
    {
      "epoch": 1.754688672168042,
      "grad_norm": 0.1335969716310501,
      "learning_rate": 4.514691841681601e-06,
      "loss": 0.0464,
      "step": 2339
    },
    {
      "epoch": 1.7554388597149289,
      "grad_norm": 0.21206989884376526,
      "learning_rate": 4.487541304750848e-06,
      "loss": 0.0368,
      "step": 2340
    },
    {
      "epoch": 1.7561890472618154,
      "grad_norm": 0.1703171730041504,
      "learning_rate": 4.4604688172050605e-06,
      "loss": 0.0515,
      "step": 2341
    },
    {
      "epoch": 1.7569392348087023,
      "grad_norm": 0.26870593428611755,
      "learning_rate": 4.433474425470902e-06,
      "loss": 0.0706,
      "step": 2342
    },
    {
      "epoch": 1.7576894223555888,
      "grad_norm": 0.21271054446697235,
      "learning_rate": 4.406558175841097e-06,
      "loss": 0.0601,
      "step": 2343
    },
    {
      "epoch": 1.7584396099024757,
      "grad_norm": 0.19085806608200073,
      "learning_rate": 4.379720114474351e-06,
      "loss": 0.0365,
      "step": 2344
    },
    {
      "epoch": 1.7591897974493622,
      "grad_norm": 0.1780657321214676,
      "learning_rate": 4.352960287395303e-06,
      "loss": 0.0348,
      "step": 2345
    },
    {
      "epoch": 1.7599399849962492,
      "grad_norm": 0.14623548090457916,
      "learning_rate": 4.3262787404944165e-06,
      "loss": 0.0317,
      "step": 2346
    },
    {
      "epoch": 1.7606901725431356,
      "grad_norm": 0.13500511646270752,
      "learning_rate": 4.299675519527929e-06,
      "loss": 0.0256,
      "step": 2347
    },
    {
      "epoch": 1.7614403600900226,
      "grad_norm": 0.2227623611688614,
      "learning_rate": 4.273150670117743e-06,
      "loss": 0.0452,
      "step": 2348
    },
    {
      "epoch": 1.762190547636909,
      "grad_norm": 0.15160562098026276,
      "learning_rate": 4.246704237751342e-06,
      "loss": 0.0331,
      "step": 2349
    },
    {
      "epoch": 1.762940735183796,
      "grad_norm": 0.20232978463172913,
      "learning_rate": 4.220336267781777e-06,
      "loss": 0.0396,
      "step": 2350
    },
    {
      "epoch": 1.7636909227306825,
      "grad_norm": 0.1632014513015747,
      "learning_rate": 4.19404680542751e-06,
      "loss": 0.0444,
      "step": 2351
    },
    {
      "epoch": 1.7644411102775694,
      "grad_norm": 0.1549803912639618,
      "learning_rate": 4.167835895772382e-06,
      "loss": 0.0361,
      "step": 2352
    },
    {
      "epoch": 1.7651912978244562,
      "grad_norm": 0.24175381660461426,
      "learning_rate": 4.141703583765522e-06,
      "loss": 0.0434,
      "step": 2353
    },
    {
      "epoch": 1.7659414853713429,
      "grad_norm": 0.15828168392181396,
      "learning_rate": 4.11564991422127e-06,
      "loss": 0.0329,
      "step": 2354
    },
    {
      "epoch": 1.7666916729182296,
      "grad_norm": 0.26973676681518555,
      "learning_rate": 4.0896749318191095e-06,
      "loss": 0.0677,
      "step": 2355
    },
    {
      "epoch": 1.7674418604651163,
      "grad_norm": 0.2405744045972824,
      "learning_rate": 4.06377868110358e-06,
      "loss": 0.0429,
      "step": 2356
    },
    {
      "epoch": 1.768192048012003,
      "grad_norm": 0.1941184252500534,
      "learning_rate": 4.037961206484186e-06,
      "loss": 0.0471,
      "step": 2357
    },
    {
      "epoch": 1.7689422355588897,
      "grad_norm": 0.2016299068927765,
      "learning_rate": 4.0122225522353675e-06,
      "loss": 0.0427,
      "step": 2358
    },
    {
      "epoch": 1.7696924231057765,
      "grad_norm": 0.21357449889183044,
      "learning_rate": 3.986562762496376e-06,
      "loss": 0.0443,
      "step": 2359
    },
    {
      "epoch": 1.7704426106526632,
      "grad_norm": 0.279895156621933,
      "learning_rate": 3.9609818812712255e-06,
      "loss": 0.0583,
      "step": 2360
    },
    {
      "epoch": 1.77119279819955,
      "grad_norm": 0.132369726896286,
      "learning_rate": 3.935479952428611e-06,
      "loss": 0.0326,
      "step": 2361
    },
    {
      "epoch": 1.7719429857464366,
      "grad_norm": 0.20074398815631866,
      "learning_rate": 3.91005701970183e-06,
      "loss": 0.0426,
      "step": 2362
    },
    {
      "epoch": 1.7726931732933233,
      "grad_norm": 0.19614775478839874,
      "learning_rate": 3.8847131266886935e-06,
      "loss": 0.0303,
      "step": 2363
    },
    {
      "epoch": 1.77344336084021,
      "grad_norm": 0.14640925824642181,
      "learning_rate": 3.859448316851505e-06,
      "loss": 0.0429,
      "step": 2364
    },
    {
      "epoch": 1.7741935483870968,
      "grad_norm": 0.22066819667816162,
      "learning_rate": 3.834262633516916e-06,
      "loss": 0.0357,
      "step": 2365
    },
    {
      "epoch": 1.7749437359339835,
      "grad_norm": 0.17144808173179626,
      "learning_rate": 3.8091561198758897e-06,
      "loss": 0.0365,
      "step": 2366
    },
    {
      "epoch": 1.7756939234808702,
      "grad_norm": 0.16311363875865936,
      "learning_rate": 3.784128818983618e-06,
      "loss": 0.0312,
      "step": 2367
    },
    {
      "epoch": 1.776444111027757,
      "grad_norm": 0.14726579189300537,
      "learning_rate": 3.7591807737594743e-06,
      "loss": 0.033,
      "step": 2368
    },
    {
      "epoch": 1.7771942985746436,
      "grad_norm": 0.18850120902061462,
      "learning_rate": 3.734312026986897e-06,
      "loss": 0.0429,
      "step": 2369
    },
    {
      "epoch": 1.7779444861215303,
      "grad_norm": 0.19177739322185516,
      "learning_rate": 3.7095226213133272e-06,
      "loss": 0.059,
      "step": 2370
    },
    {
      "epoch": 1.7786946736684173,
      "grad_norm": 0.14402103424072266,
      "learning_rate": 3.6848125992501592e-06,
      "loss": 0.044,
      "step": 2371
    },
    {
      "epoch": 1.7794448612153038,
      "grad_norm": 0.18489781022071838,
      "learning_rate": 3.6601820031726517e-06,
      "loss": 0.0368,
      "step": 2372
    },
    {
      "epoch": 1.7801950487621907,
      "grad_norm": 0.1771378219127655,
      "learning_rate": 3.6356308753198454e-06,
      "loss": 0.0348,
      "step": 2373
    },
    {
      "epoch": 1.7809452363090772,
      "grad_norm": 0.1410515457391739,
      "learning_rate": 3.6111592577945217e-06,
      "loss": 0.032,
      "step": 2374
    },
    {
      "epoch": 1.7816954238559641,
      "grad_norm": 0.18638338148593903,
      "learning_rate": 3.586767192563073e-06,
      "loss": 0.0566,
      "step": 2375
    },
    {
      "epoch": 1.7824456114028506,
      "grad_norm": 0.21808239817619324,
      "learning_rate": 3.562454721455505e-06,
      "loss": 0.0447,
      "step": 2376
    },
    {
      "epoch": 1.7831957989497376,
      "grad_norm": 0.17375057935714722,
      "learning_rate": 3.538221886165299e-06,
      "loss": 0.0377,
      "step": 2377
    },
    {
      "epoch": 1.783945986496624,
      "grad_norm": 0.1853272020816803,
      "learning_rate": 3.514068728249398e-06,
      "loss": 0.0344,
      "step": 2378
    },
    {
      "epoch": 1.784696174043511,
      "grad_norm": 0.2159195989370346,
      "learning_rate": 3.489995289128073e-06,
      "loss": 0.0672,
      "step": 2379
    },
    {
      "epoch": 1.7854463615903975,
      "grad_norm": 0.1763429343700409,
      "learning_rate": 3.4660016100849126e-06,
      "loss": 0.0539,
      "step": 2380
    },
    {
      "epoch": 1.7861965491372844,
      "grad_norm": 0.13722042739391327,
      "learning_rate": 3.442087732266697e-06,
      "loss": 0.039,
      "step": 2381
    },
    {
      "epoch": 1.786946736684171,
      "grad_norm": 0.180751770734787,
      "learning_rate": 3.418253696683399e-06,
      "loss": 0.0433,
      "step": 2382
    },
    {
      "epoch": 1.7876969242310579,
      "grad_norm": 0.23207427561283112,
      "learning_rate": 3.3944995442080185e-06,
      "loss": 0.043,
      "step": 2383
    },
    {
      "epoch": 1.7884471117779444,
      "grad_norm": 0.1683093160390854,
      "learning_rate": 3.3708253155766033e-06,
      "loss": 0.0429,
      "step": 2384
    },
    {
      "epoch": 1.7891972993248313,
      "grad_norm": 0.19512560963630676,
      "learning_rate": 3.347231051388117e-06,
      "loss": 0.0385,
      "step": 2385
    },
    {
      "epoch": 1.7899474868717178,
      "grad_norm": 0.152205228805542,
      "learning_rate": 3.323716792104403e-06,
      "loss": 0.0306,
      "step": 2386
    },
    {
      "epoch": 1.7906976744186047,
      "grad_norm": 0.18546339869499207,
      "learning_rate": 3.3002825780500957e-06,
      "loss": 0.0524,
      "step": 2387
    },
    {
      "epoch": 1.7914478619654912,
      "grad_norm": 0.14806154370307922,
      "learning_rate": 3.276928449412564e-06,
      "loss": 0.0349,
      "step": 2388
    },
    {
      "epoch": 1.7921980495123782,
      "grad_norm": 0.2536125183105469,
      "learning_rate": 3.253654446241844e-06,
      "loss": 0.0513,
      "step": 2389
    },
    {
      "epoch": 1.7929482370592649,
      "grad_norm": 0.16887833178043365,
      "learning_rate": 3.2304606084505585e-06,
      "loss": 0.032,
      "step": 2390
    },
    {
      "epoch": 1.7936984246061516,
      "grad_norm": 0.2701931297779083,
      "learning_rate": 3.2073469758138577e-06,
      "loss": 0.0623,
      "step": 2391
    },
    {
      "epoch": 1.7944486121530383,
      "grad_norm": 0.14021989703178406,
      "learning_rate": 3.18431358796934e-06,
      "loss": 0.0415,
      "step": 2392
    },
    {
      "epoch": 1.795198799699925,
      "grad_norm": 0.14851094782352448,
      "learning_rate": 3.161360484416992e-06,
      "loss": 0.0322,
      "step": 2393
    },
    {
      "epoch": 1.7959489872468117,
      "grad_norm": 0.13025152683258057,
      "learning_rate": 3.1384877045191384e-06,
      "loss": 0.0304,
      "step": 2394
    },
    {
      "epoch": 1.7966991747936985,
      "grad_norm": 0.156709685921669,
      "learning_rate": 3.1156952875003365e-06,
      "loss": 0.0383,
      "step": 2395
    },
    {
      "epoch": 1.7974493623405852,
      "grad_norm": 0.12184840440750122,
      "learning_rate": 3.0929832724473416e-06,
      "loss": 0.0297,
      "step": 2396
    },
    {
      "epoch": 1.7981995498874719,
      "grad_norm": 0.20353339612483978,
      "learning_rate": 3.0703516983090207e-06,
      "loss": 0.0351,
      "step": 2397
    },
    {
      "epoch": 1.7989497374343586,
      "grad_norm": 0.26167547702789307,
      "learning_rate": 3.0478006038962947e-06,
      "loss": 0.0587,
      "step": 2398
    },
    {
      "epoch": 1.7996999249812453,
      "grad_norm": 0.1540125608444214,
      "learning_rate": 3.0253300278820783e-06,
      "loss": 0.0255,
      "step": 2399
    },
    {
      "epoch": 1.800450112528132,
      "grad_norm": 0.1433778554201126,
      "learning_rate": 3.002940008801186e-06,
      "loss": 0.0317,
      "step": 2400
    },
    {
      "epoch": 1.800450112528132,
      "eval_loss": 0.06825650483369827,
      "eval_runtime": 2.6517,
      "eval_samples_per_second": 20.364,
      "eval_steps_per_second": 5.28,
      "step": 2400
    },
    {
      "epoch": 1.8012003000750187,
      "grad_norm": 0.22410492599010468,
      "learning_rate": 2.9806305850502923e-06,
      "loss": 0.0542,
      "step": 2401
    },
    {
      "epoch": 1.8019504876219055,
      "grad_norm": 0.18377956748008728,
      "learning_rate": 2.9584017948878717e-06,
      "loss": 0.0591,
      "step": 2402
    },
    {
      "epoch": 1.8027006751687922,
      "grad_norm": 0.24742087721824646,
      "learning_rate": 2.9362536764341085e-06,
      "loss": 0.0444,
      "step": 2403
    },
    {
      "epoch": 1.803450862715679,
      "grad_norm": 0.25326278805732727,
      "learning_rate": 2.9141862676708486e-06,
      "loss": 0.1,
      "step": 2404
    },
    {
      "epoch": 1.8042010502625656,
      "grad_norm": 0.2263568788766861,
      "learning_rate": 2.8921996064415147e-06,
      "loss": 0.046,
      "step": 2405
    },
    {
      "epoch": 1.8049512378094523,
      "grad_norm": 0.2560195326805115,
      "learning_rate": 2.870293730451068e-06,
      "loss": 0.0505,
      "step": 2406
    },
    {
      "epoch": 1.805701425356339,
      "grad_norm": 0.1759989708662033,
      "learning_rate": 2.8484686772659308e-06,
      "loss": 0.0553,
      "step": 2407
    },
    {
      "epoch": 1.8064516129032258,
      "grad_norm": 0.18385818600654602,
      "learning_rate": 2.826724484313925e-06,
      "loss": 0.0432,
      "step": 2408
    },
    {
      "epoch": 1.8072018004501125,
      "grad_norm": 0.16650594770908356,
      "learning_rate": 2.8050611888841947e-06,
      "loss": 0.0517,
      "step": 2409
    },
    {
      "epoch": 1.8079519879969994,
      "grad_norm": 0.13436725735664368,
      "learning_rate": 2.7834788281271616e-06,
      "loss": 0.0384,
      "step": 2410
    },
    {
      "epoch": 1.808702175543886,
      "grad_norm": 0.18747705221176147,
      "learning_rate": 2.7619774390544473e-06,
      "loss": 0.0517,
      "step": 2411
    },
    {
      "epoch": 1.8094523630907728,
      "grad_norm": 0.13398858904838562,
      "learning_rate": 2.740557058538823e-06,
      "loss": 0.0302,
      "step": 2412
    },
    {
      "epoch": 1.8102025506376593,
      "grad_norm": 0.16094616055488586,
      "learning_rate": 2.7192177233141215e-06,
      "loss": 0.04,
      "step": 2413
    },
    {
      "epoch": 1.8109527381845463,
      "grad_norm": 0.2032438963651657,
      "learning_rate": 2.697959469975203e-06,
      "loss": 0.0617,
      "step": 2414
    },
    {
      "epoch": 1.8117029257314328,
      "grad_norm": 0.19508475065231323,
      "learning_rate": 2.6767823349778843e-06,
      "loss": 0.0347,
      "step": 2415
    },
    {
      "epoch": 1.8124531132783197,
      "grad_norm": 0.16064900159835815,
      "learning_rate": 2.65568635463887e-06,
      "loss": 0.0547,
      "step": 2416
    },
    {
      "epoch": 1.8132033008252062,
      "grad_norm": 0.15190844237804413,
      "learning_rate": 2.634671565135677e-06,
      "loss": 0.0505,
      "step": 2417
    },
    {
      "epoch": 1.8139534883720931,
      "grad_norm": 0.19380758702754974,
      "learning_rate": 2.613738002506605e-06,
      "loss": 0.0381,
      "step": 2418
    },
    {
      "epoch": 1.8147036759189796,
      "grad_norm": 0.19494427740573883,
      "learning_rate": 2.592885702650655e-06,
      "loss": 0.0454,
      "step": 2419
    },
    {
      "epoch": 1.8154538634658666,
      "grad_norm": 0.21446380019187927,
      "learning_rate": 2.572114701327466e-06,
      "loss": 0.0406,
      "step": 2420
    },
    {
      "epoch": 1.816204051012753,
      "grad_norm": 0.18711164593696594,
      "learning_rate": 2.551425034157262e-06,
      "loss": 0.036,
      "step": 2421
    },
    {
      "epoch": 1.81695423855964,
      "grad_norm": 0.18074631690979004,
      "learning_rate": 2.5308167366207724e-06,
      "loss": 0.0567,
      "step": 2422
    },
    {
      "epoch": 1.8177044261065265,
      "grad_norm": 0.13808584213256836,
      "learning_rate": 2.510289844059216e-06,
      "loss": 0.0336,
      "step": 2423
    },
    {
      "epoch": 1.8184546136534134,
      "grad_norm": 0.13104617595672607,
      "learning_rate": 2.48984439167419e-06,
      "loss": 0.0345,
      "step": 2424
    },
    {
      "epoch": 1.8192048012003,
      "grad_norm": 0.1713109016418457,
      "learning_rate": 2.4694804145276305e-06,
      "loss": 0.0499,
      "step": 2425
    },
    {
      "epoch": 1.8199549887471869,
      "grad_norm": 0.21308591961860657,
      "learning_rate": 2.449197947541737e-06,
      "loss": 0.0423,
      "step": 2426
    },
    {
      "epoch": 1.8207051762940734,
      "grad_norm": 0.27069994807243347,
      "learning_rate": 2.4289970254989635e-06,
      "loss": 0.0557,
      "step": 2427
    },
    {
      "epoch": 1.8214553638409603,
      "grad_norm": 0.2805314362049103,
      "learning_rate": 2.408877683041888e-06,
      "loss": 0.051,
      "step": 2428
    },
    {
      "epoch": 1.822205551387847,
      "grad_norm": 0.15674597024917603,
      "learning_rate": 2.388839954673222e-06,
      "loss": 0.0334,
      "step": 2429
    },
    {
      "epoch": 1.8229557389347337,
      "grad_norm": 0.17512424290180206,
      "learning_rate": 2.3688838747556674e-06,
      "loss": 0.0473,
      "step": 2430
    },
    {
      "epoch": 1.8237059264816204,
      "grad_norm": 0.15832626819610596,
      "learning_rate": 2.3490094775119597e-06,
      "loss": 0.0324,
      "step": 2431
    },
    {
      "epoch": 1.8244561140285072,
      "grad_norm": 0.1268201768398285,
      "learning_rate": 2.3292167970247193e-06,
      "loss": 0.0212,
      "step": 2432
    },
    {
      "epoch": 1.8252063015753939,
      "grad_norm": 0.2088364213705063,
      "learning_rate": 2.30950586723645e-06,
      "loss": 0.0473,
      "step": 2433
    },
    {
      "epoch": 1.8259564891222806,
      "grad_norm": 0.14922069013118744,
      "learning_rate": 2.2898767219494634e-06,
      "loss": 0.0403,
      "step": 2434
    },
    {
      "epoch": 1.8267066766691673,
      "grad_norm": 0.2907104194164276,
      "learning_rate": 2.270329394825793e-06,
      "loss": 0.0745,
      "step": 2435
    },
    {
      "epoch": 1.827456864216054,
      "grad_norm": 0.17265377938747406,
      "learning_rate": 2.2508639193871805e-06,
      "loss": 0.0498,
      "step": 2436
    },
    {
      "epoch": 1.8282070517629407,
      "grad_norm": 0.18557190895080566,
      "learning_rate": 2.2314803290150287e-06,
      "loss": 0.0367,
      "step": 2437
    },
    {
      "epoch": 1.8289572393098275,
      "grad_norm": 0.18080484867095947,
      "learning_rate": 2.2121786569502535e-06,
      "loss": 0.0378,
      "step": 2438
    },
    {
      "epoch": 1.8297074268567142,
      "grad_norm": 0.12578532099723816,
      "learning_rate": 2.192958936293338e-06,
      "loss": 0.0325,
      "step": 2439
    },
    {
      "epoch": 1.8304576144036009,
      "grad_norm": 0.19788827002048492,
      "learning_rate": 2.1738212000042e-06,
      "loss": 0.0461,
      "step": 2440
    },
    {
      "epoch": 1.8312078019504876,
      "grad_norm": 0.22011855244636536,
      "learning_rate": 2.1547654809021877e-06,
      "loss": 0.036,
      "step": 2441
    },
    {
      "epoch": 1.8319579894973743,
      "grad_norm": 0.14238829910755157,
      "learning_rate": 2.135791811665977e-06,
      "loss": 0.0434,
      "step": 2442
    },
    {
      "epoch": 1.832708177044261,
      "grad_norm": 0.14220482110977173,
      "learning_rate": 2.1169002248335346e-06,
      "loss": 0.031,
      "step": 2443
    },
    {
      "epoch": 1.8334583645911477,
      "grad_norm": 0.1597825586795807,
      "learning_rate": 2.098090752802073e-06,
      "loss": 0.0429,
      "step": 2444
    },
    {
      "epoch": 1.8342085521380345,
      "grad_norm": 0.19599536061286926,
      "learning_rate": 2.0793634278279907e-06,
      "loss": 0.0487,
      "step": 2445
    },
    {
      "epoch": 1.8349587396849212,
      "grad_norm": 0.27415987849235535,
      "learning_rate": 2.0607182820268133e-06,
      "loss": 0.0499,
      "step": 2446
    },
    {
      "epoch": 1.835708927231808,
      "grad_norm": 0.1647443026304245,
      "learning_rate": 2.042155347373109e-06,
      "loss": 0.0412,
      "step": 2447
    },
    {
      "epoch": 1.8364591147786946,
      "grad_norm": 0.22096386551856995,
      "learning_rate": 2.023674655700497e-06,
      "loss": 0.0383,
      "step": 2448
    },
    {
      "epoch": 1.8372093023255816,
      "grad_norm": 0.23002421855926514,
      "learning_rate": 2.0052762387015424e-06,
      "loss": 0.0431,
      "step": 2449
    },
    {
      "epoch": 1.837959489872468,
      "grad_norm": 0.18029820919036865,
      "learning_rate": 1.986960127927717e-06,
      "loss": 0.0334,
      "step": 2450
    },
    {
      "epoch": 1.838709677419355,
      "grad_norm": 0.20091702044010162,
      "learning_rate": 1.9687263547893407e-06,
      "loss": 0.0469,
      "step": 2451
    },
    {
      "epoch": 1.8394598649662415,
      "grad_norm": 0.2061494141817093,
      "learning_rate": 1.9505749505555503e-06,
      "loss": 0.0374,
      "step": 2452
    },
    {
      "epoch": 1.8402100525131284,
      "grad_norm": 0.20793215930461884,
      "learning_rate": 1.932505946354213e-06,
      "loss": 0.0526,
      "step": 2453
    },
    {
      "epoch": 1.840960240060015,
      "grad_norm": 0.19572290778160095,
      "learning_rate": 1.9145193731718858e-06,
      "loss": 0.04,
      "step": 2454
    },
    {
      "epoch": 1.8417104276069018,
      "grad_norm": 0.3863771855831146,
      "learning_rate": 1.8966152618537846e-06,
      "loss": 0.0791,
      "step": 2455
    },
    {
      "epoch": 1.8424606151537883,
      "grad_norm": 0.22557200491428375,
      "learning_rate": 1.8787936431036824e-06,
      "loss": 0.0407,
      "step": 2456
    },
    {
      "epoch": 1.8432108027006753,
      "grad_norm": 0.2332926243543625,
      "learning_rate": 1.8610545474839036e-06,
      "loss": 0.0614,
      "step": 2457
    },
    {
      "epoch": 1.8439609902475618,
      "grad_norm": 0.1692853420972824,
      "learning_rate": 1.8433980054152533e-06,
      "loss": 0.054,
      "step": 2458
    },
    {
      "epoch": 1.8447111777944487,
      "grad_norm": 0.1144590973854065,
      "learning_rate": 1.8258240471769662e-06,
      "loss": 0.0306,
      "step": 2459
    },
    {
      "epoch": 1.8454613653413352,
      "grad_norm": 0.2206297367811203,
      "learning_rate": 1.8083327029066399e-06,
      "loss": 0.0582,
      "step": 2460
    },
    {
      "epoch": 1.8462115528882221,
      "grad_norm": 0.20330913364887238,
      "learning_rate": 1.7909240026002138e-06,
      "loss": 0.0565,
      "step": 2461
    },
    {
      "epoch": 1.8469617404351086,
      "grad_norm": 0.15369729697704315,
      "learning_rate": 1.773597976111896e-06,
      "loss": 0.0292,
      "step": 2462
    },
    {
      "epoch": 1.8477119279819956,
      "grad_norm": 0.1898067593574524,
      "learning_rate": 1.7563546531541132e-06,
      "loss": 0.0451,
      "step": 2463
    },
    {
      "epoch": 1.848462115528882,
      "grad_norm": 0.14799977838993073,
      "learning_rate": 1.7391940632974667e-06,
      "loss": 0.0399,
      "step": 2464
    },
    {
      "epoch": 1.849212303075769,
      "grad_norm": 0.20645496249198914,
      "learning_rate": 1.7221162359706776e-06,
      "loss": 0.0455,
      "step": 2465
    },
    {
      "epoch": 1.8499624906226555,
      "grad_norm": 0.22738315165042877,
      "learning_rate": 1.705121200460541e-06,
      "loss": 0.0611,
      "step": 2466
    },
    {
      "epoch": 1.8507126781695424,
      "grad_norm": 0.24919851124286652,
      "learning_rate": 1.6882089859118766e-06,
      "loss": 0.0677,
      "step": 2467
    },
    {
      "epoch": 1.8514628657164292,
      "grad_norm": 0.18418778479099274,
      "learning_rate": 1.6713796213274457e-06,
      "loss": 0.0346,
      "step": 2468
    },
    {
      "epoch": 1.8522130532633159,
      "grad_norm": 0.20278531312942505,
      "learning_rate": 1.6546331355679623e-06,
      "loss": 0.0409,
      "step": 2469
    },
    {
      "epoch": 1.8529632408102026,
      "grad_norm": 0.14719854295253754,
      "learning_rate": 1.6379695573520093e-06,
      "loss": 0.0343,
      "step": 2470
    },
    {
      "epoch": 1.8537134283570893,
      "grad_norm": 0.210566908121109,
      "learning_rate": 1.621388915255967e-06,
      "loss": 0.0559,
      "step": 2471
    },
    {
      "epoch": 1.854463615903976,
      "grad_norm": 0.14764787256717682,
      "learning_rate": 1.604891237714018e-06,
      "loss": 0.0371,
      "step": 2472
    },
    {
      "epoch": 1.8552138034508627,
      "grad_norm": 0.14555113017559052,
      "learning_rate": 1.5884765530180478e-06,
      "loss": 0.0334,
      "step": 2473
    },
    {
      "epoch": 1.8559639909977494,
      "grad_norm": 0.17617011070251465,
      "learning_rate": 1.5721448893176228e-06,
      "loss": 0.0324,
      "step": 2474
    },
    {
      "epoch": 1.8567141785446362,
      "grad_norm": 0.18802230060100555,
      "learning_rate": 1.5558962746199335e-06,
      "loss": 0.0385,
      "step": 2475
    },
    {
      "epoch": 1.8574643660915229,
      "grad_norm": 0.1850004941225052,
      "learning_rate": 1.5397307367897684e-06,
      "loss": 0.0481,
      "step": 2476
    },
    {
      "epoch": 1.8582145536384096,
      "grad_norm": 0.19387805461883545,
      "learning_rate": 1.5236483035494297e-06,
      "loss": 0.0421,
      "step": 2477
    },
    {
      "epoch": 1.8589647411852963,
      "grad_norm": 0.1989627182483673,
      "learning_rate": 1.5076490024786893e-06,
      "loss": 0.0485,
      "step": 2478
    },
    {
      "epoch": 1.859714928732183,
      "grad_norm": 0.17542366683483124,
      "learning_rate": 1.4917328610147885e-06,
      "loss": 0.0465,
      "step": 2479
    },
    {
      "epoch": 1.8604651162790697,
      "grad_norm": 0.24785301089286804,
      "learning_rate": 1.4758999064523493e-06,
      "loss": 0.0493,
      "step": 2480
    },
    {
      "epoch": 1.8612153038259565,
      "grad_norm": 0.18737287819385529,
      "learning_rate": 1.4601501659433137e-06,
      "loss": 0.0435,
      "step": 2481
    },
    {
      "epoch": 1.8619654913728432,
      "grad_norm": 0.19134843349456787,
      "learning_rate": 1.444483666496943e-06,
      "loss": 0.0556,
      "step": 2482
    },
    {
      "epoch": 1.86271567891973,
      "grad_norm": 0.1329682469367981,
      "learning_rate": 1.4289004349797409e-06,
      "loss": 0.0333,
      "step": 2483
    },
    {
      "epoch": 1.8634658664666166,
      "grad_norm": 0.18284755945205688,
      "learning_rate": 1.4134004981154137e-06,
      "loss": 0.0554,
      "step": 2484
    },
    {
      "epoch": 1.8642160540135033,
      "grad_norm": 0.21650445461273193,
      "learning_rate": 1.3979838824848378e-06,
      "loss": 0.061,
      "step": 2485
    },
    {
      "epoch": 1.8649662415603903,
      "grad_norm": 0.22950206696987152,
      "learning_rate": 1.382650614525971e-06,
      "loss": 0.0483,
      "step": 2486
    },
    {
      "epoch": 1.8657164291072768,
      "grad_norm": 0.20865564048290253,
      "learning_rate": 1.3674007205338678e-06,
      "loss": 0.0511,
      "step": 2487
    },
    {
      "epoch": 1.8664666166541637,
      "grad_norm": 0.16846178472042084,
      "learning_rate": 1.3522342266605925e-06,
      "loss": 0.037,
      "step": 2488
    },
    {
      "epoch": 1.8672168042010502,
      "grad_norm": 0.19286170601844788,
      "learning_rate": 1.3371511589152008e-06,
      "loss": 0.0417,
      "step": 2489
    },
    {
      "epoch": 1.8679669917479371,
      "grad_norm": 0.12854991853237152,
      "learning_rate": 1.3221515431636522e-06,
      "loss": 0.0246,
      "step": 2490
    },
    {
      "epoch": 1.8687171792948236,
      "grad_norm": 0.2654399275779724,
      "learning_rate": 1.307235405128815e-06,
      "loss": 0.0648,
      "step": 2491
    },
    {
      "epoch": 1.8694673668417106,
      "grad_norm": 0.24505434930324554,
      "learning_rate": 1.292402770390394e-06,
      "loss": 0.0498,
      "step": 2492
    },
    {
      "epoch": 1.870217554388597,
      "grad_norm": 0.14539310336112976,
      "learning_rate": 1.2776536643849145e-06,
      "loss": 0.0391,
      "step": 2493
    },
    {
      "epoch": 1.870967741935484,
      "grad_norm": 0.15054713189601898,
      "learning_rate": 1.2629881124056274e-06,
      "loss": 0.0292,
      "step": 2494
    },
    {
      "epoch": 1.8717179294823705,
      "grad_norm": 0.228153795003891,
      "learning_rate": 1.2484061396025038e-06,
      "loss": 0.047,
      "step": 2495
    },
    {
      "epoch": 1.8724681170292574,
      "grad_norm": 0.16526199877262115,
      "learning_rate": 1.2339077709822067e-06,
      "loss": 0.0393,
      "step": 2496
    },
    {
      "epoch": 1.873218304576144,
      "grad_norm": 0.19541572034358978,
      "learning_rate": 1.2194930314080032e-06,
      "loss": 0.0607,
      "step": 2497
    },
    {
      "epoch": 1.8739684921230308,
      "grad_norm": 0.39976009726524353,
      "learning_rate": 1.2051619455997476e-06,
      "loss": 0.0486,
      "step": 2498
    },
    {
      "epoch": 1.8747186796699173,
      "grad_norm": 0.5087877511978149,
      "learning_rate": 1.1909145381338472e-06,
      "loss": 0.0762,
      "step": 2499
    },
    {
      "epoch": 1.8754688672168043,
      "grad_norm": 0.2975837290287018,
      "learning_rate": 1.1767508334431964e-06,
      "loss": 0.0469,
      "step": 2500
    },
    {
      "epoch": 1.8762190547636908,
      "grad_norm": 0.1391455978155136,
      "learning_rate": 1.1626708558171606e-06,
      "loss": 0.0439,
      "step": 2501
    },
    {
      "epoch": 1.8769692423105777,
      "grad_norm": 0.19531936943531036,
      "learning_rate": 1.1486746294015193e-06,
      "loss": 0.0499,
      "step": 2502
    },
    {
      "epoch": 1.8777194298574642,
      "grad_norm": 0.1554703414440155,
      "learning_rate": 1.134762178198412e-06,
      "loss": 0.0311,
      "step": 2503
    },
    {
      "epoch": 1.8784696174043511,
      "grad_norm": 0.22662873566150665,
      "learning_rate": 1.1209335260663256e-06,
      "loss": 0.0577,
      "step": 2504
    },
    {
      "epoch": 1.8792198049512379,
      "grad_norm": 0.17107880115509033,
      "learning_rate": 1.1071886967200352e-06,
      "loss": 0.0569,
      "step": 2505
    },
    {
      "epoch": 1.8799699924981246,
      "grad_norm": 0.1529870182275772,
      "learning_rate": 1.0935277137305744e-06,
      "loss": 0.0281,
      "step": 2506
    },
    {
      "epoch": 1.8807201800450113,
      "grad_norm": 0.17268817126750946,
      "learning_rate": 1.0799506005251814e-06,
      "loss": 0.0457,
      "step": 2507
    },
    {
      "epoch": 1.881470367591898,
      "grad_norm": 0.198935866355896,
      "learning_rate": 1.06645738038727e-06,
      "loss": 0.0576,
      "step": 2508
    },
    {
      "epoch": 1.8822205551387847,
      "grad_norm": 0.17006203532218933,
      "learning_rate": 1.053048076456381e-06,
      "loss": 0.0376,
      "step": 2509
    },
    {
      "epoch": 1.8829707426856714,
      "grad_norm": 0.17668931186199188,
      "learning_rate": 1.0397227117281528e-06,
      "loss": 0.0359,
      "step": 2510
    },
    {
      "epoch": 1.8837209302325582,
      "grad_norm": 0.17520001530647278,
      "learning_rate": 1.0264813090542725e-06,
      "loss": 0.0323,
      "step": 2511
    },
    {
      "epoch": 1.8844711177794449,
      "grad_norm": 0.1917439103126526,
      "learning_rate": 1.0133238911424426e-06,
      "loss": 0.0444,
      "step": 2512
    },
    {
      "epoch": 1.8852213053263316,
      "grad_norm": 0.21084263920783997,
      "learning_rate": 1.0002504805563362e-06,
      "loss": 0.0282,
      "step": 2513
    },
    {
      "epoch": 1.8859714928732183,
      "grad_norm": 0.12074773758649826,
      "learning_rate": 9.872610997155695e-07,
      "loss": 0.025,
      "step": 2514
    },
    {
      "epoch": 1.886721680420105,
      "grad_norm": 0.143777534365654,
      "learning_rate": 9.743557708956575e-07,
      "loss": 0.03,
      "step": 2515
    },
    {
      "epoch": 1.8874718679669917,
      "grad_norm": 0.16410160064697266,
      "learning_rate": 9.615345162279521e-07,
      "loss": 0.0295,
      "step": 2516
    },
    {
      "epoch": 1.8882220555138785,
      "grad_norm": 0.3826543688774109,
      "learning_rate": 9.48797357699649e-07,
      "loss": 0.0719,
      "step": 2517
    },
    {
      "epoch": 1.8889722430607652,
      "grad_norm": 0.17290951311588287,
      "learning_rate": 9.361443171537254e-07,
      "loss": 0.0338,
      "step": 2518
    },
    {
      "epoch": 1.8897224306076519,
      "grad_norm": 0.18427728116512299,
      "learning_rate": 9.235754162889021e-07,
      "loss": 0.0512,
      "step": 2519
    },
    {
      "epoch": 1.8904726181545386,
      "grad_norm": 0.2464424967765808,
      "learning_rate": 9.110906766595872e-07,
      "loss": 0.0512,
      "step": 2520
    },
    {
      "epoch": 1.8912228057014253,
      "grad_norm": 0.27707773447036743,
      "learning_rate": 8.986901196759046e-07,
      "loss": 0.0741,
      "step": 2521
    },
    {
      "epoch": 1.891972993248312,
      "grad_norm": 0.09056626260280609,
      "learning_rate": 8.863737666035765e-07,
      "loss": 0.0228,
      "step": 2522
    },
    {
      "epoch": 1.8927231807951987,
      "grad_norm": 0.25095298886299133,
      "learning_rate": 8.741416385639412e-07,
      "loss": 0.0811,
      "step": 2523
    },
    {
      "epoch": 1.8934733683420855,
      "grad_norm": 0.1794692426919937,
      "learning_rate": 8.619937565338854e-07,
      "loss": 0.0416,
      "step": 2524
    },
    {
      "epoch": 1.8942235558889724,
      "grad_norm": 0.16589999198913574,
      "learning_rate": 8.499301413458338e-07,
      "loss": 0.0418,
      "step": 2525
    },
    {
      "epoch": 1.894973743435859,
      "grad_norm": 0.2033083289861679,
      "learning_rate": 8.37950813687699e-07,
      "loss": 0.0453,
      "step": 2526
    },
    {
      "epoch": 1.8957239309827458,
      "grad_norm": 0.24125005304813385,
      "learning_rate": 8.26055794102848e-07,
      "loss": 0.0712,
      "step": 2527
    },
    {
      "epoch": 1.8964741185296323,
      "grad_norm": 0.1965634971857071,
      "learning_rate": 8.142451029900744e-07,
      "loss": 0.0377,
      "step": 2528
    },
    {
      "epoch": 1.8972243060765193,
      "grad_norm": 0.1771164983510971,
      "learning_rate": 8.025187606035434e-07,
      "loss": 0.0465,
      "step": 2529
    },
    {
      "epoch": 1.8979744936234058,
      "grad_norm": 0.2019067406654358,
      "learning_rate": 7.908767870527745e-07,
      "loss": 0.049,
      "step": 2530
    },
    {
      "epoch": 1.8987246811702927,
      "grad_norm": 0.1835569441318512,
      "learning_rate": 7.793192023026142e-07,
      "loss": 0.038,
      "step": 2531
    },
    {
      "epoch": 1.8994748687171792,
      "grad_norm": 0.2343943566083908,
      "learning_rate": 7.678460261731801e-07,
      "loss": 0.0752,
      "step": 2532
    },
    {
      "epoch": 1.9002250562640661,
      "grad_norm": 0.2090568095445633,
      "learning_rate": 7.564572783398339e-07,
      "loss": 0.0338,
      "step": 2533
    },
    {
      "epoch": 1.9009752438109526,
      "grad_norm": 0.18029850721359253,
      "learning_rate": 7.451529783331523e-07,
      "loss": 0.0318,
      "step": 2534
    },
    {
      "epoch": 1.9017254313578396,
      "grad_norm": 0.2427482008934021,
      "learning_rate": 7.339331455389175e-07,
      "loss": 0.0441,
      "step": 2535
    },
    {
      "epoch": 1.902475618904726,
      "grad_norm": 0.1632808893918991,
      "learning_rate": 7.227977991980217e-07,
      "loss": 0.0463,
      "step": 2536
    },
    {
      "epoch": 1.903225806451613,
      "grad_norm": 0.1503705233335495,
      "learning_rate": 7.117469584064895e-07,
      "loss": 0.0347,
      "step": 2537
    },
    {
      "epoch": 1.9039759939984995,
      "grad_norm": 0.19075416028499603,
      "learning_rate": 7.007806421154284e-07,
      "loss": 0.0268,
      "step": 2538
    },
    {
      "epoch": 1.9047261815453864,
      "grad_norm": 0.18026676774024963,
      "learning_rate": 6.898988691309893e-07,
      "loss": 0.0393,
      "step": 2539
    },
    {
      "epoch": 1.905476369092273,
      "grad_norm": 0.14918452501296997,
      "learning_rate": 6.791016581143395e-07,
      "loss": 0.0342,
      "step": 2540
    },
    {
      "epoch": 1.9062265566391599,
      "grad_norm": 0.2030014842748642,
      "learning_rate": 6.683890275816341e-07,
      "loss": 0.0436,
      "step": 2541
    },
    {
      "epoch": 1.9069767441860463,
      "grad_norm": 0.23770801723003387,
      "learning_rate": 6.577609959039776e-07,
      "loss": 0.0575,
      "step": 2542
    },
    {
      "epoch": 1.9077269317329333,
      "grad_norm": 0.2590319812297821,
      "learning_rate": 6.472175813074022e-07,
      "loss": 0.0656,
      "step": 2543
    },
    {
      "epoch": 1.90847711927982,
      "grad_norm": 0.14560408890247345,
      "learning_rate": 6.367588018728166e-07,
      "loss": 0.0402,
      "step": 2544
    },
    {
      "epoch": 1.9092273068267067,
      "grad_norm": 0.14348311722278595,
      "learning_rate": 6.263846755360126e-07,
      "loss": 0.0282,
      "step": 2545
    },
    {
      "epoch": 1.9099774943735934,
      "grad_norm": 0.1561354547739029,
      "learning_rate": 6.16095220087587e-07,
      "loss": 0.0336,
      "step": 2546
    },
    {
      "epoch": 1.9107276819204801,
      "grad_norm": 0.14579859375953674,
      "learning_rate": 6.05890453172936e-07,
      "loss": 0.0324,
      "step": 2547
    },
    {
      "epoch": 1.9114778694673669,
      "grad_norm": 0.20945526659488678,
      "learning_rate": 5.957703922922386e-07,
      "loss": 0.0547,
      "step": 2548
    },
    {
      "epoch": 1.9122280570142536,
      "grad_norm": 0.14774468541145325,
      "learning_rate": 5.857350548004015e-07,
      "loss": 0.0306,
      "step": 2549
    },
    {
      "epoch": 1.9129782445611403,
      "grad_norm": 0.1546287089586258,
      "learning_rate": 5.757844579070359e-07,
      "loss": 0.0278,
      "step": 2550
    },
    {
      "epoch": 1.913728432108027,
      "grad_norm": 0.13874396681785583,
      "learning_rate": 5.65918618676442e-07,
      "loss": 0.0287,
      "step": 2551
    },
    {
      "epoch": 1.9144786196549137,
      "grad_norm": 0.26407912373542786,
      "learning_rate": 5.561375540275581e-07,
      "loss": 0.0651,
      "step": 2552
    },
    {
      "epoch": 1.9152288072018004,
      "grad_norm": 0.19224029779434204,
      "learning_rate": 5.464412807339558e-07,
      "loss": 0.0499,
      "step": 2553
    },
    {
      "epoch": 1.9159789947486872,
      "grad_norm": 0.20610445737838745,
      "learning_rate": 5.368298154237727e-07,
      "loss": 0.0553,
      "step": 2554
    },
    {
      "epoch": 1.9167291822955739,
      "grad_norm": 0.19327408075332642,
      "learning_rate": 5.273031745797352e-07,
      "loss": 0.0324,
      "step": 2555
    },
    {
      "epoch": 1.9174793698424606,
      "grad_norm": 0.17420758306980133,
      "learning_rate": 5.17861374539097e-07,
      "loss": 0.0289,
      "step": 2556
    },
    {
      "epoch": 1.9182295573893473,
      "grad_norm": 0.2689003646373749,
      "learning_rate": 5.085044314936116e-07,
      "loss": 0.0423,
      "step": 2557
    },
    {
      "epoch": 1.918979744936234,
      "grad_norm": 0.1577097475528717,
      "learning_rate": 4.992323614895156e-07,
      "loss": 0.049,
      "step": 2558
    },
    {
      "epoch": 1.9197299324831207,
      "grad_norm": 0.222988560795784,
      "learning_rate": 4.900451804274898e-07,
      "loss": 0.0452,
      "step": 2559
    },
    {
      "epoch": 1.9204801200300075,
      "grad_norm": 0.20198635756969452,
      "learning_rate": 4.809429040626535e-07,
      "loss": 0.0437,
      "step": 2560
    },
    {
      "epoch": 1.9212303075768942,
      "grad_norm": 0.12086933106184006,
      "learning_rate": 4.719255480045148e-07,
      "loss": 0.0269,
      "step": 2561
    },
    {
      "epoch": 1.921980495123781,
      "grad_norm": 0.12228506803512573,
      "learning_rate": 4.6299312771694304e-07,
      "loss": 0.031,
      "step": 2562
    },
    {
      "epoch": 1.9227306826706676,
      "grad_norm": 0.2840142250061035,
      "learning_rate": 4.5414565851816806e-07,
      "loss": 0.0577,
      "step": 2563
    },
    {
      "epoch": 1.9234808702175545,
      "grad_norm": 0.19320012629032135,
      "learning_rate": 4.453831555807253e-07,
      "loss": 0.0426,
      "step": 2564
    },
    {
      "epoch": 1.924231057764441,
      "grad_norm": 0.14760589599609375,
      "learning_rate": 4.36705633931439e-07,
      "loss": 0.0313,
      "step": 2565
    },
    {
      "epoch": 1.924981245311328,
      "grad_norm": 0.18715044856071472,
      "learning_rate": 4.281131084514167e-07,
      "loss": 0.0486,
      "step": 2566
    },
    {
      "epoch": 1.9257314328582145,
      "grad_norm": 0.2232694923877716,
      "learning_rate": 4.196055938759824e-07,
      "loss": 0.0533,
      "step": 2567
    },
    {
      "epoch": 1.9264816204051014,
      "grad_norm": 0.21138830482959747,
      "learning_rate": 4.111831047946879e-07,
      "loss": 0.0425,
      "step": 2568
    },
    {
      "epoch": 1.927231807951988,
      "grad_norm": 0.29661694169044495,
      "learning_rate": 4.0284565565127384e-07,
      "loss": 0.0364,
      "step": 2569
    },
    {
      "epoch": 1.9279819954988748,
      "grad_norm": 0.23025383055210114,
      "learning_rate": 3.9459326074364756e-07,
      "loss": 0.0547,
      "step": 2570
    },
    {
      "epoch": 1.9287321830457613,
      "grad_norm": 0.13459983468055725,
      "learning_rate": 3.8642593422384965e-07,
      "loss": 0.032,
      "step": 2571
    },
    {
      "epoch": 1.9294823705926483,
      "grad_norm": 0.1266406923532486,
      "learning_rate": 3.7834369009804303e-07,
      "loss": 0.0324,
      "step": 2572
    },
    {
      "epoch": 1.9302325581395348,
      "grad_norm": 0.1843937188386917,
      "learning_rate": 3.703465422264796e-07,
      "loss": 0.0391,
      "step": 2573
    },
    {
      "epoch": 1.9309827456864217,
      "grad_norm": 0.15218639373779297,
      "learning_rate": 3.624345043234778e-07,
      "loss": 0.0339,
      "step": 2574
    },
    {
      "epoch": 1.9317329332333082,
      "grad_norm": 0.1591964066028595,
      "learning_rate": 3.5460758995741194e-07,
      "loss": 0.0285,
      "step": 2575
    },
    {
      "epoch": 1.9324831207801951,
      "grad_norm": 0.23621922731399536,
      "learning_rate": 3.468658125506563e-07,
      "loss": 0.0491,
      "step": 2576
    },
    {
      "epoch": 1.9332333083270816,
      "grad_norm": 0.14313896000385284,
      "learning_rate": 3.3920918537960754e-07,
      "loss": 0.0282,
      "step": 2577
    },
    {
      "epoch": 1.9339834958739686,
      "grad_norm": 0.18378767371177673,
      "learning_rate": 3.3163772157462357e-07,
      "loss": 0.0484,
      "step": 2578
    },
    {
      "epoch": 1.934733683420855,
      "grad_norm": 0.2192428708076477,
      "learning_rate": 3.241514341200236e-07,
      "loss": 0.0731,
      "step": 2579
    },
    {
      "epoch": 1.935483870967742,
      "grad_norm": 0.14537541568279266,
      "learning_rate": 3.1675033585404355e-07,
      "loss": 0.0304,
      "step": 2580
    },
    {
      "epoch": 1.9362340585146287,
      "grad_norm": 0.23168228566646576,
      "learning_rate": 3.0943443946884755e-07,
      "loss": 0.055,
      "step": 2581
    },
    {
      "epoch": 1.9369842460615154,
      "grad_norm": 0.15877023339271545,
      "learning_rate": 3.0220375751047194e-07,
      "loss": 0.0348,
      "step": 2582
    },
    {
      "epoch": 1.9377344336084021,
      "grad_norm": 0.21073149144649506,
      "learning_rate": 2.950583023788256e-07,
      "loss": 0.0505,
      "step": 2583
    },
    {
      "epoch": 1.9384846211552889,
      "grad_norm": 0.18246489763259888,
      "learning_rate": 2.879980863276621e-07,
      "loss": 0.0394,
      "step": 2584
    },
    {
      "epoch": 1.9392348087021756,
      "grad_norm": 0.15483985841274261,
      "learning_rate": 2.8102312146455755e-07,
      "loss": 0.0373,
      "step": 2585
    },
    {
      "epoch": 1.9399849962490623,
      "grad_norm": 0.21077927947044373,
      "learning_rate": 2.7413341975088824e-07,
      "loss": 0.0482,
      "step": 2586
    },
    {
      "epoch": 1.940735183795949,
      "grad_norm": 0.16964302957057953,
      "learning_rate": 2.6732899300180857e-07,
      "loss": 0.0316,
      "step": 2587
    },
    {
      "epoch": 1.9414853713428357,
      "grad_norm": 0.18473991751670837,
      "learning_rate": 2.606098528862566e-07,
      "loss": 0.0476,
      "step": 2588
    },
    {
      "epoch": 1.9422355588897224,
      "grad_norm": 0.1389751136302948,
      "learning_rate": 2.5397601092687627e-07,
      "loss": 0.0484,
      "step": 2589
    },
    {
      "epoch": 1.9429857464366092,
      "grad_norm": 0.1501632183790207,
      "learning_rate": 2.474274785000619e-07,
      "loss": 0.0339,
      "step": 2590
    },
    {
      "epoch": 1.9437359339834959,
      "grad_norm": 0.17686618864536285,
      "learning_rate": 2.40964266835908e-07,
      "loss": 0.0554,
      "step": 2591
    },
    {
      "epoch": 1.9444861215303826,
      "grad_norm": 0.13381214439868927,
      "learning_rate": 2.3458638701817636e-07,
      "loss": 0.0356,
      "step": 2592
    },
    {
      "epoch": 1.9452363090772693,
      "grad_norm": 0.15234093368053436,
      "learning_rate": 2.2829384998430681e-07,
      "loss": 0.034,
      "step": 2593
    },
    {
      "epoch": 1.945986496624156,
      "grad_norm": 0.1863105297088623,
      "learning_rate": 2.2208666652537846e-07,
      "loss": 0.0522,
      "step": 2594
    },
    {
      "epoch": 1.9467366841710427,
      "grad_norm": 0.1748838722705841,
      "learning_rate": 2.1596484728610421e-07,
      "loss": 0.0573,
      "step": 2595
    },
    {
      "epoch": 1.9474868717179294,
      "grad_norm": 0.23012542724609375,
      "learning_rate": 2.099284027647974e-07,
      "loss": 0.0449,
      "step": 2596
    },
    {
      "epoch": 1.9482370592648162,
      "grad_norm": 0.23254768550395966,
      "learning_rate": 2.039773433133718e-07,
      "loss": 0.0359,
      "step": 2597
    },
    {
      "epoch": 1.9489872468117029,
      "grad_norm": 0.24744318425655365,
      "learning_rate": 1.9811167913729723e-07,
      "loss": 0.0441,
      "step": 2598
    },
    {
      "epoch": 1.9497374343585896,
      "grad_norm": 0.12150046974420547,
      "learning_rate": 1.923314202956217e-07,
      "loss": 0.0217,
      "step": 2599
    },
    {
      "epoch": 1.9504876219054763,
      "grad_norm": 0.18914827704429626,
      "learning_rate": 1.8663657670091595e-07,
      "loss": 0.0489,
      "step": 2600
    },
    {
      "epoch": 1.9504876219054763,
      "eval_loss": 0.06815290451049805,
      "eval_runtime": 2.6569,
      "eval_samples_per_second": 20.324,
      "eval_steps_per_second": 5.269,
      "step": 2600
    },
    {
      "epoch": 1.9512378094523632,
      "grad_norm": 0.13075482845306396,
      "learning_rate": 1.810271581192735e-07,
      "loss": 0.026,
      "step": 2601
    },
    {
      "epoch": 1.9519879969992497,
      "grad_norm": 0.15196098387241364,
      "learning_rate": 1.755031741702995e-07,
      "loss": 0.0361,
      "step": 2602
    },
    {
      "epoch": 1.9527381845461367,
      "grad_norm": 0.19363312423229218,
      "learning_rate": 1.7006463432707177e-07,
      "loss": 0.0398,
      "step": 2603
    },
    {
      "epoch": 1.9534883720930232,
      "grad_norm": 0.13908173143863678,
      "learning_rate": 1.6471154791616317e-07,
      "loss": 0.0375,
      "step": 2604
    },
    {
      "epoch": 1.9542385596399101,
      "grad_norm": 0.15720535814762115,
      "learning_rate": 1.59443924117586e-07,
      "loss": 0.031,
      "step": 2605
    },
    {
      "epoch": 1.9549887471867966,
      "grad_norm": 0.1469358503818512,
      "learning_rate": 1.5426177196479207e-07,
      "loss": 0.0327,
      "step": 2606
    },
    {
      "epoch": 1.9557389347336835,
      "grad_norm": 0.20019660890102386,
      "learning_rate": 1.4916510034466702e-07,
      "loss": 0.073,
      "step": 2607
    },
    {
      "epoch": 1.95648912228057,
      "grad_norm": 0.15746329724788666,
      "learning_rate": 1.441539179974971e-07,
      "loss": 0.0534,
      "step": 2608
    },
    {
      "epoch": 1.957239309827457,
      "grad_norm": 0.20289811491966248,
      "learning_rate": 1.3922823351697479e-07,
      "loss": 0.0331,
      "step": 2609
    },
    {
      "epoch": 1.9579894973743435,
      "grad_norm": 0.2068672776222229,
      "learning_rate": 1.343880553501542e-07,
      "loss": 0.0318,
      "step": 2610
    },
    {
      "epoch": 1.9587396849212304,
      "grad_norm": 0.18808381259441376,
      "learning_rate": 1.2963339179746238e-07,
      "loss": 0.0378,
      "step": 2611
    },
    {
      "epoch": 1.959489872468117,
      "grad_norm": 0.20876070857048035,
      "learning_rate": 1.2496425101268804e-07,
      "loss": 0.064,
      "step": 2612
    },
    {
      "epoch": 1.9602400600150038,
      "grad_norm": 0.1189245879650116,
      "learning_rate": 1.2038064100294843e-07,
      "loss": 0.0237,
      "step": 2613
    },
    {
      "epoch": 1.9609902475618903,
      "grad_norm": 0.19520774483680725,
      "learning_rate": 1.158825696286725e-07,
      "loss": 0.0393,
      "step": 2614
    },
    {
      "epoch": 1.9617404351087773,
      "grad_norm": 0.1765730082988739,
      "learning_rate": 1.114700446036232e-07,
      "loss": 0.0356,
      "step": 2615
    },
    {
      "epoch": 1.9624906226556638,
      "grad_norm": 0.119378462433815,
      "learning_rate": 1.0714307349483089e-07,
      "loss": 0.031,
      "step": 2616
    },
    {
      "epoch": 1.9632408102025507,
      "grad_norm": 0.15803013741970062,
      "learning_rate": 1.029016637226432e-07,
      "loss": 0.0343,
      "step": 2617
    },
    {
      "epoch": 1.9639909977494372,
      "grad_norm": 0.1664055436849594,
      "learning_rate": 9.874582256064192e-08,
      "loss": 0.0353,
      "step": 2618
    },
    {
      "epoch": 1.9647411852963241,
      "grad_norm": 0.20048417150974274,
      "learning_rate": 9.46755571356983e-08,
      "loss": 0.0374,
      "step": 2619
    },
    {
      "epoch": 1.9654913728432108,
      "grad_norm": 0.21543757617473602,
      "learning_rate": 9.069087442791224e-08,
      "loss": 0.0706,
      "step": 2620
    },
    {
      "epoch": 1.9662415603900976,
      "grad_norm": 0.2072872519493103,
      "learning_rate": 8.679178127062871e-08,
      "loss": 0.0635,
      "step": 2621
    },
    {
      "epoch": 1.9669917479369843,
      "grad_norm": 0.19162490963935852,
      "learning_rate": 8.297828435039346e-08,
      "loss": 0.0351,
      "step": 2622
    },
    {
      "epoch": 1.967741935483871,
      "grad_norm": 0.17158183455467224,
      "learning_rate": 7.925039020699187e-08,
      "loss": 0.0377,
      "step": 2623
    },
    {
      "epoch": 1.9684921230307577,
      "grad_norm": 0.2609858810901642,
      "learning_rate": 7.56081052333879e-08,
      "loss": 0.0788,
      "step": 2624
    },
    {
      "epoch": 1.9692423105776444,
      "grad_norm": 0.19607380032539368,
      "learning_rate": 7.205143567574624e-08,
      "loss": 0.0528,
      "step": 2625
    },
    {
      "epoch": 1.9699924981245311,
      "grad_norm": 0.24832478165626526,
      "learning_rate": 6.858038763340458e-08,
      "loss": 0.0519,
      "step": 2626
    },
    {
      "epoch": 1.9707426856714179,
      "grad_norm": 0.24293671548366547,
      "learning_rate": 6.519496705886252e-08,
      "loss": 0.0559,
      "step": 2627
    },
    {
      "epoch": 1.9714928732183046,
      "grad_norm": 0.16258803009986877,
      "learning_rate": 6.189517975778713e-08,
      "loss": 0.0385,
      "step": 2628
    },
    {
      "epoch": 1.9722430607651913,
      "grad_norm": 0.2014719396829605,
      "learning_rate": 5.8681031388990724e-08,
      "loss": 0.0331,
      "step": 2629
    },
    {
      "epoch": 1.972993248312078,
      "grad_norm": 0.170815110206604,
      "learning_rate": 5.555252746441975e-08,
      "loss": 0.0476,
      "step": 2630
    },
    {
      "epoch": 1.9737434358589647,
      "grad_norm": 0.16516125202178955,
      "learning_rate": 5.25096733491548e-08,
      "loss": 0.0373,
      "step": 2631
    },
    {
      "epoch": 1.9744936234058514,
      "grad_norm": 0.1762745976448059,
      "learning_rate": 4.9552474261377326e-08,
      "loss": 0.0372,
      "step": 2632
    },
    {
      "epoch": 1.9752438109527382,
      "grad_norm": 0.1631755232810974,
      "learning_rate": 4.6680935272408465e-08,
      "loss": 0.0271,
      "step": 2633
    },
    {
      "epoch": 1.9759939984996249,
      "grad_norm": 0.16810788214206696,
      "learning_rate": 4.3895061306648e-08,
      "loss": 0.0458,
      "step": 2634
    },
    {
      "epoch": 1.9767441860465116,
      "grad_norm": 0.2130126953125,
      "learning_rate": 4.119485714159099e-08,
      "loss": 0.0483,
      "step": 2635
    },
    {
      "epoch": 1.9774943735933983,
      "grad_norm": 0.18357796967029572,
      "learning_rate": 3.8580327407827796e-08,
      "loss": 0.0476,
      "step": 2636
    },
    {
      "epoch": 1.978244561140285,
      "grad_norm": 0.23056019842624664,
      "learning_rate": 3.605147658901631e-08,
      "loss": 0.0403,
      "step": 2637
    },
    {
      "epoch": 1.978994748687172,
      "grad_norm": 0.24155446887016296,
      "learning_rate": 3.360830902189305e-08,
      "loss": 0.0845,
      "step": 2638
    },
    {
      "epoch": 1.9797449362340584,
      "grad_norm": 0.2481115311384201,
      "learning_rate": 3.125082889623987e-08,
      "loss": 0.0467,
      "step": 2639
    },
    {
      "epoch": 1.9804951237809454,
      "grad_norm": 0.3033842444419861,
      "learning_rate": 2.8979040254911717e-08,
      "loss": 0.0677,
      "step": 2640
    },
    {
      "epoch": 1.9812453113278319,
      "grad_norm": 0.1688585877418518,
      "learning_rate": 2.67929469937922e-08,
      "loss": 0.0429,
      "step": 2641
    },
    {
      "epoch": 1.9819954988747188,
      "grad_norm": 0.16313163936138153,
      "learning_rate": 2.4692552861826925e-08,
      "loss": 0.0346,
      "step": 2642
    },
    {
      "epoch": 1.9827456864216053,
      "grad_norm": 0.19703680276870728,
      "learning_rate": 2.2677861460984607e-08,
      "loss": 0.0381,
      "step": 2643
    },
    {
      "epoch": 1.9834958739684923,
      "grad_norm": 0.17000724375247955,
      "learning_rate": 2.074887624625155e-08,
      "loss": 0.0499,
      "step": 2644
    },
    {
      "epoch": 1.9842460615153787,
      "grad_norm": 0.15490411221981049,
      "learning_rate": 1.890560052565937e-08,
      "loss": 0.0384,
      "step": 2645
    },
    {
      "epoch": 1.9849962490622657,
      "grad_norm": 0.14818815886974335,
      "learning_rate": 1.7148037460235078e-08,
      "loss": 0.032,
      "step": 2646
    },
    {
      "epoch": 1.9857464366091522,
      "grad_norm": 0.29421237111091614,
      "learning_rate": 1.5476190064034334e-08,
      "loss": 0.0621,
      "step": 2647
    },
    {
      "epoch": 1.9864966241560391,
      "grad_norm": 0.15614813566207886,
      "learning_rate": 1.3890061204108185e-08,
      "loss": 0.0451,
      "step": 2648
    },
    {
      "epoch": 1.9872468117029256,
      "grad_norm": 0.24815501272678375,
      "learning_rate": 1.2389653600508588e-08,
      "loss": 0.0516,
      "step": 2649
    },
    {
      "epoch": 1.9879969992498125,
      "grad_norm": 0.2233879268169403,
      "learning_rate": 1.0974969826288428e-08,
      "loss": 0.0523,
      "step": 2650
    },
    {
      "epoch": 1.988747186796699,
      "grad_norm": 0.2333759218454361,
      "learning_rate": 9.646012307490405e-09,
      "loss": 0.0365,
      "step": 2651
    },
    {
      "epoch": 1.989497374343586,
      "grad_norm": 0.32244834303855896,
      "learning_rate": 8.402783323147034e-09,
      "loss": 0.0595,
      "step": 2652
    },
    {
      "epoch": 1.9902475618904725,
      "grad_norm": 0.21089965105056763,
      "learning_rate": 7.245285005275104e-09,
      "loss": 0.0433,
      "step": 2653
    },
    {
      "epoch": 1.9909977494373594,
      "grad_norm": 0.14350849390029907,
      "learning_rate": 6.1735193388701155e-09,
      "loss": 0.0364,
      "step": 2654
    },
    {
      "epoch": 1.991747936984246,
      "grad_norm": 0.1588405966758728,
      "learning_rate": 5.187488161895182e-09,
      "loss": 0.0222,
      "step": 2655
    },
    {
      "epoch": 1.9924981245311328,
      "grad_norm": 0.24085837602615356,
      "learning_rate": 4.28719316531434e-09,
      "loss": 0.0548,
      "step": 2656
    },
    {
      "epoch": 1.9932483120780196,
      "grad_norm": 0.18003256618976593,
      "learning_rate": 3.4726358930259328e-09,
      "loss": 0.051,
      "step": 2657
    },
    {
      "epoch": 1.9939984996249063,
      "grad_norm": 0.1899539977312088,
      "learning_rate": 2.743817741929222e-09,
      "loss": 0.0621,
      "step": 2658
    },
    {
      "epoch": 1.994748687171793,
      "grad_norm": 0.24146965146064758,
      "learning_rate": 2.1007399618688807e-09,
      "loss": 0.0355,
      "step": 2659
    },
    {
      "epoch": 1.9954988747186797,
      "grad_norm": 0.17824961245059967,
      "learning_rate": 1.543403655662745e-09,
      "loss": 0.0399,
      "step": 2660
    },
    {
      "epoch": 1.9962490622655664,
      "grad_norm": 0.18570539355278015,
      "learning_rate": 1.0718097790907156e-09,
      "loss": 0.0383,
      "step": 2661
    },
    {
      "epoch": 1.9969992498124531,
      "grad_norm": 0.12949424982070923,
      "learning_rate": 6.859591408836519e-10,
      "loss": 0.0241,
      "step": 2662
    },
    {
      "epoch": 1.9977494373593399,
      "grad_norm": 0.22268816828727722,
      "learning_rate": 3.8585240273447677e-10,
      "loss": 0.0601,
      "step": 2663
    },
    {
      "epoch": 1.9984996249062266,
      "grad_norm": 0.1774711310863495,
      "learning_rate": 1.7149007930927773e-10,
      "loss": 0.0538,
      "step": 2664
    },
    {
      "epoch": 1.9992498124531133,
      "grad_norm": 0.1890869289636612,
      "learning_rate": 4.2872538208449386e-11,
      "loss": 0.0301,
      "step": 2665
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.1974925547838211,
      "learning_rate": 0.0,
      "loss": 0.051,
      "step": 2666
    },
    {
      "epoch": 2.0,
      "step": 2666,
      "total_flos": 4.5995952202737254e+17,
      "train_loss": 0.05864244963993323,
      "train_runtime": 1863.1592,
      "train_samples_per_second": 5.72,
      "train_steps_per_second": 1.431
    }
  ],
  "logging_steps": 1,
  "max_steps": 2666,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.5995952202737254e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}