{
  "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": 0.06649772077798843,
      "learning_rate": 3.7453183520599253e-07,
      "loss": 0.1491,
      "step": 1
    },
    {
      "epoch": 0.0015003750937734434,
      "grad_norm": 0.1297842562198639,
      "learning_rate": 7.490636704119851e-07,
      "loss": 0.2357,
      "step": 2
    },
    {
      "epoch": 0.002250562640660165,
      "grad_norm": 0.0687314048409462,
      "learning_rate": 1.1235955056179775e-06,
      "loss": 0.114,
      "step": 3
    },
    {
      "epoch": 0.003000750187546887,
      "grad_norm": 0.1251819133758545,
      "learning_rate": 1.4981273408239701e-06,
      "loss": 0.2133,
      "step": 4
    },
    {
      "epoch": 0.0037509377344336083,
      "grad_norm": 0.060200925916433334,
      "learning_rate": 1.8726591760299627e-06,
      "loss": 0.113,
      "step": 5
    },
    {
      "epoch": 0.00450112528132033,
      "grad_norm": 0.12672510743141174,
      "learning_rate": 2.247191011235955e-06,
      "loss": 0.2232,
      "step": 6
    },
    {
      "epoch": 0.005251312828207052,
      "grad_norm": 0.10037848353385925,
      "learning_rate": 2.621722846441948e-06,
      "loss": 0.1297,
      "step": 7
    },
    {
      "epoch": 0.006001500375093774,
      "grad_norm": 0.07913671433925629,
      "learning_rate": 2.9962546816479402e-06,
      "loss": 0.1174,
      "step": 8
    },
    {
      "epoch": 0.006751687921980495,
      "grad_norm": 0.08461928367614746,
      "learning_rate": 3.3707865168539327e-06,
      "loss": 0.131,
      "step": 9
    },
    {
      "epoch": 0.007501875468867217,
      "grad_norm": 0.09129065275192261,
      "learning_rate": 3.7453183520599255e-06,
      "loss": 0.1834,
      "step": 10
    },
    {
      "epoch": 0.008252063015753939,
      "grad_norm": 0.0547911673784256,
      "learning_rate": 4.1198501872659175e-06,
      "loss": 0.078,
      "step": 11
    },
    {
      "epoch": 0.00900225056264066,
      "grad_norm": 0.11671412736177444,
      "learning_rate": 4.49438202247191e-06,
      "loss": 0.1937,
      "step": 12
    },
    {
      "epoch": 0.009752438109527382,
      "grad_norm": 0.11381004750728607,
      "learning_rate": 4.868913857677903e-06,
      "loss": 0.1882,
      "step": 13
    },
    {
      "epoch": 0.010502625656414103,
      "grad_norm": 0.4827083945274353,
      "learning_rate": 5.243445692883896e-06,
      "loss": 0.3005,
      "step": 14
    },
    {
      "epoch": 0.011252813203300824,
      "grad_norm": 0.0948605015873909,
      "learning_rate": 5.617977528089888e-06,
      "loss": 0.1403,
      "step": 15
    },
    {
      "epoch": 0.012003000750187547,
      "grad_norm": 0.17905683815479279,
      "learning_rate": 5.9925093632958805e-06,
      "loss": 0.1826,
      "step": 16
    },
    {
      "epoch": 0.012753188297074268,
      "grad_norm": 0.04930970072746277,
      "learning_rate": 6.367041198501873e-06,
      "loss": 0.0794,
      "step": 17
    },
    {
      "epoch": 0.01350337584396099,
      "grad_norm": 0.05358840152621269,
      "learning_rate": 6.741573033707865e-06,
      "loss": 0.0867,
      "step": 18
    },
    {
      "epoch": 0.014253563390847712,
      "grad_norm": 0.058819886296987534,
      "learning_rate": 7.116104868913858e-06,
      "loss": 0.1035,
      "step": 19
    },
    {
      "epoch": 0.015003750937734433,
      "grad_norm": 0.31954073905944824,
      "learning_rate": 7.490636704119851e-06,
      "loss": 0.2974,
      "step": 20
    },
    {
      "epoch": 0.015753938484621154,
      "grad_norm": 0.12035220861434937,
      "learning_rate": 7.865168539325843e-06,
      "loss": 0.142,
      "step": 21
    },
    {
      "epoch": 0.016504126031507877,
      "grad_norm": 0.1415967345237732,
      "learning_rate": 8.239700374531835e-06,
      "loss": 0.2324,
      "step": 22
    },
    {
      "epoch": 0.0172543135783946,
      "grad_norm": 0.11801585555076599,
      "learning_rate": 8.614232209737828e-06,
      "loss": 0.1633,
      "step": 23
    },
    {
      "epoch": 0.01800450112528132,
      "grad_norm": 0.1823718100786209,
      "learning_rate": 8.98876404494382e-06,
      "loss": 0.1321,
      "step": 24
    },
    {
      "epoch": 0.018754688672168042,
      "grad_norm": 0.16873396933078766,
      "learning_rate": 9.363295880149813e-06,
      "loss": 0.1453,
      "step": 25
    },
    {
      "epoch": 0.019504876219054765,
      "grad_norm": 0.12931793928146362,
      "learning_rate": 9.737827715355806e-06,
      "loss": 0.169,
      "step": 26
    },
    {
      "epoch": 0.020255063765941484,
      "grad_norm": 0.13040228188037872,
      "learning_rate": 1.0112359550561798e-05,
      "loss": 0.1103,
      "step": 27
    },
    {
      "epoch": 0.021005251312828207,
      "grad_norm": 0.21055294573307037,
      "learning_rate": 1.0486891385767791e-05,
      "loss": 0.2282,
      "step": 28
    },
    {
      "epoch": 0.02175543885971493,
      "grad_norm": 0.09372907131910324,
      "learning_rate": 1.0861423220973783e-05,
      "loss": 0.1398,
      "step": 29
    },
    {
      "epoch": 0.02250562640660165,
      "grad_norm": 0.15932485461235046,
      "learning_rate": 1.1235955056179776e-05,
      "loss": 0.1226,
      "step": 30
    },
    {
      "epoch": 0.023255813953488372,
      "grad_norm": 0.09856501966714859,
      "learning_rate": 1.161048689138577e-05,
      "loss": 0.1182,
      "step": 31
    },
    {
      "epoch": 0.024006001500375095,
      "grad_norm": 0.0559878945350647,
      "learning_rate": 1.1985018726591761e-05,
      "loss": 0.0681,
      "step": 32
    },
    {
      "epoch": 0.024756189047261814,
      "grad_norm": 0.1646910458803177,
      "learning_rate": 1.2359550561797752e-05,
      "loss": 0.1503,
      "step": 33
    },
    {
      "epoch": 0.025506376594148537,
      "grad_norm": 0.09920473396778107,
      "learning_rate": 1.2734082397003746e-05,
      "loss": 0.1493,
      "step": 34
    },
    {
      "epoch": 0.02625656414103526,
      "grad_norm": 0.21710704267024994,
      "learning_rate": 1.3108614232209737e-05,
      "loss": 0.2232,
      "step": 35
    },
    {
      "epoch": 0.02700675168792198,
      "grad_norm": 0.06066899746656418,
      "learning_rate": 1.348314606741573e-05,
      "loss": 0.0856,
      "step": 36
    },
    {
      "epoch": 0.0277569392348087,
      "grad_norm": 0.12835277616977692,
      "learning_rate": 1.3857677902621724e-05,
      "loss": 0.146,
      "step": 37
    },
    {
      "epoch": 0.028507126781695424,
      "grad_norm": 0.10923058539628983,
      "learning_rate": 1.4232209737827715e-05,
      "loss": 0.1094,
      "step": 38
    },
    {
      "epoch": 0.029257314328582147,
      "grad_norm": 0.24330253899097443,
      "learning_rate": 1.4606741573033709e-05,
      "loss": 0.2436,
      "step": 39
    },
    {
      "epoch": 0.030007501875468866,
      "grad_norm": 0.10736766457557678,
      "learning_rate": 1.4981273408239702e-05,
      "loss": 0.1085,
      "step": 40
    },
    {
      "epoch": 0.03075768942235559,
      "grad_norm": 0.05089711770415306,
      "learning_rate": 1.5355805243445692e-05,
      "loss": 0.0871,
      "step": 41
    },
    {
      "epoch": 0.03150787696924231,
      "grad_norm": 0.13075260818004608,
      "learning_rate": 1.5730337078651687e-05,
      "loss": 0.1558,
      "step": 42
    },
    {
      "epoch": 0.03225806451612903,
      "grad_norm": 0.09256450831890106,
      "learning_rate": 1.610486891385768e-05,
      "loss": 0.1214,
      "step": 43
    },
    {
      "epoch": 0.033008252063015754,
      "grad_norm": 0.13702747225761414,
      "learning_rate": 1.647940074906367e-05,
      "loss": 0.1186,
      "step": 44
    },
    {
      "epoch": 0.03375843960990248,
      "grad_norm": 0.1761491447687149,
      "learning_rate": 1.6853932584269665e-05,
      "loss": 0.1627,
      "step": 45
    },
    {
      "epoch": 0.0345086271567892,
      "grad_norm": 0.12984786927700043,
      "learning_rate": 1.7228464419475657e-05,
      "loss": 0.1335,
      "step": 46
    },
    {
      "epoch": 0.035258814703675916,
      "grad_norm": 0.09949901700019836,
      "learning_rate": 1.760299625468165e-05,
      "loss": 0.1188,
      "step": 47
    },
    {
      "epoch": 0.03600900225056264,
      "grad_norm": 0.2773364782333374,
      "learning_rate": 1.797752808988764e-05,
      "loss": 0.2703,
      "step": 48
    },
    {
      "epoch": 0.03675918979744936,
      "grad_norm": 0.3009333610534668,
      "learning_rate": 1.8352059925093635e-05,
      "loss": 0.2168,
      "step": 49
    },
    {
      "epoch": 0.037509377344336084,
      "grad_norm": 0.08068252354860306,
      "learning_rate": 1.8726591760299626e-05,
      "loss": 0.1113,
      "step": 50
    },
    {
      "epoch": 0.03825956489122281,
      "grad_norm": 0.08510925620794296,
      "learning_rate": 1.9101123595505618e-05,
      "loss": 0.0675,
      "step": 51
    },
    {
      "epoch": 0.03900975243810953,
      "grad_norm": 0.16944704949855804,
      "learning_rate": 1.9475655430711613e-05,
      "loss": 0.1452,
      "step": 52
    },
    {
      "epoch": 0.03975993998499625,
      "grad_norm": 0.21001164615154266,
      "learning_rate": 1.9850187265917604e-05,
      "loss": 0.2077,
      "step": 53
    },
    {
      "epoch": 0.04051012753188297,
      "grad_norm": 0.21587851643562317,
      "learning_rate": 2.0224719101123596e-05,
      "loss": 0.14,
      "step": 54
    },
    {
      "epoch": 0.04126031507876969,
      "grad_norm": 0.19861550629138947,
      "learning_rate": 2.059925093632959e-05,
      "loss": 0.1491,
      "step": 55
    },
    {
      "epoch": 0.042010502625656414,
      "grad_norm": 0.05874146148562431,
      "learning_rate": 2.0973782771535582e-05,
      "loss": 0.0653,
      "step": 56
    },
    {
      "epoch": 0.04276069017254314,
      "grad_norm": 0.2626660466194153,
      "learning_rate": 2.1348314606741574e-05,
      "loss": 0.1116,
      "step": 57
    },
    {
      "epoch": 0.04351087771942986,
      "grad_norm": 0.3618119955062866,
      "learning_rate": 2.1722846441947566e-05,
      "loss": 0.2157,
      "step": 58
    },
    {
      "epoch": 0.04426106526631658,
      "grad_norm": 0.12022845447063446,
      "learning_rate": 2.209737827715356e-05,
      "loss": 0.0955,
      "step": 59
    },
    {
      "epoch": 0.0450112528132033,
      "grad_norm": 0.07745017111301422,
      "learning_rate": 2.2471910112359552e-05,
      "loss": 0.0777,
      "step": 60
    },
    {
      "epoch": 0.04576144036009002,
      "grad_norm": 0.12869754433631897,
      "learning_rate": 2.2846441947565544e-05,
      "loss": 0.1211,
      "step": 61
    },
    {
      "epoch": 0.046511627906976744,
      "grad_norm": 0.13848523795604706,
      "learning_rate": 2.322097378277154e-05,
      "loss": 0.0762,
      "step": 62
    },
    {
      "epoch": 0.047261815453863466,
      "grad_norm": 0.1735474169254303,
      "learning_rate": 2.359550561797753e-05,
      "loss": 0.1488,
      "step": 63
    },
    {
      "epoch": 0.04801200300075019,
      "grad_norm": 0.1833038032054901,
      "learning_rate": 2.3970037453183522e-05,
      "loss": 0.1407,
      "step": 64
    },
    {
      "epoch": 0.04876219054763691,
      "grad_norm": 0.08675671368837357,
      "learning_rate": 2.4344569288389517e-05,
      "loss": 0.0663,
      "step": 65
    },
    {
      "epoch": 0.04951237809452363,
      "grad_norm": 0.4732707738876343,
      "learning_rate": 2.4719101123595505e-05,
      "loss": 0.0857,
      "step": 66
    },
    {
      "epoch": 0.05026256564141035,
      "grad_norm": 0.07791633903980255,
      "learning_rate": 2.50936329588015e-05,
      "loss": 0.0585,
      "step": 67
    },
    {
      "epoch": 0.05101275318829707,
      "grad_norm": 0.11928839981555939,
      "learning_rate": 2.546816479400749e-05,
      "loss": 0.0748,
      "step": 68
    },
    {
      "epoch": 0.051762940735183796,
      "grad_norm": 0.089127317070961,
      "learning_rate": 2.5842696629213486e-05,
      "loss": 0.09,
      "step": 69
    },
    {
      "epoch": 0.05251312828207052,
      "grad_norm": 0.10768590867519379,
      "learning_rate": 2.6217228464419475e-05,
      "loss": 0.0804,
      "step": 70
    },
    {
      "epoch": 0.05326331582895724,
      "grad_norm": 0.21062488853931427,
      "learning_rate": 2.6591760299625466e-05,
      "loss": 0.1271,
      "step": 71
    },
    {
      "epoch": 0.05401350337584396,
      "grad_norm": 0.11834652721881866,
      "learning_rate": 2.696629213483146e-05,
      "loss": 0.0821,
      "step": 72
    },
    {
      "epoch": 0.05476369092273068,
      "grad_norm": 0.09663186967372894,
      "learning_rate": 2.7340823970037456e-05,
      "loss": 0.0593,
      "step": 73
    },
    {
      "epoch": 0.0555138784696174,
      "grad_norm": 0.22409404814243317,
      "learning_rate": 2.7715355805243448e-05,
      "loss": 0.1147,
      "step": 74
    },
    {
      "epoch": 0.056264066016504126,
      "grad_norm": 0.13819625973701477,
      "learning_rate": 2.8089887640449443e-05,
      "loss": 0.1101,
      "step": 75
    },
    {
      "epoch": 0.05701425356339085,
      "grad_norm": 0.25516584515571594,
      "learning_rate": 2.846441947565543e-05,
      "loss": 0.1276,
      "step": 76
    },
    {
      "epoch": 0.05776444111027757,
      "grad_norm": 0.0935330018401146,
      "learning_rate": 2.8838951310861422e-05,
      "loss": 0.0749,
      "step": 77
    },
    {
      "epoch": 0.058514628657164294,
      "grad_norm": 0.11693164706230164,
      "learning_rate": 2.9213483146067417e-05,
      "loss": 0.0756,
      "step": 78
    },
    {
      "epoch": 0.05926481620405101,
      "grad_norm": 0.1455622911453247,
      "learning_rate": 2.958801498127341e-05,
      "loss": 0.0847,
      "step": 79
    },
    {
      "epoch": 0.06001500375093773,
      "grad_norm": 0.09133189916610718,
      "learning_rate": 2.9962546816479404e-05,
      "loss": 0.056,
      "step": 80
    },
    {
      "epoch": 0.060765191297824456,
      "grad_norm": 0.12661296129226685,
      "learning_rate": 3.0337078651685396e-05,
      "loss": 0.0823,
      "step": 81
    },
    {
      "epoch": 0.06151537884471118,
      "grad_norm": 0.13131161034107208,
      "learning_rate": 3.0711610486891384e-05,
      "loss": 0.0772,
      "step": 82
    },
    {
      "epoch": 0.0622655663915979,
      "grad_norm": 0.2745862305164337,
      "learning_rate": 3.108614232209738e-05,
      "loss": 0.1155,
      "step": 83
    },
    {
      "epoch": 0.06301575393848462,
      "grad_norm": 0.11527538299560547,
      "learning_rate": 3.1460674157303374e-05,
      "loss": 0.0655,
      "step": 84
    },
    {
      "epoch": 0.06376594148537135,
      "grad_norm": 0.12666447460651398,
      "learning_rate": 3.183520599250936e-05,
      "loss": 0.0656,
      "step": 85
    },
    {
      "epoch": 0.06451612903225806,
      "grad_norm": 0.09897023439407349,
      "learning_rate": 3.220973782771536e-05,
      "loss": 0.0659,
      "step": 86
    },
    {
      "epoch": 0.06526631657914479,
      "grad_norm": 0.15928857028484344,
      "learning_rate": 3.258426966292135e-05,
      "loss": 0.0857,
      "step": 87
    },
    {
      "epoch": 0.06601650412603151,
      "grad_norm": 0.14418122172355652,
      "learning_rate": 3.295880149812734e-05,
      "loss": 0.0903,
      "step": 88
    },
    {
      "epoch": 0.06676669167291822,
      "grad_norm": 0.15921813249588013,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0811,
      "step": 89
    },
    {
      "epoch": 0.06751687921980495,
      "grad_norm": 0.18934333324432373,
      "learning_rate": 3.370786516853933e-05,
      "loss": 0.0836,
      "step": 90
    },
    {
      "epoch": 0.06826706676669167,
      "grad_norm": 0.10202574729919434,
      "learning_rate": 3.408239700374532e-05,
      "loss": 0.0464,
      "step": 91
    },
    {
      "epoch": 0.0690172543135784,
      "grad_norm": 0.12126179784536362,
      "learning_rate": 3.445692883895131e-05,
      "loss": 0.052,
      "step": 92
    },
    {
      "epoch": 0.06976744186046512,
      "grad_norm": 0.06997974961996078,
      "learning_rate": 3.483146067415731e-05,
      "loss": 0.0423,
      "step": 93
    },
    {
      "epoch": 0.07051762940735183,
      "grad_norm": 0.1367831528186798,
      "learning_rate": 3.52059925093633e-05,
      "loss": 0.054,
      "step": 94
    },
    {
      "epoch": 0.07126781695423856,
      "grad_norm": 0.1674944907426834,
      "learning_rate": 3.558052434456929e-05,
      "loss": 0.0635,
      "step": 95
    },
    {
      "epoch": 0.07201800450112528,
      "grad_norm": 0.1916777491569519,
      "learning_rate": 3.595505617977528e-05,
      "loss": 0.0476,
      "step": 96
    },
    {
      "epoch": 0.072768192048012,
      "grad_norm": 0.5634555220603943,
      "learning_rate": 3.6329588014981274e-05,
      "loss": 0.0509,
      "step": 97
    },
    {
      "epoch": 0.07351837959489872,
      "grad_norm": 0.24589844048023224,
      "learning_rate": 3.670411985018727e-05,
      "loss": 0.0702,
      "step": 98
    },
    {
      "epoch": 0.07426856714178545,
      "grad_norm": 0.19461844861507416,
      "learning_rate": 3.7078651685393264e-05,
      "loss": 0.0684,
      "step": 99
    },
    {
      "epoch": 0.07501875468867217,
      "grad_norm": 0.17358846962451935,
      "learning_rate": 3.745318352059925e-05,
      "loss": 0.072,
      "step": 100
    },
    {
      "epoch": 0.07576894223555888,
      "grad_norm": 0.1436762511730194,
      "learning_rate": 3.782771535580524e-05,
      "loss": 0.0573,
      "step": 101
    },
    {
      "epoch": 0.07651912978244561,
      "grad_norm": 0.17795026302337646,
      "learning_rate": 3.8202247191011236e-05,
      "loss": 0.0662,
      "step": 102
    },
    {
      "epoch": 0.07726931732933233,
      "grad_norm": 0.1983191967010498,
      "learning_rate": 3.857677902621723e-05,
      "loss": 0.0559,
      "step": 103
    },
    {
      "epoch": 0.07801950487621906,
      "grad_norm": 0.24022872745990753,
      "learning_rate": 3.8951310861423226e-05,
      "loss": 0.0724,
      "step": 104
    },
    {
      "epoch": 0.07876969242310577,
      "grad_norm": 0.16081267595291138,
      "learning_rate": 3.9325842696629214e-05,
      "loss": 0.0762,
      "step": 105
    },
    {
      "epoch": 0.0795198799699925,
      "grad_norm": 0.31475764513015747,
      "learning_rate": 3.970037453183521e-05,
      "loss": 0.0581,
      "step": 106
    },
    {
      "epoch": 0.08027006751687922,
      "grad_norm": 0.23797044157981873,
      "learning_rate": 4.00749063670412e-05,
      "loss": 0.0546,
      "step": 107
    },
    {
      "epoch": 0.08102025506376594,
      "grad_norm": 0.18666911125183105,
      "learning_rate": 4.044943820224719e-05,
      "loss": 0.0604,
      "step": 108
    },
    {
      "epoch": 0.08177044261065267,
      "grad_norm": 0.10255415737628937,
      "learning_rate": 4.082397003745319e-05,
      "loss": 0.0404,
      "step": 109
    },
    {
      "epoch": 0.08252063015753938,
      "grad_norm": 0.2886727750301361,
      "learning_rate": 4.119850187265918e-05,
      "loss": 0.058,
      "step": 110
    },
    {
      "epoch": 0.08327081770442611,
      "grad_norm": 0.1744941771030426,
      "learning_rate": 4.157303370786517e-05,
      "loss": 0.0696,
      "step": 111
    },
    {
      "epoch": 0.08402100525131283,
      "grad_norm": 0.2490309327840805,
      "learning_rate": 4.1947565543071165e-05,
      "loss": 0.0479,
      "step": 112
    },
    {
      "epoch": 0.08477119279819954,
      "grad_norm": 0.2167392075061798,
      "learning_rate": 4.232209737827715e-05,
      "loss": 0.0539,
      "step": 113
    },
    {
      "epoch": 0.08552138034508627,
      "grad_norm": 0.16126973927021027,
      "learning_rate": 4.269662921348315e-05,
      "loss": 0.044,
      "step": 114
    },
    {
      "epoch": 0.08627156789197299,
      "grad_norm": 0.17505522072315216,
      "learning_rate": 4.307116104868914e-05,
      "loss": 0.0548,
      "step": 115
    },
    {
      "epoch": 0.08702175543885972,
      "grad_norm": 0.27888697385787964,
      "learning_rate": 4.344569288389513e-05,
      "loss": 0.0681,
      "step": 116
    },
    {
      "epoch": 0.08777194298574643,
      "grad_norm": 0.13575726747512817,
      "learning_rate": 4.3820224719101126e-05,
      "loss": 0.0385,
      "step": 117
    },
    {
      "epoch": 0.08852213053263316,
      "grad_norm": 0.143665611743927,
      "learning_rate": 4.419475655430712e-05,
      "loss": 0.0442,
      "step": 118
    },
    {
      "epoch": 0.08927231807951988,
      "grad_norm": 0.3198046386241913,
      "learning_rate": 4.456928838951311e-05,
      "loss": 0.0796,
      "step": 119
    },
    {
      "epoch": 0.0900225056264066,
      "grad_norm": 0.11711857467889786,
      "learning_rate": 4.4943820224719104e-05,
      "loss": 0.0325,
      "step": 120
    },
    {
      "epoch": 0.09077269317329333,
      "grad_norm": 0.19504939019680023,
      "learning_rate": 4.531835205992509e-05,
      "loss": 0.0454,
      "step": 121
    },
    {
      "epoch": 0.09152288072018004,
      "grad_norm": 0.16343529522418976,
      "learning_rate": 4.569288389513109e-05,
      "loss": 0.0423,
      "step": 122
    },
    {
      "epoch": 0.09227306826706677,
      "grad_norm": 0.0722413882613182,
      "learning_rate": 4.606741573033708e-05,
      "loss": 0.0364,
      "step": 123
    },
    {
      "epoch": 0.09302325581395349,
      "grad_norm": 0.20592856407165527,
      "learning_rate": 4.644194756554308e-05,
      "loss": 0.0628,
      "step": 124
    },
    {
      "epoch": 0.09377344336084022,
      "grad_norm": 0.19177493453025818,
      "learning_rate": 4.6816479400749066e-05,
      "loss": 0.0543,
      "step": 125
    },
    {
      "epoch": 0.09452363090772693,
      "grad_norm": 0.24182191491127014,
      "learning_rate": 4.719101123595506e-05,
      "loss": 0.0493,
      "step": 126
    },
    {
      "epoch": 0.09527381845461365,
      "grad_norm": 0.18809586763381958,
      "learning_rate": 4.756554307116105e-05,
      "loss": 0.0531,
      "step": 127
    },
    {
      "epoch": 0.09602400600150038,
      "grad_norm": 0.16542655229568481,
      "learning_rate": 4.7940074906367044e-05,
      "loss": 0.0345,
      "step": 128
    },
    {
      "epoch": 0.0967741935483871,
      "grad_norm": 0.12689034640789032,
      "learning_rate": 4.831460674157304e-05,
      "loss": 0.0255,
      "step": 129
    },
    {
      "epoch": 0.09752438109527382,
      "grad_norm": 0.14506380259990692,
      "learning_rate": 4.8689138576779034e-05,
      "loss": 0.0432,
      "step": 130
    },
    {
      "epoch": 0.09827456864216054,
      "grad_norm": 0.24236801266670227,
      "learning_rate": 4.906367041198502e-05,
      "loss": 0.0869,
      "step": 131
    },
    {
      "epoch": 0.09902475618904726,
      "grad_norm": 0.23130767047405243,
      "learning_rate": 4.943820224719101e-05,
      "loss": 0.0498,
      "step": 132
    },
    {
      "epoch": 0.09977494373593399,
      "grad_norm": 0.06554227322340012,
      "learning_rate": 4.9812734082397005e-05,
      "loss": 0.0256,
      "step": 133
    },
    {
      "epoch": 0.1005251312828207,
      "grad_norm": 0.14324985444545746,
      "learning_rate": 5.0187265917603e-05,
      "loss": 0.0467,
      "step": 134
    },
    {
      "epoch": 0.10127531882970743,
      "grad_norm": 0.37844982743263245,
      "learning_rate": 5.0561797752808995e-05,
      "loss": 0.0656,
      "step": 135
    },
    {
      "epoch": 0.10202550637659415,
      "grad_norm": 0.20668180286884308,
      "learning_rate": 5.093632958801498e-05,
      "loss": 0.0444,
      "step": 136
    },
    {
      "epoch": 0.10277569392348088,
      "grad_norm": 0.10723793506622314,
      "learning_rate": 5.131086142322098e-05,
      "loss": 0.0214,
      "step": 137
    },
    {
      "epoch": 0.10352588147036759,
      "grad_norm": 0.19435639679431915,
      "learning_rate": 5.168539325842697e-05,
      "loss": 0.0547,
      "step": 138
    },
    {
      "epoch": 0.10427606901725431,
      "grad_norm": 0.19808021187782288,
      "learning_rate": 5.205992509363297e-05,
      "loss": 0.0372,
      "step": 139
    },
    {
      "epoch": 0.10502625656414104,
      "grad_norm": 0.17958632111549377,
      "learning_rate": 5.243445692883895e-05,
      "loss": 0.0416,
      "step": 140
    },
    {
      "epoch": 0.10577644411102775,
      "grad_norm": 0.09122506529092789,
      "learning_rate": 5.2808988764044944e-05,
      "loss": 0.0324,
      "step": 141
    },
    {
      "epoch": 0.10652663165791448,
      "grad_norm": 0.13046078383922577,
      "learning_rate": 5.318352059925093e-05,
      "loss": 0.0304,
      "step": 142
    },
    {
      "epoch": 0.1072768192048012,
      "grad_norm": 0.2404167801141739,
      "learning_rate": 5.355805243445693e-05,
      "loss": 0.0572,
      "step": 143
    },
    {
      "epoch": 0.10802700675168792,
      "grad_norm": 0.2153376042842865,
      "learning_rate": 5.393258426966292e-05,
      "loss": 0.0443,
      "step": 144
    },
    {
      "epoch": 0.10877719429857464,
      "grad_norm": 0.2301207333803177,
      "learning_rate": 5.430711610486892e-05,
      "loss": 0.0619,
      "step": 145
    },
    {
      "epoch": 0.10952738184546136,
      "grad_norm": 0.2225511372089386,
      "learning_rate": 5.468164794007491e-05,
      "loss": 0.041,
      "step": 146
    },
    {
      "epoch": 0.11027756939234809,
      "grad_norm": 0.3045608699321747,
      "learning_rate": 5.50561797752809e-05,
      "loss": 0.0608,
      "step": 147
    },
    {
      "epoch": 0.1110277569392348,
      "grad_norm": 0.47876784205436707,
      "learning_rate": 5.5430711610486895e-05,
      "loss": 0.1024,
      "step": 148
    },
    {
      "epoch": 0.11177794448612154,
      "grad_norm": 0.1713646650314331,
      "learning_rate": 5.580524344569289e-05,
      "loss": 0.0512,
      "step": 149
    },
    {
      "epoch": 0.11252813203300825,
      "grad_norm": 0.1879701018333435,
      "learning_rate": 5.6179775280898885e-05,
      "loss": 0.0342,
      "step": 150
    },
    {
      "epoch": 0.11327831957989497,
      "grad_norm": 0.2024771124124527,
      "learning_rate": 5.6554307116104874e-05,
      "loss": 0.0421,
      "step": 151
    },
    {
      "epoch": 0.1140285071267817,
      "grad_norm": 0.18217679858207703,
      "learning_rate": 5.692883895131086e-05,
      "loss": 0.0344,
      "step": 152
    },
    {
      "epoch": 0.11477869467366841,
      "grad_norm": 0.13079820573329926,
      "learning_rate": 5.730337078651685e-05,
      "loss": 0.0328,
      "step": 153
    },
    {
      "epoch": 0.11552888222055514,
      "grad_norm": 0.15242129564285278,
      "learning_rate": 5.7677902621722845e-05,
      "loss": 0.0392,
      "step": 154
    },
    {
      "epoch": 0.11627906976744186,
      "grad_norm": 0.08799269795417786,
      "learning_rate": 5.805243445692884e-05,
      "loss": 0.0217,
      "step": 155
    },
    {
      "epoch": 0.11702925731432859,
      "grad_norm": 0.15637950599193573,
      "learning_rate": 5.8426966292134835e-05,
      "loss": 0.0356,
      "step": 156
    },
    {
      "epoch": 0.1177794448612153,
      "grad_norm": 0.11212990432977676,
      "learning_rate": 5.880149812734082e-05,
      "loss": 0.0298,
      "step": 157
    },
    {
      "epoch": 0.11852963240810202,
      "grad_norm": 0.1721002459526062,
      "learning_rate": 5.917602996254682e-05,
      "loss": 0.0511,
      "step": 158
    },
    {
      "epoch": 0.11927981995498875,
      "grad_norm": 0.23305867612361908,
      "learning_rate": 5.955056179775281e-05,
      "loss": 0.0543,
      "step": 159
    },
    {
      "epoch": 0.12003000750187547,
      "grad_norm": 0.24314911663532257,
      "learning_rate": 5.992509363295881e-05,
      "loss": 0.0542,
      "step": 160
    },
    {
      "epoch": 0.1207801950487622,
      "grad_norm": 0.12653297185897827,
      "learning_rate": 6.02996254681648e-05,
      "loss": 0.0361,
      "step": 161
    },
    {
      "epoch": 0.12153038259564891,
      "grad_norm": 0.1558983474969864,
      "learning_rate": 6.067415730337079e-05,
      "loss": 0.0554,
      "step": 162
    },
    {
      "epoch": 0.12228057014253563,
      "grad_norm": 0.15075267851352692,
      "learning_rate": 6.104868913857679e-05,
      "loss": 0.0477,
      "step": 163
    },
    {
      "epoch": 0.12303075768942236,
      "grad_norm": 0.14883558452129364,
      "learning_rate": 6.142322097378277e-05,
      "loss": 0.041,
      "step": 164
    },
    {
      "epoch": 0.12378094523630907,
      "grad_norm": 0.09038081020116806,
      "learning_rate": 6.179775280898876e-05,
      "loss": 0.0316,
      "step": 165
    },
    {
      "epoch": 0.1245311327831958,
      "grad_norm": 0.07693412154912949,
      "learning_rate": 6.217228464419476e-05,
      "loss": 0.0223,
      "step": 166
    },
    {
      "epoch": 0.12528132033008252,
      "grad_norm": 0.10217053443193436,
      "learning_rate": 6.254681647940075e-05,
      "loss": 0.0251,
      "step": 167
    },
    {
      "epoch": 0.12603150787696923,
      "grad_norm": 0.12066909670829773,
      "learning_rate": 6.292134831460675e-05,
      "loss": 0.0319,
      "step": 168
    },
    {
      "epoch": 0.12678169542385595,
      "grad_norm": 0.15891261398792267,
      "learning_rate": 6.329588014981274e-05,
      "loss": 0.0384,
      "step": 169
    },
    {
      "epoch": 0.1275318829707427,
      "grad_norm": 0.16246218979358673,
      "learning_rate": 6.367041198501872e-05,
      "loss": 0.0364,
      "step": 170
    },
    {
      "epoch": 0.1282820705176294,
      "grad_norm": 0.16500765085220337,
      "learning_rate": 6.404494382022472e-05,
      "loss": 0.0316,
      "step": 171
    },
    {
      "epoch": 0.12903225806451613,
      "grad_norm": 0.15567168593406677,
      "learning_rate": 6.441947565543071e-05,
      "loss": 0.0369,
      "step": 172
    },
    {
      "epoch": 0.12978244561140284,
      "grad_norm": 0.2111051082611084,
      "learning_rate": 6.479400749063671e-05,
      "loss": 0.0425,
      "step": 173
    },
    {
      "epoch": 0.13053263315828958,
      "grad_norm": 0.20215699076652527,
      "learning_rate": 6.51685393258427e-05,
      "loss": 0.0392,
      "step": 174
    },
    {
      "epoch": 0.1312828207051763,
      "grad_norm": 0.1345527321100235,
      "learning_rate": 6.55430711610487e-05,
      "loss": 0.0267,
      "step": 175
    },
    {
      "epoch": 0.13203300825206302,
      "grad_norm": 0.14389894902706146,
      "learning_rate": 6.591760299625468e-05,
      "loss": 0.023,
      "step": 176
    },
    {
      "epoch": 0.13278319579894973,
      "grad_norm": 0.23610113561153412,
      "learning_rate": 6.629213483146067e-05,
      "loss": 0.0596,
      "step": 177
    },
    {
      "epoch": 0.13353338334583645,
      "grad_norm": 0.20613926649093628,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0343,
      "step": 178
    },
    {
      "epoch": 0.1342835708927232,
      "grad_norm": 0.28699466586112976,
      "learning_rate": 6.704119850187266e-05,
      "loss": 0.0434,
      "step": 179
    },
    {
      "epoch": 0.1350337584396099,
      "grad_norm": 0.2657482326030731,
      "learning_rate": 6.741573033707866e-05,
      "loss": 0.0388,
      "step": 180
    },
    {
      "epoch": 0.13578394598649662,
      "grad_norm": 0.3304208219051361,
      "learning_rate": 6.779026217228464e-05,
      "loss": 0.0564,
      "step": 181
    },
    {
      "epoch": 0.13653413353338334,
      "grad_norm": 0.20125621557235718,
      "learning_rate": 6.816479400749064e-05,
      "loss": 0.031,
      "step": 182
    },
    {
      "epoch": 0.13728432108027006,
      "grad_norm": 0.15342605113983154,
      "learning_rate": 6.853932584269663e-05,
      "loss": 0.0332,
      "step": 183
    },
    {
      "epoch": 0.1380345086271568,
      "grad_norm": 0.2653309106826782,
      "learning_rate": 6.891385767790263e-05,
      "loss": 0.0589,
      "step": 184
    },
    {
      "epoch": 0.13878469617404351,
      "grad_norm": 0.2295987457036972,
      "learning_rate": 6.928838951310862e-05,
      "loss": 0.039,
      "step": 185
    },
    {
      "epoch": 0.13953488372093023,
      "grad_norm": 0.16181237995624542,
      "learning_rate": 6.966292134831462e-05,
      "loss": 0.0294,
      "step": 186
    },
    {
      "epoch": 0.14028507126781695,
      "grad_norm": 0.08588402718305588,
      "learning_rate": 7.003745318352061e-05,
      "loss": 0.0288,
      "step": 187
    },
    {
      "epoch": 0.14103525881470366,
      "grad_norm": 0.15989354252815247,
      "learning_rate": 7.04119850187266e-05,
      "loss": 0.0533,
      "step": 188
    },
    {
      "epoch": 0.1417854463615904,
      "grad_norm": 0.21237890422344208,
      "learning_rate": 7.078651685393259e-05,
      "loss": 0.0543,
      "step": 189
    },
    {
      "epoch": 0.14253563390847712,
      "grad_norm": 0.23640727996826172,
      "learning_rate": 7.116104868913858e-05,
      "loss": 0.0511,
      "step": 190
    },
    {
      "epoch": 0.14328582145536384,
      "grad_norm": 0.22190897166728973,
      "learning_rate": 7.153558052434456e-05,
      "loss": 0.0518,
      "step": 191
    },
    {
      "epoch": 0.14403600900225055,
      "grad_norm": 0.22951455414295197,
      "learning_rate": 7.191011235955056e-05,
      "loss": 0.0448,
      "step": 192
    },
    {
      "epoch": 0.1447861965491373,
      "grad_norm": 0.15378020703792572,
      "learning_rate": 7.228464419475655e-05,
      "loss": 0.0351,
      "step": 193
    },
    {
      "epoch": 0.145536384096024,
      "grad_norm": 0.21683332324028015,
      "learning_rate": 7.265917602996255e-05,
      "loss": 0.0571,
      "step": 194
    },
    {
      "epoch": 0.14628657164291073,
      "grad_norm": 0.15624549984931946,
      "learning_rate": 7.303370786516854e-05,
      "loss": 0.0418,
      "step": 195
    },
    {
      "epoch": 0.14703675918979744,
      "grad_norm": 0.14242206513881683,
      "learning_rate": 7.340823970037454e-05,
      "loss": 0.0407,
      "step": 196
    },
    {
      "epoch": 0.14778694673668416,
      "grad_norm": 0.10074103623628616,
      "learning_rate": 7.378277153558053e-05,
      "loss": 0.0312,
      "step": 197
    },
    {
      "epoch": 0.1485371342835709,
      "grad_norm": 0.24729660153388977,
      "learning_rate": 7.415730337078653e-05,
      "loss": 0.0394,
      "step": 198
    },
    {
      "epoch": 0.14928732183045762,
      "grad_norm": 0.16826917231082916,
      "learning_rate": 7.453183520599252e-05,
      "loss": 0.0345,
      "step": 199
    },
    {
      "epoch": 0.15003750937734434,
      "grad_norm": 0.2644681930541992,
      "learning_rate": 7.49063670411985e-05,
      "loss": 0.0528,
      "step": 200
    },
    {
      "epoch": 0.15003750937734434,
      "eval_loss": 0.049570418894290924,
      "eval_runtime": 5.1436,
      "eval_samples_per_second": 10.498,
      "eval_steps_per_second": 2.722,
      "step": 200
    },
    {
      "epoch": 0.15078769692423105,
      "grad_norm": 0.17306017875671387,
      "learning_rate": 7.52808988764045e-05,
      "loss": 0.0529,
      "step": 201
    },
    {
      "epoch": 0.15153788447111777,
      "grad_norm": 0.13111412525177002,
      "learning_rate": 7.565543071161048e-05,
      "loss": 0.0404,
      "step": 202
    },
    {
      "epoch": 0.1522880720180045,
      "grad_norm": 0.12245412915945053,
      "learning_rate": 7.602996254681648e-05,
      "loss": 0.0325,
      "step": 203
    },
    {
      "epoch": 0.15303825956489123,
      "grad_norm": 0.15576554834842682,
      "learning_rate": 7.640449438202247e-05,
      "loss": 0.0393,
      "step": 204
    },
    {
      "epoch": 0.15378844711177794,
      "grad_norm": 0.22586987912654877,
      "learning_rate": 7.677902621722847e-05,
      "loss": 0.036,
      "step": 205
    },
    {
      "epoch": 0.15453863465866466,
      "grad_norm": 0.19062842428684235,
      "learning_rate": 7.715355805243446e-05,
      "loss": 0.0454,
      "step": 206
    },
    {
      "epoch": 0.15528882220555137,
      "grad_norm": 0.18759313225746155,
      "learning_rate": 7.752808988764046e-05,
      "loss": 0.0362,
      "step": 207
    },
    {
      "epoch": 0.15603900975243812,
      "grad_norm": 0.33924275636672974,
      "learning_rate": 7.790262172284645e-05,
      "loss": 0.0645,
      "step": 208
    },
    {
      "epoch": 0.15678919729932483,
      "grad_norm": 0.124433733522892,
      "learning_rate": 7.827715355805245e-05,
      "loss": 0.0331,
      "step": 209
    },
    {
      "epoch": 0.15753938484621155,
      "grad_norm": 0.10698409378528595,
      "learning_rate": 7.865168539325843e-05,
      "loss": 0.0341,
      "step": 210
    },
    {
      "epoch": 0.15828957239309827,
      "grad_norm": 0.1937510073184967,
      "learning_rate": 7.902621722846442e-05,
      "loss": 0.0535,
      "step": 211
    },
    {
      "epoch": 0.159039759939985,
      "grad_norm": 0.25406792759895325,
      "learning_rate": 7.940074906367042e-05,
      "loss": 0.0622,
      "step": 212
    },
    {
      "epoch": 0.15978994748687173,
      "grad_norm": 0.16593150794506073,
      "learning_rate": 7.97752808988764e-05,
      "loss": 0.0371,
      "step": 213
    },
    {
      "epoch": 0.16054013503375844,
      "grad_norm": 0.17848573625087738,
      "learning_rate": 8.01498127340824e-05,
      "loss": 0.0414,
      "step": 214
    },
    {
      "epoch": 0.16129032258064516,
      "grad_norm": 0.1251983791589737,
      "learning_rate": 8.052434456928839e-05,
      "loss": 0.0355,
      "step": 215
    },
    {
      "epoch": 0.16204051012753187,
      "grad_norm": 0.2389572262763977,
      "learning_rate": 8.089887640449438e-05,
      "loss": 0.0433,
      "step": 216
    },
    {
      "epoch": 0.16279069767441862,
      "grad_norm": 0.13656052947044373,
      "learning_rate": 8.127340823970038e-05,
      "loss": 0.0394,
      "step": 217
    },
    {
      "epoch": 0.16354088522130533,
      "grad_norm": 0.1780516654253006,
      "learning_rate": 8.164794007490637e-05,
      "loss": 0.0553,
      "step": 218
    },
    {
      "epoch": 0.16429107276819205,
      "grad_norm": 0.13755282759666443,
      "learning_rate": 8.202247191011237e-05,
      "loss": 0.0348,
      "step": 219
    },
    {
      "epoch": 0.16504126031507876,
      "grad_norm": 0.35480576753616333,
      "learning_rate": 8.239700374531836e-05,
      "loss": 0.0675,
      "step": 220
    },
    {
      "epoch": 0.16579144786196548,
      "grad_norm": 0.13914087414741516,
      "learning_rate": 8.277153558052434e-05,
      "loss": 0.0403,
      "step": 221
    },
    {
      "epoch": 0.16654163540885222,
      "grad_norm": 0.1394328624010086,
      "learning_rate": 8.314606741573034e-05,
      "loss": 0.0321,
      "step": 222
    },
    {
      "epoch": 0.16729182295573894,
      "grad_norm": 0.10997188836336136,
      "learning_rate": 8.352059925093633e-05,
      "loss": 0.0386,
      "step": 223
    },
    {
      "epoch": 0.16804201050262565,
      "grad_norm": 0.10793869197368622,
      "learning_rate": 8.389513108614233e-05,
      "loss": 0.0323,
      "step": 224
    },
    {
      "epoch": 0.16879219804951237,
      "grad_norm": 0.119425468146801,
      "learning_rate": 8.426966292134831e-05,
      "loss": 0.0337,
      "step": 225
    },
    {
      "epoch": 0.1695423855963991,
      "grad_norm": 0.21268026530742645,
      "learning_rate": 8.46441947565543e-05,
      "loss": 0.0314,
      "step": 226
    },
    {
      "epoch": 0.17029257314328583,
      "grad_norm": 0.1542912721633911,
      "learning_rate": 8.50187265917603e-05,
      "loss": 0.0429,
      "step": 227
    },
    {
      "epoch": 0.17104276069017255,
      "grad_norm": 0.1182723417878151,
      "learning_rate": 8.53932584269663e-05,
      "loss": 0.0344,
      "step": 228
    },
    {
      "epoch": 0.17179294823705926,
      "grad_norm": 0.14600324630737305,
      "learning_rate": 8.576779026217229e-05,
      "loss": 0.0291,
      "step": 229
    },
    {
      "epoch": 0.17254313578394598,
      "grad_norm": 0.10251737385988235,
      "learning_rate": 8.614232209737829e-05,
      "loss": 0.0358,
      "step": 230
    },
    {
      "epoch": 0.17329332333083272,
      "grad_norm": 0.1304892748594284,
      "learning_rate": 8.651685393258427e-05,
      "loss": 0.0427,
      "step": 231
    },
    {
      "epoch": 0.17404351087771944,
      "grad_norm": 0.2422615885734558,
      "learning_rate": 8.689138576779026e-05,
      "loss": 0.0751,
      "step": 232
    },
    {
      "epoch": 0.17479369842460615,
      "grad_norm": 0.24723097681999207,
      "learning_rate": 8.726591760299626e-05,
      "loss": 0.0543,
      "step": 233
    },
    {
      "epoch": 0.17554388597149287,
      "grad_norm": 0.15995217859745026,
      "learning_rate": 8.764044943820225e-05,
      "loss": 0.04,
      "step": 234
    },
    {
      "epoch": 0.17629407351837958,
      "grad_norm": 0.17250633239746094,
      "learning_rate": 8.801498127340825e-05,
      "loss": 0.0568,
      "step": 235
    },
    {
      "epoch": 0.17704426106526633,
      "grad_norm": 0.1270926296710968,
      "learning_rate": 8.838951310861424e-05,
      "loss": 0.0427,
      "step": 236
    },
    {
      "epoch": 0.17779444861215304,
      "grad_norm": 0.22896043956279755,
      "learning_rate": 8.876404494382022e-05,
      "loss": 0.0633,
      "step": 237
    },
    {
      "epoch": 0.17854463615903976,
      "grad_norm": 0.09588818997144699,
      "learning_rate": 8.913857677902622e-05,
      "loss": 0.0362,
      "step": 238
    },
    {
      "epoch": 0.17929482370592648,
      "grad_norm": 0.10814792662858963,
      "learning_rate": 8.951310861423221e-05,
      "loss": 0.0361,
      "step": 239
    },
    {
      "epoch": 0.1800450112528132,
      "grad_norm": 0.10366350412368774,
      "learning_rate": 8.988764044943821e-05,
      "loss": 0.0279,
      "step": 240
    },
    {
      "epoch": 0.18079519879969994,
      "grad_norm": 0.17119480669498444,
      "learning_rate": 9.02621722846442e-05,
      "loss": 0.047,
      "step": 241
    },
    {
      "epoch": 0.18154538634658665,
      "grad_norm": 0.13574019074440002,
      "learning_rate": 9.063670411985018e-05,
      "loss": 0.0395,
      "step": 242
    },
    {
      "epoch": 0.18229557389347337,
      "grad_norm": 0.16211435198783875,
      "learning_rate": 9.101123595505618e-05,
      "loss": 0.0473,
      "step": 243
    },
    {
      "epoch": 0.18304576144036008,
      "grad_norm": 0.21162639558315277,
      "learning_rate": 9.138576779026217e-05,
      "loss": 0.0666,
      "step": 244
    },
    {
      "epoch": 0.1837959489872468,
      "grad_norm": 0.18772201240062714,
      "learning_rate": 9.176029962546817e-05,
      "loss": 0.0421,
      "step": 245
    },
    {
      "epoch": 0.18454613653413354,
      "grad_norm": 0.11692129820585251,
      "learning_rate": 9.213483146067416e-05,
      "loss": 0.0393,
      "step": 246
    },
    {
      "epoch": 0.18529632408102026,
      "grad_norm": 0.17434294521808624,
      "learning_rate": 9.250936329588016e-05,
      "loss": 0.0417,
      "step": 247
    },
    {
      "epoch": 0.18604651162790697,
      "grad_norm": 0.10198646783828735,
      "learning_rate": 9.288389513108615e-05,
      "loss": 0.0236,
      "step": 248
    },
    {
      "epoch": 0.1867966991747937,
      "grad_norm": 0.13234616816043854,
      "learning_rate": 9.325842696629214e-05,
      "loss": 0.0229,
      "step": 249
    },
    {
      "epoch": 0.18754688672168043,
      "grad_norm": 0.08883968740701675,
      "learning_rate": 9.363295880149813e-05,
      "loss": 0.0274,
      "step": 250
    },
    {
      "epoch": 0.18829707426856715,
      "grad_norm": 0.16140736639499664,
      "learning_rate": 9.400749063670413e-05,
      "loss": 0.0357,
      "step": 251
    },
    {
      "epoch": 0.18904726181545387,
      "grad_norm": 0.1592225581407547,
      "learning_rate": 9.438202247191012e-05,
      "loss": 0.0338,
      "step": 252
    },
    {
      "epoch": 0.18979744936234058,
      "grad_norm": 0.21853357553482056,
      "learning_rate": 9.47565543071161e-05,
      "loss": 0.0502,
      "step": 253
    },
    {
      "epoch": 0.1905476369092273,
      "grad_norm": 0.11385892331600189,
      "learning_rate": 9.51310861423221e-05,
      "loss": 0.0283,
      "step": 254
    },
    {
      "epoch": 0.19129782445611404,
      "grad_norm": 0.12962888181209564,
      "learning_rate": 9.550561797752809e-05,
      "loss": 0.0355,
      "step": 255
    },
    {
      "epoch": 0.19204801200300076,
      "grad_norm": 0.15307216346263885,
      "learning_rate": 9.588014981273409e-05,
      "loss": 0.0412,
      "step": 256
    },
    {
      "epoch": 0.19279819954988747,
      "grad_norm": 0.20015685260295868,
      "learning_rate": 9.625468164794008e-05,
      "loss": 0.0371,
      "step": 257
    },
    {
      "epoch": 0.1935483870967742,
      "grad_norm": 0.2568797469139099,
      "learning_rate": 9.662921348314608e-05,
      "loss": 0.0465,
      "step": 258
    },
    {
      "epoch": 0.1942985746436609,
      "grad_norm": 0.09511401504278183,
      "learning_rate": 9.700374531835207e-05,
      "loss": 0.029,
      "step": 259
    },
    {
      "epoch": 0.19504876219054765,
      "grad_norm": 0.19397354125976562,
      "learning_rate": 9.737827715355807e-05,
      "loss": 0.0422,
      "step": 260
    },
    {
      "epoch": 0.19579894973743436,
      "grad_norm": 0.15164721012115479,
      "learning_rate": 9.775280898876405e-05,
      "loss": 0.0454,
      "step": 261
    },
    {
      "epoch": 0.19654913728432108,
      "grad_norm": 0.11148524284362793,
      "learning_rate": 9.812734082397004e-05,
      "loss": 0.0347,
      "step": 262
    },
    {
      "epoch": 0.1972993248312078,
      "grad_norm": 0.09554921835660934,
      "learning_rate": 9.850187265917602e-05,
      "loss": 0.0259,
      "step": 263
    },
    {
      "epoch": 0.1980495123780945,
      "grad_norm": 0.35856032371520996,
      "learning_rate": 9.887640449438202e-05,
      "loss": 0.0481,
      "step": 264
    },
    {
      "epoch": 0.19879969992498125,
      "grad_norm": 0.15284886956214905,
      "learning_rate": 9.925093632958801e-05,
      "loss": 0.0475,
      "step": 265
    },
    {
      "epoch": 0.19954988747186797,
      "grad_norm": 0.0865863636136055,
      "learning_rate": 9.962546816479401e-05,
      "loss": 0.0302,
      "step": 266
    },
    {
      "epoch": 0.2003000750187547,
      "grad_norm": 0.1828830987215042,
      "learning_rate": 0.0001,
      "loss": 0.0514,
      "step": 267
    },
    {
      "epoch": 0.2010502625656414,
      "grad_norm": 0.08779073506593704,
      "learning_rate": 9.99999571274618e-05,
      "loss": 0.0198,
      "step": 268
    },
    {
      "epoch": 0.20180045011252815,
      "grad_norm": 0.1077849492430687,
      "learning_rate": 9.999982850992069e-05,
      "loss": 0.0313,
      "step": 269
    },
    {
      "epoch": 0.20255063765941486,
      "grad_norm": 0.17151670157909393,
      "learning_rate": 9.999961414759727e-05,
      "loss": 0.0454,
      "step": 270
    },
    {
      "epoch": 0.20330082520630158,
      "grad_norm": 0.09756535291671753,
      "learning_rate": 9.999931404085912e-05,
      "loss": 0.0206,
      "step": 271
    },
    {
      "epoch": 0.2040510127531883,
      "grad_norm": 0.18158778548240662,
      "learning_rate": 9.999892819022092e-05,
      "loss": 0.0455,
      "step": 272
    },
    {
      "epoch": 0.204801200300075,
      "grad_norm": 0.10948951542377472,
      "learning_rate": 9.999845659634435e-05,
      "loss": 0.0337,
      "step": 273
    },
    {
      "epoch": 0.20555138784696175,
      "grad_norm": 0.13912096619606018,
      "learning_rate": 9.999789926003814e-05,
      "loss": 0.0423,
      "step": 274
    },
    {
      "epoch": 0.20630157539384847,
      "grad_norm": 0.08084128797054291,
      "learning_rate": 9.999725618225808e-05,
      "loss": 0.0336,
      "step": 275
    },
    {
      "epoch": 0.20705176294073518,
      "grad_norm": 0.1288767158985138,
      "learning_rate": 9.999652736410698e-05,
      "loss": 0.0317,
      "step": 276
    },
    {
      "epoch": 0.2078019504876219,
      "grad_norm": 0.14830724895000458,
      "learning_rate": 9.999571280683468e-05,
      "loss": 0.0465,
      "step": 277
    },
    {
      "epoch": 0.20855213803450862,
      "grad_norm": 0.14257365465164185,
      "learning_rate": 9.99948125118381e-05,
      "loss": 0.0442,
      "step": 278
    },
    {
      "epoch": 0.20930232558139536,
      "grad_norm": 0.09027249366044998,
      "learning_rate": 9.999382648066113e-05,
      "loss": 0.0319,
      "step": 279
    },
    {
      "epoch": 0.21005251312828208,
      "grad_norm": 0.10757441818714142,
      "learning_rate": 9.999275471499472e-05,
      "loss": 0.0397,
      "step": 280
    },
    {
      "epoch": 0.2108027006751688,
      "grad_norm": 0.12700489163398743,
      "learning_rate": 9.999159721667685e-05,
      "loss": 0.0374,
      "step": 281
    },
    {
      "epoch": 0.2115528882220555,
      "grad_norm": 0.10365627706050873,
      "learning_rate": 9.999035398769252e-05,
      "loss": 0.0346,
      "step": 282
    },
    {
      "epoch": 0.21230307576894222,
      "grad_norm": 0.14833049476146698,
      "learning_rate": 9.998902503017372e-05,
      "loss": 0.054,
      "step": 283
    },
    {
      "epoch": 0.21305326331582897,
      "grad_norm": 0.21913230419158936,
      "learning_rate": 9.99876103463995e-05,
      "loss": 0.0505,
      "step": 284
    },
    {
      "epoch": 0.21380345086271568,
      "grad_norm": 0.173373743891716,
      "learning_rate": 9.998610993879589e-05,
      "loss": 0.0362,
      "step": 285
    },
    {
      "epoch": 0.2145536384096024,
      "grad_norm": 0.14817701280117035,
      "learning_rate": 9.998452380993597e-05,
      "loss": 0.0339,
      "step": 286
    },
    {
      "epoch": 0.21530382595648911,
      "grad_norm": 0.14449955523014069,
      "learning_rate": 9.998285196253977e-05,
      "loss": 0.0427,
      "step": 287
    },
    {
      "epoch": 0.21605401350337583,
      "grad_norm": 0.11007577925920486,
      "learning_rate": 9.998109439947434e-05,
      "loss": 0.0241,
      "step": 288
    },
    {
      "epoch": 0.21680420105026257,
      "grad_norm": 0.10484791547060013,
      "learning_rate": 9.997925112375375e-05,
      "loss": 0.029,
      "step": 289
    },
    {
      "epoch": 0.2175543885971493,
      "grad_norm": 0.27984753251075745,
      "learning_rate": 9.997732213853902e-05,
      "loss": 0.0493,
      "step": 290
    },
    {
      "epoch": 0.218304576144036,
      "grad_norm": 0.4144909977912903,
      "learning_rate": 9.997530744713817e-05,
      "loss": 0.075,
      "step": 291
    },
    {
      "epoch": 0.21905476369092272,
      "grad_norm": 0.14818748831748962,
      "learning_rate": 9.997320705300621e-05,
      "loss": 0.0465,
      "step": 292
    },
    {
      "epoch": 0.21980495123780946,
      "grad_norm": 0.11120837181806564,
      "learning_rate": 9.997102095974508e-05,
      "loss": 0.0297,
      "step": 293
    },
    {
      "epoch": 0.22055513878469618,
      "grad_norm": 0.10603461414575577,
      "learning_rate": 9.996874917110378e-05,
      "loss": 0.0335,
      "step": 294
    },
    {
      "epoch": 0.2213053263315829,
      "grad_norm": 0.11525816470384598,
      "learning_rate": 9.996639169097811e-05,
      "loss": 0.0357,
      "step": 295
    },
    {
      "epoch": 0.2220555138784696,
      "grad_norm": 0.26272228360176086,
      "learning_rate": 9.996394852341098e-05,
      "loss": 0.0636,
      "step": 296
    },
    {
      "epoch": 0.22280570142535633,
      "grad_norm": 0.12792189419269562,
      "learning_rate": 9.996141967259218e-05,
      "loss": 0.0389,
      "step": 297
    },
    {
      "epoch": 0.22355588897224307,
      "grad_norm": 0.13640092313289642,
      "learning_rate": 9.995880514285841e-05,
      "loss": 0.043,
      "step": 298
    },
    {
      "epoch": 0.2243060765191298,
      "grad_norm": 0.08026789128780365,
      "learning_rate": 9.995610493869336e-05,
      "loss": 0.0333,
      "step": 299
    },
    {
      "epoch": 0.2250562640660165,
      "grad_norm": 0.12414755672216415,
      "learning_rate": 9.99533190647276e-05,
      "loss": 0.0428,
      "step": 300
    },
    {
      "epoch": 0.22580645161290322,
      "grad_norm": 0.1047615334391594,
      "learning_rate": 9.995044752573864e-05,
      "loss": 0.0301,
      "step": 301
    },
    {
      "epoch": 0.22655663915978994,
      "grad_norm": 0.09507095813751221,
      "learning_rate": 9.994749032665085e-05,
      "loss": 0.0299,
      "step": 302
    },
    {
      "epoch": 0.22730682670667668,
      "grad_norm": 0.1469980925321579,
      "learning_rate": 9.994444747253559e-05,
      "loss": 0.0408,
      "step": 303
    },
    {
      "epoch": 0.2280570142535634,
      "grad_norm": 0.06208149716258049,
      "learning_rate": 9.9941318968611e-05,
      "loss": 0.0199,
      "step": 304
    },
    {
      "epoch": 0.2288072018004501,
      "grad_norm": 0.12007199227809906,
      "learning_rate": 9.993810482024221e-05,
      "loss": 0.0442,
      "step": 305
    },
    {
      "epoch": 0.22955738934733683,
      "grad_norm": 0.10592085868120193,
      "learning_rate": 9.993480503294114e-05,
      "loss": 0.0375,
      "step": 306
    },
    {
      "epoch": 0.23030757689422354,
      "grad_norm": 0.16347835958003998,
      "learning_rate": 9.993141961236661e-05,
      "loss": 0.0341,
      "step": 307
    },
    {
      "epoch": 0.23105776444111029,
      "grad_norm": 0.10255320370197296,
      "learning_rate": 9.992794856432426e-05,
      "loss": 0.0312,
      "step": 308
    },
    {
      "epoch": 0.231807951987997,
      "grad_norm": 0.18392932415008545,
      "learning_rate": 9.992439189476661e-05,
      "loss": 0.0518,
      "step": 309
    },
    {
      "epoch": 0.23255813953488372,
      "grad_norm": 0.08901845663785934,
      "learning_rate": 9.992074960979301e-05,
      "loss": 0.0268,
      "step": 310
    },
    {
      "epoch": 0.23330832708177043,
      "grad_norm": 0.14844289422035217,
      "learning_rate": 9.991702171564961e-05,
      "loss": 0.0391,
      "step": 311
    },
    {
      "epoch": 0.23405851462865718,
      "grad_norm": 0.18847529590129852,
      "learning_rate": 9.991320821872939e-05,
      "loss": 0.0418,
      "step": 312
    },
    {
      "epoch": 0.2348087021755439,
      "grad_norm": 0.12273112684488297,
      "learning_rate": 9.990930912557209e-05,
      "loss": 0.0403,
      "step": 313
    },
    {
      "epoch": 0.2355588897224306,
      "grad_norm": 0.08602188527584076,
      "learning_rate": 9.990532444286431e-05,
      "loss": 0.0231,
      "step": 314
    },
    {
      "epoch": 0.23630907726931732,
      "grad_norm": 0.18314284086227417,
      "learning_rate": 9.990125417743937e-05,
      "loss": 0.0357,
      "step": 315
    },
    {
      "epoch": 0.23705926481620404,
      "grad_norm": 0.11995340138673782,
      "learning_rate": 9.989709833627736e-05,
      "loss": 0.0341,
      "step": 316
    },
    {
      "epoch": 0.23780945236309078,
      "grad_norm": 0.20460061728954315,
      "learning_rate": 9.989285692650518e-05,
      "loss": 0.0464,
      "step": 317
    },
    {
      "epoch": 0.2385596399099775,
      "grad_norm": 0.1889379769563675,
      "learning_rate": 9.98885299553964e-05,
      "loss": 0.0517,
      "step": 318
    },
    {
      "epoch": 0.23930982745686422,
      "grad_norm": 0.09578392654657364,
      "learning_rate": 9.988411743037134e-05,
      "loss": 0.027,
      "step": 319
    },
    {
      "epoch": 0.24006001500375093,
      "grad_norm": 0.12053310871124268,
      "learning_rate": 9.987961935899706e-05,
      "loss": 0.0319,
      "step": 320
    },
    {
      "epoch": 0.24081020255063765,
      "grad_norm": 0.21278013288974762,
      "learning_rate": 9.987503574898731e-05,
      "loss": 0.0468,
      "step": 321
    },
    {
      "epoch": 0.2415603900975244,
      "grad_norm": 0.08968649059534073,
      "learning_rate": 9.987036660820255e-05,
      "loss": 0.025,
      "step": 322
    },
    {
      "epoch": 0.2423105776444111,
      "grad_norm": 0.12514202296733856,
      "learning_rate": 9.986561194464985e-05,
      "loss": 0.0367,
      "step": 323
    },
    {
      "epoch": 0.24306076519129782,
      "grad_norm": 0.10158783197402954,
      "learning_rate": 9.986077176648303e-05,
      "loss": 0.0317,
      "step": 324
    },
    {
      "epoch": 0.24381095273818454,
      "grad_norm": 0.1138206347823143,
      "learning_rate": 9.985584608200251e-05,
      "loss": 0.0312,
      "step": 325
    },
    {
      "epoch": 0.24456114028507125,
      "grad_norm": 0.07617976516485214,
      "learning_rate": 9.985083489965534e-05,
      "loss": 0.0213,
      "step": 326
    },
    {
      "epoch": 0.245311327831958,
      "grad_norm": 0.07596145570278168,
      "learning_rate": 9.984573822803521e-05,
      "loss": 0.024,
      "step": 327
    },
    {
      "epoch": 0.24606151537884471,
      "grad_norm": 0.11473222076892853,
      "learning_rate": 9.984055607588242e-05,
      "loss": 0.0321,
      "step": 328
    },
    {
      "epoch": 0.24681170292573143,
      "grad_norm": 0.10172963887453079,
      "learning_rate": 9.983528845208384e-05,
      "loss": 0.0298,
      "step": 329
    },
    {
      "epoch": 0.24756189047261815,
      "grad_norm": 0.16156165301799774,
      "learning_rate": 9.982993536567293e-05,
      "loss": 0.0401,
      "step": 330
    },
    {
      "epoch": 0.2483120780195049,
      "grad_norm": 0.09021109342575073,
      "learning_rate": 9.98244968258297e-05,
      "loss": 0.0255,
      "step": 331
    },
    {
      "epoch": 0.2490622655663916,
      "grad_norm": 0.14055666327476501,
      "learning_rate": 9.981897284188073e-05,
      "loss": 0.0484,
      "step": 332
    },
    {
      "epoch": 0.24981245311327832,
      "grad_norm": 0.1804196536540985,
      "learning_rate": 9.981336342329909e-05,
      "loss": 0.0328,
      "step": 333
    },
    {
      "epoch": 0.25056264066016504,
      "grad_norm": 0.10542705655097961,
      "learning_rate": 9.980766857970438e-05,
      "loss": 0.0335,
      "step": 334
    },
    {
      "epoch": 0.25131282820705175,
      "grad_norm": 0.23658208549022675,
      "learning_rate": 9.98018883208627e-05,
      "loss": 0.0511,
      "step": 335
    },
    {
      "epoch": 0.25206301575393847,
      "grad_norm": 0.08560968190431595,
      "learning_rate": 9.979602265668664e-05,
      "loss": 0.023,
      "step": 336
    },
    {
      "epoch": 0.2528132033008252,
      "grad_norm": 0.1381518393754959,
      "learning_rate": 9.979007159723521e-05,
      "loss": 0.0399,
      "step": 337
    },
    {
      "epoch": 0.2535633908477119,
      "grad_norm": 0.12532010674476624,
      "learning_rate": 9.97840351527139e-05,
      "loss": 0.0321,
      "step": 338
    },
    {
      "epoch": 0.25431357839459867,
      "grad_norm": 0.1004038155078888,
      "learning_rate": 9.977791333347462e-05,
      "loss": 0.034,
      "step": 339
    },
    {
      "epoch": 0.2550637659414854,
      "grad_norm": 0.16208575665950775,
      "learning_rate": 9.97717061500157e-05,
      "loss": 0.0403,
      "step": 340
    },
    {
      "epoch": 0.2558139534883721,
      "grad_norm": 0.06728474050760269,
      "learning_rate": 9.976541361298184e-05,
      "loss": 0.0233,
      "step": 341
    },
    {
      "epoch": 0.2565641410352588,
      "grad_norm": 0.17869628965854645,
      "learning_rate": 9.97590357331641e-05,
      "loss": 0.0395,
      "step": 342
    },
    {
      "epoch": 0.25731432858214554,
      "grad_norm": 0.1622343361377716,
      "learning_rate": 9.975257252149994e-05,
      "loss": 0.0381,
      "step": 343
    },
    {
      "epoch": 0.25806451612903225,
      "grad_norm": 0.16202475130558014,
      "learning_rate": 9.974602398907313e-05,
      "loss": 0.0506,
      "step": 344
    },
    {
      "epoch": 0.25881470367591897,
      "grad_norm": 0.17167797684669495,
      "learning_rate": 9.973939014711375e-05,
      "loss": 0.0503,
      "step": 345
    },
    {
      "epoch": 0.2595648912228057,
      "grad_norm": 0.2033884972333908,
      "learning_rate": 9.973267100699819e-05,
      "loss": 0.0424,
      "step": 346
    },
    {
      "epoch": 0.2603150787696924,
      "grad_norm": 0.16165803372859955,
      "learning_rate": 9.972586658024911e-05,
      "loss": 0.0471,
      "step": 347
    },
    {
      "epoch": 0.26106526631657917,
      "grad_norm": 0.11009008437395096,
      "learning_rate": 9.971897687853544e-05,
      "loss": 0.0302,
      "step": 348
    },
    {
      "epoch": 0.2618154538634659,
      "grad_norm": 0.17024879157543182,
      "learning_rate": 9.971200191367234e-05,
      "loss": 0.0363,
      "step": 349
    },
    {
      "epoch": 0.2625656414103526,
      "grad_norm": 0.1899632215499878,
      "learning_rate": 9.970494169762117e-05,
      "loss": 0.051,
      "step": 350
    },
    {
      "epoch": 0.2633158289572393,
      "grad_norm": 0.18265698850154877,
      "learning_rate": 9.969779624248954e-05,
      "loss": 0.0434,
      "step": 351
    },
    {
      "epoch": 0.26406601650412603,
      "grad_norm": 0.21099521219730377,
      "learning_rate": 9.969056556053116e-05,
      "loss": 0.0552,
      "step": 352
    },
    {
      "epoch": 0.26481620405101275,
      "grad_norm": 0.12291638553142548,
      "learning_rate": 9.968324966414597e-05,
      "loss": 0.0413,
      "step": 353
    },
    {
      "epoch": 0.26556639159789946,
      "grad_norm": 0.1895735114812851,
      "learning_rate": 9.967584856588e-05,
      "loss": 0.0581,
      "step": 354
    },
    {
      "epoch": 0.2663165791447862,
      "grad_norm": 0.1934838891029358,
      "learning_rate": 9.966836227842538e-05,
      "loss": 0.0438,
      "step": 355
    },
    {
      "epoch": 0.2670667666916729,
      "grad_norm": 0.11946920305490494,
      "learning_rate": 9.96607908146204e-05,
      "loss": 0.0367,
      "step": 356
    },
    {
      "epoch": 0.2678169542385596,
      "grad_norm": 0.10245028138160706,
      "learning_rate": 9.965313418744935e-05,
      "loss": 0.0318,
      "step": 357
    },
    {
      "epoch": 0.2685671417854464,
      "grad_norm": 0.07515406608581543,
      "learning_rate": 9.964539241004261e-05,
      "loss": 0.0243,
      "step": 358
    },
    {
      "epoch": 0.2693173293323331,
      "grad_norm": 0.2560199797153473,
      "learning_rate": 9.963756549567654e-05,
      "loss": 0.0573,
      "step": 359
    },
    {
      "epoch": 0.2700675168792198,
      "grad_norm": 0.160162553191185,
      "learning_rate": 9.962965345777353e-05,
      "loss": 0.0478,
      "step": 360
    },
    {
      "epoch": 0.27081770442610653,
      "grad_norm": 0.13862910866737366,
      "learning_rate": 9.962165630990196e-05,
      "loss": 0.0525,
      "step": 361
    },
    {
      "epoch": 0.27156789197299325,
      "grad_norm": 0.1421460658311844,
      "learning_rate": 9.961357406577617e-05,
      "loss": 0.0388,
      "step": 362
    },
    {
      "epoch": 0.27231807951987996,
      "grad_norm": 0.07698662579059601,
      "learning_rate": 9.960540673925636e-05,
      "loss": 0.0265,
      "step": 363
    },
    {
      "epoch": 0.2730682670667667,
      "grad_norm": 0.12471983581781387,
      "learning_rate": 9.959715434434873e-05,
      "loss": 0.0497,
      "step": 364
    },
    {
      "epoch": 0.2738184546136534,
      "grad_norm": 0.10801269859075546,
      "learning_rate": 9.958881689520531e-05,
      "loss": 0.0328,
      "step": 365
    },
    {
      "epoch": 0.2745686421605401,
      "grad_norm": 0.12907591462135315,
      "learning_rate": 9.958039440612402e-05,
      "loss": 0.0411,
      "step": 366
    },
    {
      "epoch": 0.2753188297074269,
      "grad_norm": 0.15533341467380524,
      "learning_rate": 9.957188689154859e-05,
      "loss": 0.0404,
      "step": 367
    },
    {
      "epoch": 0.2760690172543136,
      "grad_norm": 0.11384189873933792,
      "learning_rate": 9.956329436606857e-05,
      "loss": 0.0385,
      "step": 368
    },
    {
      "epoch": 0.2768192048012003,
      "grad_norm": 0.11162270605564117,
      "learning_rate": 9.955461684441928e-05,
      "loss": 0.0416,
      "step": 369
    },
    {
      "epoch": 0.27756939234808703,
      "grad_norm": 0.19340857863426208,
      "learning_rate": 9.954585434148183e-05,
      "loss": 0.0458,
      "step": 370
    },
    {
      "epoch": 0.27831957989497375,
      "grad_norm": 0.10425236821174622,
      "learning_rate": 9.953700687228306e-05,
      "loss": 0.0336,
      "step": 371
    },
    {
      "epoch": 0.27906976744186046,
      "grad_norm": 0.1434914767742157,
      "learning_rate": 9.952807445199549e-05,
      "loss": 0.0354,
      "step": 372
    },
    {
      "epoch": 0.2798199549887472,
      "grad_norm": 0.08324921876192093,
      "learning_rate": 9.951905709593735e-05,
      "loss": 0.0241,
      "step": 373
    },
    {
      "epoch": 0.2805701425356339,
      "grad_norm": 0.1119835302233696,
      "learning_rate": 9.950995481957251e-05,
      "loss": 0.0345,
      "step": 374
    },
    {
      "epoch": 0.2813203300825206,
      "grad_norm": 0.09572914987802505,
      "learning_rate": 9.950076763851049e-05,
      "loss": 0.0305,
      "step": 375
    },
    {
      "epoch": 0.2820705176294073,
      "grad_norm": 0.10482971370220184,
      "learning_rate": 9.949149556850638e-05,
      "loss": 0.0358,
      "step": 376
    },
    {
      "epoch": 0.2828207051762941,
      "grad_norm": 0.14237084984779358,
      "learning_rate": 9.94821386254609e-05,
      "loss": 0.0338,
      "step": 377
    },
    {
      "epoch": 0.2835708927231808,
      "grad_norm": 0.10478667169809341,
      "learning_rate": 9.947269682542027e-05,
      "loss": 0.0249,
      "step": 378
    },
    {
      "epoch": 0.2843210802700675,
      "grad_norm": 0.11596731841564178,
      "learning_rate": 9.946317018457622e-05,
      "loss": 0.0286,
      "step": 379
    },
    {
      "epoch": 0.28507126781695424,
      "grad_norm": 0.11092496663331985,
      "learning_rate": 9.945355871926605e-05,
      "loss": 0.0333,
      "step": 380
    },
    {
      "epoch": 0.28582145536384096,
      "grad_norm": 0.14135751128196716,
      "learning_rate": 9.944386244597244e-05,
      "loss": 0.0348,
      "step": 381
    },
    {
      "epoch": 0.2865716429107277,
      "grad_norm": 0.1096055880188942,
      "learning_rate": 9.943408138132357e-05,
      "loss": 0.0382,
      "step": 382
    },
    {
      "epoch": 0.2873218304576144,
      "grad_norm": 0.07303545624017715,
      "learning_rate": 9.942421554209297e-05,
      "loss": 0.0237,
      "step": 383
    },
    {
      "epoch": 0.2880720180045011,
      "grad_norm": 0.1861746609210968,
      "learning_rate": 9.94142649451996e-05,
      "loss": 0.0462,
      "step": 384
    },
    {
      "epoch": 0.2888222055513878,
      "grad_norm": 0.11650864779949188,
      "learning_rate": 9.940422960770776e-05,
      "loss": 0.0374,
      "step": 385
    },
    {
      "epoch": 0.2895723930982746,
      "grad_norm": 0.13004302978515625,
      "learning_rate": 9.939410954682706e-05,
      "loss": 0.0275,
      "step": 386
    },
    {
      "epoch": 0.2903225806451613,
      "grad_norm": 0.13394056260585785,
      "learning_rate": 9.938390477991242e-05,
      "loss": 0.0402,
      "step": 387
    },
    {
      "epoch": 0.291072768192048,
      "grad_norm": 0.11594215780496597,
      "learning_rate": 9.937361532446399e-05,
      "loss": 0.0274,
      "step": 388
    },
    {
      "epoch": 0.29182295573893474,
      "grad_norm": 0.08473474532365799,
      "learning_rate": 9.936324119812719e-05,
      "loss": 0.0264,
      "step": 389
    },
    {
      "epoch": 0.29257314328582146,
      "grad_norm": 0.14290671050548553,
      "learning_rate": 9.93527824186926e-05,
      "loss": 0.0294,
      "step": 390
    },
    {
      "epoch": 0.2933233308327082,
      "grad_norm": 0.13471192121505737,
      "learning_rate": 9.934223900409603e-05,
      "loss": 0.029,
      "step": 391
    },
    {
      "epoch": 0.2940735183795949,
      "grad_norm": 0.11877185851335526,
      "learning_rate": 9.933161097241837e-05,
      "loss": 0.0295,
      "step": 392
    },
    {
      "epoch": 0.2948237059264816,
      "grad_norm": 0.14871439337730408,
      "learning_rate": 9.932089834188567e-05,
      "loss": 0.0354,
      "step": 393
    },
    {
      "epoch": 0.2955738934733683,
      "grad_norm": 0.12378165125846863,
      "learning_rate": 9.931010113086902e-05,
      "loss": 0.0348,
      "step": 394
    },
    {
      "epoch": 0.29632408102025504,
      "grad_norm": 0.15641850233078003,
      "learning_rate": 9.929921935788457e-05,
      "loss": 0.0414,
      "step": 395
    },
    {
      "epoch": 0.2970742685671418,
      "grad_norm": 0.07526502013206482,
      "learning_rate": 9.928825304159351e-05,
      "loss": 0.0277,
      "step": 396
    },
    {
      "epoch": 0.2978244561140285,
      "grad_norm": 0.13670268654823303,
      "learning_rate": 9.927720220080199e-05,
      "loss": 0.035,
      "step": 397
    },
    {
      "epoch": 0.29857464366091524,
      "grad_norm": 0.0850730836391449,
      "learning_rate": 9.926606685446109e-05,
      "loss": 0.0373,
      "step": 398
    },
    {
      "epoch": 0.29932483120780196,
      "grad_norm": 0.08862190693616867,
      "learning_rate": 9.925484702166686e-05,
      "loss": 0.0317,
      "step": 399
    },
    {
      "epoch": 0.30007501875468867,
      "grad_norm": 0.08291352540254593,
      "learning_rate": 9.924354272166017e-05,
      "loss": 0.0179,
      "step": 400
    },
    {
      "epoch": 0.30007501875468867,
      "eval_loss": 0.03630685806274414,
      "eval_runtime": 5.1209,
      "eval_samples_per_second": 10.545,
      "eval_steps_per_second": 2.734,
      "step": 400
    },
    {
      "epoch": 0.3008252063015754,
      "grad_norm": 0.1142362654209137,
      "learning_rate": 9.923215397382684e-05,
      "loss": 0.0294,
      "step": 401
    },
    {
      "epoch": 0.3015753938484621,
      "grad_norm": 0.08515766263008118,
      "learning_rate": 9.92206807976974e-05,
      "loss": 0.0208,
      "step": 402
    },
    {
      "epoch": 0.3023255813953488,
      "grad_norm": 0.1859070360660553,
      "learning_rate": 9.920912321294723e-05,
      "loss": 0.0405,
      "step": 403
    },
    {
      "epoch": 0.30307576894223553,
      "grad_norm": 0.0877511203289032,
      "learning_rate": 9.919748123939647e-05,
      "loss": 0.0276,
      "step": 404
    },
    {
      "epoch": 0.3038259564891223,
      "grad_norm": 0.14317767322063446,
      "learning_rate": 9.918575489700993e-05,
      "loss": 0.0405,
      "step": 405
    },
    {
      "epoch": 0.304576144036009,
      "grad_norm": 0.1826557219028473,
      "learning_rate": 9.917394420589716e-05,
      "loss": 0.034,
      "step": 406
    },
    {
      "epoch": 0.30532633158289574,
      "grad_norm": 0.08531735837459564,
      "learning_rate": 9.916204918631231e-05,
      "loss": 0.031,
      "step": 407
    },
    {
      "epoch": 0.30607651912978245,
      "grad_norm": 0.13493755459785461,
      "learning_rate": 9.915006985865416e-05,
      "loss": 0.0298,
      "step": 408
    },
    {
      "epoch": 0.30682670667666917,
      "grad_norm": 0.10673831403255463,
      "learning_rate": 9.913800624346612e-05,
      "loss": 0.0313,
      "step": 409
    },
    {
      "epoch": 0.3075768942235559,
      "grad_norm": 0.16792036592960358,
      "learning_rate": 9.912585836143606e-05,
      "loss": 0.0457,
      "step": 410
    },
    {
      "epoch": 0.3083270817704426,
      "grad_norm": 0.11888712644577026,
      "learning_rate": 9.911362623339642e-05,
      "loss": 0.0337,
      "step": 411
    },
    {
      "epoch": 0.3090772693173293,
      "grad_norm": 0.13570362329483032,
      "learning_rate": 9.91013098803241e-05,
      "loss": 0.0408,
      "step": 412
    },
    {
      "epoch": 0.30982745686421603,
      "grad_norm": 0.11306290328502655,
      "learning_rate": 9.908890932334042e-05,
      "loss": 0.0345,
      "step": 413
    },
    {
      "epoch": 0.31057764441110275,
      "grad_norm": 0.1427725851535797,
      "learning_rate": 9.907642458371111e-05,
      "loss": 0.0429,
      "step": 414
    },
    {
      "epoch": 0.3113278319579895,
      "grad_norm": 0.18175150454044342,
      "learning_rate": 9.906385568284629e-05,
      "loss": 0.0428,
      "step": 415
    },
    {
      "epoch": 0.31207801950487624,
      "grad_norm": 0.1440231204032898,
      "learning_rate": 9.905120264230036e-05,
      "loss": 0.0421,
      "step": 416
    },
    {
      "epoch": 0.31282820705176295,
      "grad_norm": 0.11851240694522858,
      "learning_rate": 9.903846548377206e-05,
      "loss": 0.0291,
      "step": 417
    },
    {
      "epoch": 0.31357839459864967,
      "grad_norm": 0.18592765927314758,
      "learning_rate": 9.902564422910436e-05,
      "loss": 0.0565,
      "step": 418
    },
    {
      "epoch": 0.3143285821455364,
      "grad_norm": 0.15560060739517212,
      "learning_rate": 9.901273890028444e-05,
      "loss": 0.0378,
      "step": 419
    },
    {
      "epoch": 0.3150787696924231,
      "grad_norm": 0.1025664433836937,
      "learning_rate": 9.899974951944367e-05,
      "loss": 0.0271,
      "step": 420
    },
    {
      "epoch": 0.3158289572393098,
      "grad_norm": 0.11080783605575562,
      "learning_rate": 9.898667610885757e-05,
      "loss": 0.0287,
      "step": 421
    },
    {
      "epoch": 0.31657914478619653,
      "grad_norm": 0.13097728788852692,
      "learning_rate": 9.897351869094573e-05,
      "loss": 0.0388,
      "step": 422
    },
    {
      "epoch": 0.31732933233308325,
      "grad_norm": 0.13140304386615753,
      "learning_rate": 9.896027728827185e-05,
      "loss": 0.0451,
      "step": 423
    },
    {
      "epoch": 0.31807951987997,
      "grad_norm": 0.12562090158462524,
      "learning_rate": 9.894695192354362e-05,
      "loss": 0.0295,
      "step": 424
    },
    {
      "epoch": 0.31882970742685673,
      "grad_norm": 0.25448498129844666,
      "learning_rate": 9.893354261961274e-05,
      "loss": 0.0596,
      "step": 425
    },
    {
      "epoch": 0.31957989497374345,
      "grad_norm": 0.15546709299087524,
      "learning_rate": 9.892004939947482e-05,
      "loss": 0.0326,
      "step": 426
    },
    {
      "epoch": 0.32033008252063017,
      "grad_norm": 0.17915265262126923,
      "learning_rate": 9.890647228626944e-05,
      "loss": 0.0502,
      "step": 427
    },
    {
      "epoch": 0.3210802700675169,
      "grad_norm": 0.17885051667690277,
      "learning_rate": 9.889281130327997e-05,
      "loss": 0.0376,
      "step": 428
    },
    {
      "epoch": 0.3218304576144036,
      "grad_norm": 0.1253713220357895,
      "learning_rate": 9.887906647393368e-05,
      "loss": 0.033,
      "step": 429
    },
    {
      "epoch": 0.3225806451612903,
      "grad_norm": 0.10489869117736816,
      "learning_rate": 9.88652378218016e-05,
      "loss": 0.0314,
      "step": 430
    },
    {
      "epoch": 0.32333083270817703,
      "grad_norm": 0.15992328524589539,
      "learning_rate": 9.885132537059849e-05,
      "loss": 0.0396,
      "step": 431
    },
    {
      "epoch": 0.32408102025506375,
      "grad_norm": 0.08546454459428787,
      "learning_rate": 9.883732914418285e-05,
      "loss": 0.027,
      "step": 432
    },
    {
      "epoch": 0.32483120780195046,
      "grad_norm": 0.13443313539028168,
      "learning_rate": 9.882324916655681e-05,
      "loss": 0.0361,
      "step": 433
    },
    {
      "epoch": 0.32558139534883723,
      "grad_norm": 0.21272841095924377,
      "learning_rate": 9.880908546186616e-05,
      "loss": 0.0636,
      "step": 434
    },
    {
      "epoch": 0.32633158289572395,
      "grad_norm": 0.13114754855632782,
      "learning_rate": 9.879483805440027e-05,
      "loss": 0.0352,
      "step": 435
    },
    {
      "epoch": 0.32708177044261066,
      "grad_norm": 0.11861104518175125,
      "learning_rate": 9.8780506968592e-05,
      "loss": 0.0294,
      "step": 436
    },
    {
      "epoch": 0.3278319579894974,
      "grad_norm": 0.07027596980333328,
      "learning_rate": 9.876609222901781e-05,
      "loss": 0.0179,
      "step": 437
    },
    {
      "epoch": 0.3285821455363841,
      "grad_norm": 0.1381312608718872,
      "learning_rate": 9.875159386039749e-05,
      "loss": 0.0468,
      "step": 438
    },
    {
      "epoch": 0.3293323330832708,
      "grad_norm": 0.12370853126049042,
      "learning_rate": 9.873701188759438e-05,
      "loss": 0.044,
      "step": 439
    },
    {
      "epoch": 0.3300825206301575,
      "grad_norm": 0.12424737215042114,
      "learning_rate": 9.872234633561509e-05,
      "loss": 0.0432,
      "step": 440
    },
    {
      "epoch": 0.33083270817704424,
      "grad_norm": 0.09881707280874252,
      "learning_rate": 9.87075972296096e-05,
      "loss": 0.0286,
      "step": 441
    },
    {
      "epoch": 0.33158289572393096,
      "grad_norm": 0.11166679114103317,
      "learning_rate": 9.86927645948712e-05,
      "loss": 0.0392,
      "step": 442
    },
    {
      "epoch": 0.33233308327081773,
      "grad_norm": 0.0945601686835289,
      "learning_rate": 9.867784845683637e-05,
      "loss": 0.0339,
      "step": 443
    },
    {
      "epoch": 0.33308327081770445,
      "grad_norm": 0.13213807344436646,
      "learning_rate": 9.866284884108481e-05,
      "loss": 0.0415,
      "step": 444
    },
    {
      "epoch": 0.33383345836459116,
      "grad_norm": 0.12910525500774384,
      "learning_rate": 9.864776577333941e-05,
      "loss": 0.0405,
      "step": 445
    },
    {
      "epoch": 0.3345836459114779,
      "grad_norm": 0.07470346242189407,
      "learning_rate": 9.863259927946613e-05,
      "loss": 0.0243,
      "step": 446
    },
    {
      "epoch": 0.3353338334583646,
      "grad_norm": 0.10762255638837814,
      "learning_rate": 9.861734938547405e-05,
      "loss": 0.0223,
      "step": 447
    },
    {
      "epoch": 0.3360840210052513,
      "grad_norm": 0.11334626376628876,
      "learning_rate": 9.860201611751518e-05,
      "loss": 0.0396,
      "step": 448
    },
    {
      "epoch": 0.336834208552138,
      "grad_norm": 0.1437179446220398,
      "learning_rate": 9.858659950188458e-05,
      "loss": 0.0364,
      "step": 449
    },
    {
      "epoch": 0.33758439609902474,
      "grad_norm": 0.12946391105651855,
      "learning_rate": 9.857109956502027e-05,
      "loss": 0.0362,
      "step": 450
    },
    {
      "epoch": 0.33833458364591146,
      "grad_norm": 0.14568856358528137,
      "learning_rate": 9.855551633350306e-05,
      "loss": 0.0421,
      "step": 451
    },
    {
      "epoch": 0.3390847711927982,
      "grad_norm": 0.14602363109588623,
      "learning_rate": 9.853984983405668e-05,
      "loss": 0.0438,
      "step": 452
    },
    {
      "epoch": 0.33983495873968494,
      "grad_norm": 0.11090512573719025,
      "learning_rate": 9.852410009354766e-05,
      "loss": 0.0285,
      "step": 453
    },
    {
      "epoch": 0.34058514628657166,
      "grad_norm": 0.23449061810970306,
      "learning_rate": 9.850826713898521e-05,
      "loss": 0.0442,
      "step": 454
    },
    {
      "epoch": 0.3413353338334584,
      "grad_norm": 0.10066236555576324,
      "learning_rate": 9.849235099752132e-05,
      "loss": 0.038,
      "step": 455
    },
    {
      "epoch": 0.3420855213803451,
      "grad_norm": 0.14898282289505005,
      "learning_rate": 9.847635169645058e-05,
      "loss": 0.0331,
      "step": 456
    },
    {
      "epoch": 0.3428357089272318,
      "grad_norm": 0.17724516987800598,
      "learning_rate": 9.846026926321024e-05,
      "loss": 0.0434,
      "step": 457
    },
    {
      "epoch": 0.3435858964741185,
      "grad_norm": 0.11125597357749939,
      "learning_rate": 9.844410372538006e-05,
      "loss": 0.0286,
      "step": 458
    },
    {
      "epoch": 0.34433608402100524,
      "grad_norm": 0.13614225387573242,
      "learning_rate": 9.842785511068239e-05,
      "loss": 0.0407,
      "step": 459
    },
    {
      "epoch": 0.34508627156789196,
      "grad_norm": 0.14966627955436707,
      "learning_rate": 9.841152344698197e-05,
      "loss": 0.0244,
      "step": 460
    },
    {
      "epoch": 0.34583645911477867,
      "grad_norm": 0.140168234705925,
      "learning_rate": 9.8395108762286e-05,
      "loss": 0.0316,
      "step": 461
    },
    {
      "epoch": 0.34658664666166544,
      "grad_norm": 0.10689567774534225,
      "learning_rate": 9.837861108474404e-05,
      "loss": 0.0305,
      "step": 462
    },
    {
      "epoch": 0.34733683420855216,
      "grad_norm": 0.12915848195552826,
      "learning_rate": 9.8362030442648e-05,
      "loss": 0.0201,
      "step": 463
    },
    {
      "epoch": 0.3480870217554389,
      "grad_norm": 0.12917649745941162,
      "learning_rate": 9.834536686443204e-05,
      "loss": 0.0289,
      "step": 464
    },
    {
      "epoch": 0.3488372093023256,
      "grad_norm": 0.07282043993473053,
      "learning_rate": 9.832862037867257e-05,
      "loss": 0.0236,
      "step": 465
    },
    {
      "epoch": 0.3495873968492123,
      "grad_norm": 0.2947641611099243,
      "learning_rate": 9.831179101408813e-05,
      "loss": 0.0541,
      "step": 466
    },
    {
      "epoch": 0.350337584396099,
      "grad_norm": 0.1508365124464035,
      "learning_rate": 9.829487879953946e-05,
      "loss": 0.0361,
      "step": 467
    },
    {
      "epoch": 0.35108777194298574,
      "grad_norm": 0.14020177721977234,
      "learning_rate": 9.827788376402932e-05,
      "loss": 0.0381,
      "step": 468
    },
    {
      "epoch": 0.35183795948987245,
      "grad_norm": 0.12377166002988815,
      "learning_rate": 9.826080593670253e-05,
      "loss": 0.029,
      "step": 469
    },
    {
      "epoch": 0.35258814703675917,
      "grad_norm": 0.32600975036621094,
      "learning_rate": 9.82436453468459e-05,
      "loss": 0.0506,
      "step": 470
    },
    {
      "epoch": 0.3533383345836459,
      "grad_norm": 0.22649474442005157,
      "learning_rate": 9.822640202388812e-05,
      "loss": 0.0376,
      "step": 471
    },
    {
      "epoch": 0.35408852213053266,
      "grad_norm": 0.11873234808444977,
      "learning_rate": 9.820907599739979e-05,
      "loss": 0.0351,
      "step": 472
    },
    {
      "epoch": 0.3548387096774194,
      "grad_norm": 0.1265113800764084,
      "learning_rate": 9.819166729709336e-05,
      "loss": 0.0415,
      "step": 473
    },
    {
      "epoch": 0.3555888972243061,
      "grad_norm": 0.15935584902763367,
      "learning_rate": 9.817417595282304e-05,
      "loss": 0.0452,
      "step": 474
    },
    {
      "epoch": 0.3563390847711928,
      "grad_norm": 0.16310295462608337,
      "learning_rate": 9.815660199458476e-05,
      "loss": 0.0297,
      "step": 475
    },
    {
      "epoch": 0.3570892723180795,
      "grad_norm": 0.1761324107646942,
      "learning_rate": 9.81389454525161e-05,
      "loss": 0.0462,
      "step": 476
    },
    {
      "epoch": 0.35783945986496624,
      "grad_norm": 0.13628017902374268,
      "learning_rate": 9.812120635689632e-05,
      "loss": 0.0442,
      "step": 477
    },
    {
      "epoch": 0.35858964741185295,
      "grad_norm": 0.17877016961574554,
      "learning_rate": 9.810338473814621e-05,
      "loss": 0.0537,
      "step": 478
    },
    {
      "epoch": 0.35933983495873967,
      "grad_norm": 0.10660474747419357,
      "learning_rate": 9.808548062682812e-05,
      "loss": 0.0316,
      "step": 479
    },
    {
      "epoch": 0.3600900225056264,
      "grad_norm": 0.11876129359006882,
      "learning_rate": 9.80674940536458e-05,
      "loss": 0.0343,
      "step": 480
    },
    {
      "epoch": 0.36084021005251316,
      "grad_norm": 0.08890905231237411,
      "learning_rate": 9.804942504944445e-05,
      "loss": 0.0267,
      "step": 481
    },
    {
      "epoch": 0.36159039759939987,
      "grad_norm": 0.12817738950252533,
      "learning_rate": 9.803127364521067e-05,
      "loss": 0.0354,
      "step": 482
    },
    {
      "epoch": 0.3623405851462866,
      "grad_norm": 0.06357493251562119,
      "learning_rate": 9.801303987207229e-05,
      "loss": 0.0236,
      "step": 483
    },
    {
      "epoch": 0.3630907726931733,
      "grad_norm": 0.22393284738063812,
      "learning_rate": 9.799472376129846e-05,
      "loss": 0.0551,
      "step": 484
    },
    {
      "epoch": 0.36384096024006,
      "grad_norm": 0.17032992839813232,
      "learning_rate": 9.79763253442995e-05,
      "loss": 0.0424,
      "step": 485
    },
    {
      "epoch": 0.36459114778694673,
      "grad_norm": 0.2818201780319214,
      "learning_rate": 9.795784465262689e-05,
      "loss": 0.0532,
      "step": 486
    },
    {
      "epoch": 0.36534133533383345,
      "grad_norm": 0.2822514772415161,
      "learning_rate": 9.79392817179732e-05,
      "loss": 0.0402,
      "step": 487
    },
    {
      "epoch": 0.36609152288072017,
      "grad_norm": 0.07287586480379105,
      "learning_rate": 9.792063657217201e-05,
      "loss": 0.0231,
      "step": 488
    },
    {
      "epoch": 0.3668417104276069,
      "grad_norm": 0.10715020447969437,
      "learning_rate": 9.790190924719793e-05,
      "loss": 0.0179,
      "step": 489
    },
    {
      "epoch": 0.3675918979744936,
      "grad_norm": 0.1892417073249817,
      "learning_rate": 9.788309977516648e-05,
      "loss": 0.0476,
      "step": 490
    },
    {
      "epoch": 0.36834208552138037,
      "grad_norm": 0.1349543333053589,
      "learning_rate": 9.786420818833404e-05,
      "loss": 0.0295,
      "step": 491
    },
    {
      "epoch": 0.3690922730682671,
      "grad_norm": 0.13802854716777802,
      "learning_rate": 9.784523451909782e-05,
      "loss": 0.0326,
      "step": 492
    },
    {
      "epoch": 0.3698424606151538,
      "grad_norm": 0.28222379088401794,
      "learning_rate": 9.78261787999958e-05,
      "loss": 0.0546,
      "step": 493
    },
    {
      "epoch": 0.3705926481620405,
      "grad_norm": 0.18687193095684052,
      "learning_rate": 9.780704106370667e-05,
      "loss": 0.0535,
      "step": 494
    },
    {
      "epoch": 0.37134283570892723,
      "grad_norm": 0.11292416602373123,
      "learning_rate": 9.778782134304976e-05,
      "loss": 0.0327,
      "step": 495
    },
    {
      "epoch": 0.37209302325581395,
      "grad_norm": 0.07824314385652542,
      "learning_rate": 9.776851967098499e-05,
      "loss": 0.0218,
      "step": 496
    },
    {
      "epoch": 0.37284321080270066,
      "grad_norm": 0.1915801614522934,
      "learning_rate": 9.774913608061282e-05,
      "loss": 0.0423,
      "step": 497
    },
    {
      "epoch": 0.3735933983495874,
      "grad_norm": 0.18689598143100739,
      "learning_rate": 9.772967060517421e-05,
      "loss": 0.0493,
      "step": 498
    },
    {
      "epoch": 0.3743435858964741,
      "grad_norm": 0.11407961696386337,
      "learning_rate": 9.771012327805055e-05,
      "loss": 0.0302,
      "step": 499
    },
    {
      "epoch": 0.37509377344336087,
      "grad_norm": 0.20327265560626984,
      "learning_rate": 9.769049413276355e-05,
      "loss": 0.0459,
      "step": 500
    },
    {
      "epoch": 0.3758439609902476,
      "grad_norm": 0.1307540237903595,
      "learning_rate": 9.767078320297528e-05,
      "loss": 0.0364,
      "step": 501
    },
    {
      "epoch": 0.3765941485371343,
      "grad_norm": 0.18840287625789642,
      "learning_rate": 9.765099052248805e-05,
      "loss": 0.0419,
      "step": 502
    },
    {
      "epoch": 0.377344336084021,
      "grad_norm": 0.1403331160545349,
      "learning_rate": 9.763111612524434e-05,
      "loss": 0.0356,
      "step": 503
    },
    {
      "epoch": 0.37809452363090773,
      "grad_norm": 0.09807176887989044,
      "learning_rate": 9.761116004532679e-05,
      "loss": 0.0265,
      "step": 504
    },
    {
      "epoch": 0.37884471117779445,
      "grad_norm": 0.16782943904399872,
      "learning_rate": 9.759112231695811e-05,
      "loss": 0.0425,
      "step": 505
    },
    {
      "epoch": 0.37959489872468116,
      "grad_norm": 0.14254771173000336,
      "learning_rate": 9.757100297450103e-05,
      "loss": 0.0527,
      "step": 506
    },
    {
      "epoch": 0.3803450862715679,
      "grad_norm": 0.12506116926670074,
      "learning_rate": 9.755080205245826e-05,
      "loss": 0.0248,
      "step": 507
    },
    {
      "epoch": 0.3810952738184546,
      "grad_norm": 0.12891031801700592,
      "learning_rate": 9.753051958547238e-05,
      "loss": 0.0268,
      "step": 508
    },
    {
      "epoch": 0.3818454613653413,
      "grad_norm": 0.11933682858943939,
      "learning_rate": 9.751015560832582e-05,
      "loss": 0.0384,
      "step": 509
    },
    {
      "epoch": 0.3825956489122281,
      "grad_norm": 0.14737223088741302,
      "learning_rate": 9.748971015594078e-05,
      "loss": 0.0355,
      "step": 510
    },
    {
      "epoch": 0.3833458364591148,
      "grad_norm": 0.1678527295589447,
      "learning_rate": 9.746918326337923e-05,
      "loss": 0.0389,
      "step": 511
    },
    {
      "epoch": 0.3840960240060015,
      "grad_norm": 0.16946767270565033,
      "learning_rate": 9.744857496584274e-05,
      "loss": 0.0455,
      "step": 512
    },
    {
      "epoch": 0.38484621155288823,
      "grad_norm": 0.19964195787906647,
      "learning_rate": 9.742788529867255e-05,
      "loss": 0.0417,
      "step": 513
    },
    {
      "epoch": 0.38559639909977494,
      "grad_norm": 0.10407441854476929,
      "learning_rate": 9.740711429734936e-05,
      "loss": 0.0335,
      "step": 514
    },
    {
      "epoch": 0.38634658664666166,
      "grad_norm": 0.19779853522777557,
      "learning_rate": 9.738626199749341e-05,
      "loss": 0.0372,
      "step": 515
    },
    {
      "epoch": 0.3870967741935484,
      "grad_norm": 0.22248423099517822,
      "learning_rate": 9.736532843486433e-05,
      "loss": 0.049,
      "step": 516
    },
    {
      "epoch": 0.3878469617404351,
      "grad_norm": 0.1619381308555603,
      "learning_rate": 9.734431364536114e-05,
      "loss": 0.0435,
      "step": 517
    },
    {
      "epoch": 0.3885971492873218,
      "grad_norm": 0.16517984867095947,
      "learning_rate": 9.732321766502213e-05,
      "loss": 0.0418,
      "step": 518
    },
    {
      "epoch": 0.3893473368342086,
      "grad_norm": 0.04582435265183449,
      "learning_rate": 9.730204053002481e-05,
      "loss": 0.0173,
      "step": 519
    },
    {
      "epoch": 0.3900975243810953,
      "grad_norm": 0.09578097611665726,
      "learning_rate": 9.728078227668588e-05,
      "loss": 0.0275,
      "step": 520
    },
    {
      "epoch": 0.390847711927982,
      "grad_norm": 0.11759772896766663,
      "learning_rate": 9.725944294146119e-05,
      "loss": 0.0419,
      "step": 521
    },
    {
      "epoch": 0.3915978994748687,
      "grad_norm": 0.11090448498725891,
      "learning_rate": 9.723802256094555e-05,
      "loss": 0.0372,
      "step": 522
    },
    {
      "epoch": 0.39234808702175544,
      "grad_norm": 0.16954998672008514,
      "learning_rate": 9.721652117187283e-05,
      "loss": 0.0348,
      "step": 523
    },
    {
      "epoch": 0.39309827456864216,
      "grad_norm": 0.08991193026304245,
      "learning_rate": 9.71949388111158e-05,
      "loss": 0.0333,
      "step": 524
    },
    {
      "epoch": 0.3938484621155289,
      "grad_norm": 0.144531711935997,
      "learning_rate": 9.717327551568608e-05,
      "loss": 0.0413,
      "step": 525
    },
    {
      "epoch": 0.3945986496624156,
      "grad_norm": 0.08531095832586288,
      "learning_rate": 9.715153132273407e-05,
      "loss": 0.0289,
      "step": 526
    },
    {
      "epoch": 0.3953488372093023,
      "grad_norm": 0.09487048536539078,
      "learning_rate": 9.712970626954893e-05,
      "loss": 0.0267,
      "step": 527
    },
    {
      "epoch": 0.396099024756189,
      "grad_norm": 0.08603309094905853,
      "learning_rate": 9.71078003935585e-05,
      "loss": 0.0286,
      "step": 528
    },
    {
      "epoch": 0.3968492123030758,
      "grad_norm": 0.10319013148546219,
      "learning_rate": 9.708581373232917e-05,
      "loss": 0.0334,
      "step": 529
    },
    {
      "epoch": 0.3975993998499625,
      "grad_norm": 0.15875476598739624,
      "learning_rate": 9.70637463235659e-05,
      "loss": 0.0403,
      "step": 530
    },
    {
      "epoch": 0.3983495873968492,
      "grad_norm": 0.1145065426826477,
      "learning_rate": 9.704159820511214e-05,
      "loss": 0.0393,
      "step": 531
    },
    {
      "epoch": 0.39909977494373594,
      "grad_norm": 0.11985963582992554,
      "learning_rate": 9.701936941494971e-05,
      "loss": 0.0285,
      "step": 532
    },
    {
      "epoch": 0.39984996249062266,
      "grad_norm": 0.18910159170627594,
      "learning_rate": 9.699705999119882e-05,
      "loss": 0.037,
      "step": 533
    },
    {
      "epoch": 0.4006001500375094,
      "grad_norm": 0.11701525002717972,
      "learning_rate": 9.697466997211793e-05,
      "loss": 0.0235,
      "step": 534
    },
    {
      "epoch": 0.4013503375843961,
      "grad_norm": 0.1602649986743927,
      "learning_rate": 9.69521993961037e-05,
      "loss": 0.0384,
      "step": 535
    },
    {
      "epoch": 0.4021005251312828,
      "grad_norm": 0.12196832150220871,
      "learning_rate": 9.692964830169098e-05,
      "loss": 0.0386,
      "step": 536
    },
    {
      "epoch": 0.4028507126781695,
      "grad_norm": 0.11243263632059097,
      "learning_rate": 9.690701672755266e-05,
      "loss": 0.0361,
      "step": 537
    },
    {
      "epoch": 0.4036009002250563,
      "grad_norm": 0.15681670606136322,
      "learning_rate": 9.688430471249967e-05,
      "loss": 0.0411,
      "step": 538
    },
    {
      "epoch": 0.404351087771943,
      "grad_norm": 0.09906280040740967,
      "learning_rate": 9.686151229548088e-05,
      "loss": 0.024,
      "step": 539
    },
    {
      "epoch": 0.4051012753188297,
      "grad_norm": 0.18042565882205963,
      "learning_rate": 9.683863951558301e-05,
      "loss": 0.0415,
      "step": 540
    },
    {
      "epoch": 0.40585146286571644,
      "grad_norm": 0.09993312507867813,
      "learning_rate": 9.681568641203068e-05,
      "loss": 0.0127,
      "step": 541
    },
    {
      "epoch": 0.40660165041260315,
      "grad_norm": 0.1347557008266449,
      "learning_rate": 9.679265302418615e-05,
      "loss": 0.0318,
      "step": 542
    },
    {
      "epoch": 0.40735183795948987,
      "grad_norm": 0.14245480298995972,
      "learning_rate": 9.676953939154945e-05,
      "loss": 0.0406,
      "step": 543
    },
    {
      "epoch": 0.4081020255063766,
      "grad_norm": 0.2006424218416214,
      "learning_rate": 9.674634555375817e-05,
      "loss": 0.0336,
      "step": 544
    },
    {
      "epoch": 0.4088522130532633,
      "grad_norm": 0.16163894534111023,
      "learning_rate": 9.672307155058744e-05,
      "loss": 0.048,
      "step": 545
    },
    {
      "epoch": 0.40960240060015,
      "grad_norm": 0.1653139442205429,
      "learning_rate": 9.669971742194992e-05,
      "loss": 0.0529,
      "step": 546
    },
    {
      "epoch": 0.41035258814703673,
      "grad_norm": 0.17910104990005493,
      "learning_rate": 9.667628320789562e-05,
      "loss": 0.0492,
      "step": 547
    },
    {
      "epoch": 0.4111027756939235,
      "grad_norm": 0.10363556444644928,
      "learning_rate": 9.665276894861188e-05,
      "loss": 0.0375,
      "step": 548
    },
    {
      "epoch": 0.4118529632408102,
      "grad_norm": 0.07300784438848495,
      "learning_rate": 9.66291746844234e-05,
      "loss": 0.0223,
      "step": 549
    },
    {
      "epoch": 0.41260315078769694,
      "grad_norm": 0.1361912339925766,
      "learning_rate": 9.660550045579199e-05,
      "loss": 0.0395,
      "step": 550
    },
    {
      "epoch": 0.41335333833458365,
      "grad_norm": 0.21163193881511688,
      "learning_rate": 9.65817463033166e-05,
      "loss": 0.0494,
      "step": 551
    },
    {
      "epoch": 0.41410352588147037,
      "grad_norm": 0.10939466208219528,
      "learning_rate": 9.655791226773331e-05,
      "loss": 0.0379,
      "step": 552
    },
    {
      "epoch": 0.4148537134283571,
      "grad_norm": 0.14424750208854675,
      "learning_rate": 9.65339983899151e-05,
      "loss": 0.0385,
      "step": 553
    },
    {
      "epoch": 0.4156039009752438,
      "grad_norm": 0.09869871288537979,
      "learning_rate": 9.651000471087193e-05,
      "loss": 0.0306,
      "step": 554
    },
    {
      "epoch": 0.4163540885221305,
      "grad_norm": 0.09752748161554337,
      "learning_rate": 9.64859312717506e-05,
      "loss": 0.0272,
      "step": 555
    },
    {
      "epoch": 0.41710427606901723,
      "grad_norm": 0.19783945381641388,
      "learning_rate": 9.64617781138347e-05,
      "loss": 0.0307,
      "step": 556
    },
    {
      "epoch": 0.41785446361590395,
      "grad_norm": 0.15972763299942017,
      "learning_rate": 9.643754527854451e-05,
      "loss": 0.0443,
      "step": 557
    },
    {
      "epoch": 0.4186046511627907,
      "grad_norm": 0.11784256994724274,
      "learning_rate": 9.641323280743693e-05,
      "loss": 0.0295,
      "step": 558
    },
    {
      "epoch": 0.41935483870967744,
      "grad_norm": 0.09688984602689743,
      "learning_rate": 9.638884074220548e-05,
      "loss": 0.0434,
      "step": 559
    },
    {
      "epoch": 0.42010502625656415,
      "grad_norm": 0.09953712671995163,
      "learning_rate": 9.636436912468015e-05,
      "loss": 0.0259,
      "step": 560
    },
    {
      "epoch": 0.42085521380345087,
      "grad_norm": 0.13692741096019745,
      "learning_rate": 9.633981799682735e-05,
      "loss": 0.0441,
      "step": 561
    },
    {
      "epoch": 0.4216054013503376,
      "grad_norm": 0.09490803629159927,
      "learning_rate": 9.631518740074985e-05,
      "loss": 0.0141,
      "step": 562
    },
    {
      "epoch": 0.4223555888972243,
      "grad_norm": 0.12184827029705048,
      "learning_rate": 9.629047737868669e-05,
      "loss": 0.0328,
      "step": 563
    },
    {
      "epoch": 0.423105776444111,
      "grad_norm": 0.1606912910938263,
      "learning_rate": 9.626568797301311e-05,
      "loss": 0.0454,
      "step": 564
    },
    {
      "epoch": 0.42385596399099773,
      "grad_norm": 0.09314081072807312,
      "learning_rate": 9.624081922624053e-05,
      "loss": 0.0251,
      "step": 565
    },
    {
      "epoch": 0.42460615153788445,
      "grad_norm": 0.1758616864681244,
      "learning_rate": 9.621587118101638e-05,
      "loss": 0.0317,
      "step": 566
    },
    {
      "epoch": 0.4253563390847712,
      "grad_norm": 0.12884719669818878,
      "learning_rate": 9.619084388012412e-05,
      "loss": 0.0356,
      "step": 567
    },
    {
      "epoch": 0.42610652663165793,
      "grad_norm": 0.14522381126880646,
      "learning_rate": 9.616573736648308e-05,
      "loss": 0.0344,
      "step": 568
    },
    {
      "epoch": 0.42685671417854465,
      "grad_norm": 0.11167056858539581,
      "learning_rate": 9.61405516831485e-05,
      "loss": 0.0292,
      "step": 569
    },
    {
      "epoch": 0.42760690172543137,
      "grad_norm": 0.1222648173570633,
      "learning_rate": 9.61152868733113e-05,
      "loss": 0.0246,
      "step": 570
    },
    {
      "epoch": 0.4283570892723181,
      "grad_norm": 0.11057785153388977,
      "learning_rate": 9.608994298029818e-05,
      "loss": 0.0351,
      "step": 571
    },
    {
      "epoch": 0.4291072768192048,
      "grad_norm": 0.1606517732143402,
      "learning_rate": 9.60645200475714e-05,
      "loss": 0.0384,
      "step": 572
    },
    {
      "epoch": 0.4298574643660915,
      "grad_norm": 0.1263444721698761,
      "learning_rate": 9.603901811872877e-05,
      "loss": 0.042,
      "step": 573
    },
    {
      "epoch": 0.43060765191297823,
      "grad_norm": 0.16173186898231506,
      "learning_rate": 9.601343723750363e-05,
      "loss": 0.0365,
      "step": 574
    },
    {
      "epoch": 0.43135783945986494,
      "grad_norm": 0.07905527204275131,
      "learning_rate": 9.598777744776464e-05,
      "loss": 0.0223,
      "step": 575
    },
    {
      "epoch": 0.43210802700675166,
      "grad_norm": 0.13041624426841736,
      "learning_rate": 9.596203879351582e-05,
      "loss": 0.0288,
      "step": 576
    },
    {
      "epoch": 0.43285821455363843,
      "grad_norm": 0.14066483080387115,
      "learning_rate": 9.593622131889643e-05,
      "loss": 0.0418,
      "step": 577
    },
    {
      "epoch": 0.43360840210052515,
      "grad_norm": 0.10811105370521545,
      "learning_rate": 9.591032506818089e-05,
      "loss": 0.0283,
      "step": 578
    },
    {
      "epoch": 0.43435858964741186,
      "grad_norm": 0.15928411483764648,
      "learning_rate": 9.588435008577873e-05,
      "loss": 0.0369,
      "step": 579
    },
    {
      "epoch": 0.4351087771942986,
      "grad_norm": 0.07709395885467529,
      "learning_rate": 9.585829641623448e-05,
      "loss": 0.02,
      "step": 580
    },
    {
      "epoch": 0.4358589647411853,
      "grad_norm": 0.13441017270088196,
      "learning_rate": 9.583216410422762e-05,
      "loss": 0.0369,
      "step": 581
    },
    {
      "epoch": 0.436609152288072,
      "grad_norm": 0.1272798478603363,
      "learning_rate": 9.580595319457249e-05,
      "loss": 0.0337,
      "step": 582
    },
    {
      "epoch": 0.4373593398349587,
      "grad_norm": 0.2848840653896332,
      "learning_rate": 9.577966373221823e-05,
      "loss": 0.0627,
      "step": 583
    },
    {
      "epoch": 0.43810952738184544,
      "grad_norm": 0.12250744551420212,
      "learning_rate": 9.575329576224868e-05,
      "loss": 0.0348,
      "step": 584
    },
    {
      "epoch": 0.43885971492873216,
      "grad_norm": 0.08176639676094055,
      "learning_rate": 9.572684932988227e-05,
      "loss": 0.0236,
      "step": 585
    },
    {
      "epoch": 0.43960990247561893,
      "grad_norm": 0.11825791746377945,
      "learning_rate": 9.570032448047208e-05,
      "loss": 0.0392,
      "step": 586
    },
    {
      "epoch": 0.44036009002250565,
      "grad_norm": 0.13997545838356018,
      "learning_rate": 9.567372125950559e-05,
      "loss": 0.0255,
      "step": 587
    },
    {
      "epoch": 0.44111027756939236,
      "grad_norm": 0.07865186035633087,
      "learning_rate": 9.564703971260472e-05,
      "loss": 0.0253,
      "step": 588
    },
    {
      "epoch": 0.4418604651162791,
      "grad_norm": 0.08131557703018188,
      "learning_rate": 9.562027988552567e-05,
      "loss": 0.0295,
      "step": 589
    },
    {
      "epoch": 0.4426106526631658,
      "grad_norm": 0.15946389734745026,
      "learning_rate": 9.559344182415891e-05,
      "loss": 0.0626,
      "step": 590
    },
    {
      "epoch": 0.4433608402100525,
      "grad_norm": 0.13182489573955536,
      "learning_rate": 9.55665255745291e-05,
      "loss": 0.0403,
      "step": 591
    },
    {
      "epoch": 0.4441110277569392,
      "grad_norm": 0.14255809783935547,
      "learning_rate": 9.553953118279496e-05,
      "loss": 0.0445,
      "step": 592
    },
    {
      "epoch": 0.44486121530382594,
      "grad_norm": 0.1584695428609848,
      "learning_rate": 9.551245869524916e-05,
      "loss": 0.0415,
      "step": 593
    },
    {
      "epoch": 0.44561140285071266,
      "grad_norm": 0.08315354585647583,
      "learning_rate": 9.54853081583184e-05,
      "loss": 0.0288,
      "step": 594
    },
    {
      "epoch": 0.4463615903975994,
      "grad_norm": 0.09172634780406952,
      "learning_rate": 9.545807961856317e-05,
      "loss": 0.0261,
      "step": 595
    },
    {
      "epoch": 0.44711177794448614,
      "grad_norm": 0.13752834498882294,
      "learning_rate": 9.543077312267773e-05,
      "loss": 0.0385,
      "step": 596
    },
    {
      "epoch": 0.44786196549137286,
      "grad_norm": 0.09433577209711075,
      "learning_rate": 9.540338871749002e-05,
      "loss": 0.0369,
      "step": 597
    },
    {
      "epoch": 0.4486121530382596,
      "grad_norm": 0.1977624148130417,
      "learning_rate": 9.537592644996162e-05,
      "loss": 0.0442,
      "step": 598
    },
    {
      "epoch": 0.4493623405851463,
      "grad_norm": 0.0797150656580925,
      "learning_rate": 9.534838636718759e-05,
      "loss": 0.0285,
      "step": 599
    },
    {
      "epoch": 0.450112528132033,
      "grad_norm": 0.1593594253063202,
      "learning_rate": 9.532076851639649e-05,
      "loss": 0.0451,
      "step": 600
    },
    {
      "epoch": 0.450112528132033,
      "eval_loss": 0.039720457047224045,
      "eval_runtime": 5.1368,
      "eval_samples_per_second": 10.512,
      "eval_steps_per_second": 2.725,
      "step": 600
    },
    {
      "epoch": 0.4508627156789197,
      "grad_norm": 0.05833355337381363,
      "learning_rate": 9.529307294495018e-05,
      "loss": 0.0185,
      "step": 601
    },
    {
      "epoch": 0.45161290322580644,
      "grad_norm": 0.1370234489440918,
      "learning_rate": 9.526529970034386e-05,
      "loss": 0.0461,
      "step": 602
    },
    {
      "epoch": 0.45236309077269315,
      "grad_norm": 0.11114072799682617,
      "learning_rate": 9.52374488302059e-05,
      "loss": 0.0355,
      "step": 603
    },
    {
      "epoch": 0.45311327831957987,
      "grad_norm": 0.11904539167881012,
      "learning_rate": 9.52095203822978e-05,
      "loss": 0.0278,
      "step": 604
    },
    {
      "epoch": 0.45386346586646664,
      "grad_norm": 0.14749003946781158,
      "learning_rate": 9.518151440451411e-05,
      "loss": 0.0367,
      "step": 605
    },
    {
      "epoch": 0.45461365341335336,
      "grad_norm": 0.12626592814922333,
      "learning_rate": 9.515343094488232e-05,
      "loss": 0.0326,
      "step": 606
    },
    {
      "epoch": 0.4553638409602401,
      "grad_norm": 0.1684754639863968,
      "learning_rate": 9.51252700515628e-05,
      "loss": 0.0349,
      "step": 607
    },
    {
      "epoch": 0.4561140285071268,
      "grad_norm": 0.1260501742362976,
      "learning_rate": 9.509703177284869e-05,
      "loss": 0.0336,
      "step": 608
    },
    {
      "epoch": 0.4568642160540135,
      "grad_norm": 0.21763825416564941,
      "learning_rate": 9.506871615716587e-05,
      "loss": 0.0622,
      "step": 609
    },
    {
      "epoch": 0.4576144036009002,
      "grad_norm": 0.12283841520547867,
      "learning_rate": 9.504032325307284e-05,
      "loss": 0.0307,
      "step": 610
    },
    {
      "epoch": 0.45836459114778694,
      "grad_norm": 0.17834042012691498,
      "learning_rate": 9.501185310926062e-05,
      "loss": 0.0459,
      "step": 611
    },
    {
      "epoch": 0.45911477869467365,
      "grad_norm": 0.20170065760612488,
      "learning_rate": 9.498330577455273e-05,
      "loss": 0.0412,
      "step": 612
    },
    {
      "epoch": 0.45986496624156037,
      "grad_norm": 0.1369791477918625,
      "learning_rate": 9.495468129790499e-05,
      "loss": 0.0353,
      "step": 613
    },
    {
      "epoch": 0.4606151537884471,
      "grad_norm": 0.10395601391792297,
      "learning_rate": 9.49259797284056e-05,
      "loss": 0.0361,
      "step": 614
    },
    {
      "epoch": 0.46136534133533386,
      "grad_norm": 0.12245897203683853,
      "learning_rate": 9.489720111527492e-05,
      "loss": 0.0541,
      "step": 615
    },
    {
      "epoch": 0.46211552888222057,
      "grad_norm": 0.11299577355384827,
      "learning_rate": 9.486834550786543e-05,
      "loss": 0.0309,
      "step": 616
    },
    {
      "epoch": 0.4628657164291073,
      "grad_norm": 0.08725673705339432,
      "learning_rate": 9.483941295566165e-05,
      "loss": 0.0261,
      "step": 617
    },
    {
      "epoch": 0.463615903975994,
      "grad_norm": 0.12845538556575775,
      "learning_rate": 9.481040350828006e-05,
      "loss": 0.0218,
      "step": 618
    },
    {
      "epoch": 0.4643660915228807,
      "grad_norm": 0.09690548479557037,
      "learning_rate": 9.4781317215469e-05,
      "loss": 0.0274,
      "step": 619
    },
    {
      "epoch": 0.46511627906976744,
      "grad_norm": 0.09320426732301712,
      "learning_rate": 9.475215412710864e-05,
      "loss": 0.0293,
      "step": 620
    },
    {
      "epoch": 0.46586646661665415,
      "grad_norm": 0.08421576768159866,
      "learning_rate": 9.472291429321075e-05,
      "loss": 0.0265,
      "step": 621
    },
    {
      "epoch": 0.46661665416354087,
      "grad_norm": 0.18840914964675903,
      "learning_rate": 9.469359776391879e-05,
      "loss": 0.0451,
      "step": 622
    },
    {
      "epoch": 0.4673668417104276,
      "grad_norm": 0.22850267589092255,
      "learning_rate": 9.466420458950773e-05,
      "loss": 0.0506,
      "step": 623
    },
    {
      "epoch": 0.46811702925731435,
      "grad_norm": 0.13049699366092682,
      "learning_rate": 9.463473482038395e-05,
      "loss": 0.035,
      "step": 624
    },
    {
      "epoch": 0.46886721680420107,
      "grad_norm": 0.1595403105020523,
      "learning_rate": 9.46051885070852e-05,
      "loss": 0.0391,
      "step": 625
    },
    {
      "epoch": 0.4696174043510878,
      "grad_norm": 0.149023175239563,
      "learning_rate": 9.457556570028052e-05,
      "loss": 0.0383,
      "step": 626
    },
    {
      "epoch": 0.4703675918979745,
      "grad_norm": 0.12804150581359863,
      "learning_rate": 9.454586645077011e-05,
      "loss": 0.0288,
      "step": 627
    },
    {
      "epoch": 0.4711177794448612,
      "grad_norm": 0.16897211968898773,
      "learning_rate": 9.451609080948522e-05,
      "loss": 0.055,
      "step": 628
    },
    {
      "epoch": 0.47186796699174793,
      "grad_norm": 0.10514383763074875,
      "learning_rate": 9.448623882748817e-05,
      "loss": 0.033,
      "step": 629
    },
    {
      "epoch": 0.47261815453863465,
      "grad_norm": 0.11677896976470947,
      "learning_rate": 9.445631055597217e-05,
      "loss": 0.0272,
      "step": 630
    },
    {
      "epoch": 0.47336834208552137,
      "grad_norm": 0.15391424298286438,
      "learning_rate": 9.442630604626126e-05,
      "loss": 0.0485,
      "step": 631
    },
    {
      "epoch": 0.4741185296324081,
      "grad_norm": 0.0809544175863266,
      "learning_rate": 9.43962253498102e-05,
      "loss": 0.0232,
      "step": 632
    },
    {
      "epoch": 0.4748687171792948,
      "grad_norm": 0.09526669234037399,
      "learning_rate": 9.436606851820444e-05,
      "loss": 0.0285,
      "step": 633
    },
    {
      "epoch": 0.47561890472618157,
      "grad_norm": 0.12024281173944473,
      "learning_rate": 9.433583560315999e-05,
      "loss": 0.0358,
      "step": 634
    },
    {
      "epoch": 0.4763690922730683,
      "grad_norm": 0.12453112006187439,
      "learning_rate": 9.430552665652328e-05,
      "loss": 0.0365,
      "step": 635
    },
    {
      "epoch": 0.477119279819955,
      "grad_norm": 0.10970939695835114,
      "learning_rate": 9.427514173027121e-05,
      "loss": 0.0293,
      "step": 636
    },
    {
      "epoch": 0.4778694673668417,
      "grad_norm": 0.10942260921001434,
      "learning_rate": 9.424468087651092e-05,
      "loss": 0.03,
      "step": 637
    },
    {
      "epoch": 0.47861965491372843,
      "grad_norm": 0.06871689856052399,
      "learning_rate": 9.421414414747978e-05,
      "loss": 0.019,
      "step": 638
    },
    {
      "epoch": 0.47936984246061515,
      "grad_norm": 0.08539148420095444,
      "learning_rate": 9.418353159554526e-05,
      "loss": 0.031,
      "step": 639
    },
    {
      "epoch": 0.48012003000750186,
      "grad_norm": 0.1259005069732666,
      "learning_rate": 9.415284327320489e-05,
      "loss": 0.0259,
      "step": 640
    },
    {
      "epoch": 0.4808702175543886,
      "grad_norm": 0.12335602939128876,
      "learning_rate": 9.41220792330861e-05,
      "loss": 0.0417,
      "step": 641
    },
    {
      "epoch": 0.4816204051012753,
      "grad_norm": 0.12800176441669464,
      "learning_rate": 9.40912395279462e-05,
      "loss": 0.0388,
      "step": 642
    },
    {
      "epoch": 0.48237059264816207,
      "grad_norm": 0.13285093009471893,
      "learning_rate": 9.406032421067224e-05,
      "loss": 0.0347,
      "step": 643
    },
    {
      "epoch": 0.4831207801950488,
      "grad_norm": 0.13197124004364014,
      "learning_rate": 9.402933333428097e-05,
      "loss": 0.0358,
      "step": 644
    },
    {
      "epoch": 0.4838709677419355,
      "grad_norm": 0.1399867832660675,
      "learning_rate": 9.399826695191868e-05,
      "loss": 0.0265,
      "step": 645
    },
    {
      "epoch": 0.4846211552888222,
      "grad_norm": 0.12037494033575058,
      "learning_rate": 9.396712511686114e-05,
      "loss": 0.0298,
      "step": 646
    },
    {
      "epoch": 0.48537134283570893,
      "grad_norm": 0.11193617433309555,
      "learning_rate": 9.393590788251354e-05,
      "loss": 0.0287,
      "step": 647
    },
    {
      "epoch": 0.48612153038259565,
      "grad_norm": 0.09210868179798126,
      "learning_rate": 9.390461530241037e-05,
      "loss": 0.0206,
      "step": 648
    },
    {
      "epoch": 0.48687171792948236,
      "grad_norm": 0.2255852371454239,
      "learning_rate": 9.38732474302153e-05,
      "loss": 0.0494,
      "step": 649
    },
    {
      "epoch": 0.4876219054763691,
      "grad_norm": 0.13794954121112823,
      "learning_rate": 9.384180431972119e-05,
      "loss": 0.0282,
      "step": 650
    },
    {
      "epoch": 0.4883720930232558,
      "grad_norm": 0.11250860244035721,
      "learning_rate": 9.381028602484984e-05,
      "loss": 0.0306,
      "step": 651
    },
    {
      "epoch": 0.4891222805701425,
      "grad_norm": 0.15982091426849365,
      "learning_rate": 9.377869259965202e-05,
      "loss": 0.0386,
      "step": 652
    },
    {
      "epoch": 0.4898724681170293,
      "grad_norm": 0.1598033607006073,
      "learning_rate": 9.374702409830736e-05,
      "loss": 0.0452,
      "step": 653
    },
    {
      "epoch": 0.490622655663916,
      "grad_norm": 0.20759710669517517,
      "learning_rate": 9.37152805751242e-05,
      "loss": 0.0378,
      "step": 654
    },
    {
      "epoch": 0.4913728432108027,
      "grad_norm": 0.130874902009964,
      "learning_rate": 9.36834620845396e-05,
      "loss": 0.0317,
      "step": 655
    },
    {
      "epoch": 0.49212303075768943,
      "grad_norm": 0.12012483179569244,
      "learning_rate": 9.365156868111908e-05,
      "loss": 0.0302,
      "step": 656
    },
    {
      "epoch": 0.49287321830457614,
      "grad_norm": 0.14452581107616425,
      "learning_rate": 9.361960041955672e-05,
      "loss": 0.0292,
      "step": 657
    },
    {
      "epoch": 0.49362340585146286,
      "grad_norm": 0.11331783980131149,
      "learning_rate": 9.358755735467494e-05,
      "loss": 0.0416,
      "step": 658
    },
    {
      "epoch": 0.4943735933983496,
      "grad_norm": 0.10472054779529572,
      "learning_rate": 9.355543954142446e-05,
      "loss": 0.0372,
      "step": 659
    },
    {
      "epoch": 0.4951237809452363,
      "grad_norm": 0.18476296961307526,
      "learning_rate": 9.352324703488412e-05,
      "loss": 0.0546,
      "step": 660
    },
    {
      "epoch": 0.495873968492123,
      "grad_norm": 0.15593048930168152,
      "learning_rate": 9.349097989026093e-05,
      "loss": 0.0359,
      "step": 661
    },
    {
      "epoch": 0.4966241560390098,
      "grad_norm": 0.1576404720544815,
      "learning_rate": 9.345863816288985e-05,
      "loss": 0.0433,
      "step": 662
    },
    {
      "epoch": 0.4973743435858965,
      "grad_norm": 0.10880573093891144,
      "learning_rate": 9.342622190823378e-05,
      "loss": 0.0277,
      "step": 663
    },
    {
      "epoch": 0.4981245311327832,
      "grad_norm": 0.1214156374335289,
      "learning_rate": 9.339373118188338e-05,
      "loss": 0.0298,
      "step": 664
    },
    {
      "epoch": 0.4988747186796699,
      "grad_norm": 0.07884982973337173,
      "learning_rate": 9.336116603955707e-05,
      "loss": 0.0221,
      "step": 665
    },
    {
      "epoch": 0.49962490622655664,
      "grad_norm": 0.09125591069459915,
      "learning_rate": 9.332852653710084e-05,
      "loss": 0.0289,
      "step": 666
    },
    {
      "epoch": 0.5003750937734434,
      "grad_norm": 0.18234345316886902,
      "learning_rate": 9.329581273048822e-05,
      "loss": 0.0373,
      "step": 667
    },
    {
      "epoch": 0.5011252813203301,
      "grad_norm": 0.13476620614528656,
      "learning_rate": 9.32630246758202e-05,
      "loss": 0.0298,
      "step": 668
    },
    {
      "epoch": 0.5018754688672168,
      "grad_norm": 0.11463253945112228,
      "learning_rate": 9.323016242932504e-05,
      "loss": 0.0448,
      "step": 669
    },
    {
      "epoch": 0.5026256564141035,
      "grad_norm": 0.09115209430456161,
      "learning_rate": 9.319722604735825e-05,
      "loss": 0.0261,
      "step": 670
    },
    {
      "epoch": 0.5033758439609902,
      "grad_norm": 0.1378633677959442,
      "learning_rate": 9.31642155864025e-05,
      "loss": 0.029,
      "step": 671
    },
    {
      "epoch": 0.5041260315078769,
      "grad_norm": 0.11013531684875488,
      "learning_rate": 9.313113110306748e-05,
      "loss": 0.025,
      "step": 672
    },
    {
      "epoch": 0.5048762190547637,
      "grad_norm": 0.1313575953245163,
      "learning_rate": 9.309797265408979e-05,
      "loss": 0.0355,
      "step": 673
    },
    {
      "epoch": 0.5056264066016504,
      "grad_norm": 0.0926070511341095,
      "learning_rate": 9.306474029633294e-05,
      "loss": 0.0244,
      "step": 674
    },
    {
      "epoch": 0.5063765941485371,
      "grad_norm": 0.11532235890626907,
      "learning_rate": 9.303143408678716e-05,
      "loss": 0.0347,
      "step": 675
    },
    {
      "epoch": 0.5071267816954238,
      "grad_norm": 0.11281560361385345,
      "learning_rate": 9.299805408256928e-05,
      "loss": 0.0226,
      "step": 676
    },
    {
      "epoch": 0.5078769692423106,
      "grad_norm": 0.17721490561962128,
      "learning_rate": 9.296460034092274e-05,
      "loss": 0.0347,
      "step": 677
    },
    {
      "epoch": 0.5086271567891973,
      "grad_norm": 0.09870696067810059,
      "learning_rate": 9.293107291921741e-05,
      "loss": 0.0255,
      "step": 678
    },
    {
      "epoch": 0.5093773443360841,
      "grad_norm": 0.08905252069234848,
      "learning_rate": 9.289747187494952e-05,
      "loss": 0.0202,
      "step": 679
    },
    {
      "epoch": 0.5101275318829708,
      "grad_norm": 0.18132604658603668,
      "learning_rate": 9.286379726574155e-05,
      "loss": 0.0331,
      "step": 680
    },
    {
      "epoch": 0.5108777194298575,
      "grad_norm": 0.10676857084035873,
      "learning_rate": 9.283004914934215e-05,
      "loss": 0.0293,
      "step": 681
    },
    {
      "epoch": 0.5116279069767442,
      "grad_norm": 0.13029693067073822,
      "learning_rate": 9.2796227583626e-05,
      "loss": 0.037,
      "step": 682
    },
    {
      "epoch": 0.5123780945236309,
      "grad_norm": 0.17007900774478912,
      "learning_rate": 9.276233262659375e-05,
      "loss": 0.0348,
      "step": 683
    },
    {
      "epoch": 0.5131282820705176,
      "grad_norm": 0.125155508518219,
      "learning_rate": 9.272836433637193e-05,
      "loss": 0.0345,
      "step": 684
    },
    {
      "epoch": 0.5138784696174044,
      "grad_norm": 0.1830897033214569,
      "learning_rate": 9.269432277121281e-05,
      "loss": 0.0329,
      "step": 685
    },
    {
      "epoch": 0.5146286571642911,
      "grad_norm": 0.15975606441497803,
      "learning_rate": 9.266020798949433e-05,
      "loss": 0.0311,
      "step": 686
    },
    {
      "epoch": 0.5153788447111778,
      "grad_norm": 0.12618814408779144,
      "learning_rate": 9.262602004971996e-05,
      "loss": 0.032,
      "step": 687
    },
    {
      "epoch": 0.5161290322580645,
      "grad_norm": 0.0827803909778595,
      "learning_rate": 9.259175901051867e-05,
      "loss": 0.0228,
      "step": 688
    },
    {
      "epoch": 0.5168792198049512,
      "grad_norm": 0.194037526845932,
      "learning_rate": 9.255742493064474e-05,
      "loss": 0.0437,
      "step": 689
    },
    {
      "epoch": 0.5176294073518379,
      "grad_norm": 0.19778898358345032,
      "learning_rate": 9.252301786897776e-05,
      "loss": 0.0482,
      "step": 690
    },
    {
      "epoch": 0.5183795948987246,
      "grad_norm": 0.190335214138031,
      "learning_rate": 9.248853788452247e-05,
      "loss": 0.048,
      "step": 691
    },
    {
      "epoch": 0.5191297824456114,
      "grad_norm": 0.2096565216779709,
      "learning_rate": 9.24539850364086e-05,
      "loss": 0.0326,
      "step": 692
    },
    {
      "epoch": 0.5198799699924981,
      "grad_norm": 0.19046910107135773,
      "learning_rate": 9.241935938389093e-05,
      "loss": 0.0405,
      "step": 693
    },
    {
      "epoch": 0.5206301575393848,
      "grad_norm": 0.12360625714063644,
      "learning_rate": 9.238466098634902e-05,
      "loss": 0.033,
      "step": 694
    },
    {
      "epoch": 0.5213803450862715,
      "grad_norm": 0.16132117807865143,
      "learning_rate": 9.234988990328719e-05,
      "loss": 0.0487,
      "step": 695
    },
    {
      "epoch": 0.5221305326331583,
      "grad_norm": 0.10174372047185898,
      "learning_rate": 9.231504619433445e-05,
      "loss": 0.0304,
      "step": 696
    },
    {
      "epoch": 0.522880720180045,
      "grad_norm": 0.13074737787246704,
      "learning_rate": 9.228012991924433e-05,
      "loss": 0.0397,
      "step": 697
    },
    {
      "epoch": 0.5236309077269318,
      "grad_norm": 0.10839260369539261,
      "learning_rate": 9.224514113789477e-05,
      "loss": 0.0317,
      "step": 698
    },
    {
      "epoch": 0.5243810952738185,
      "grad_norm": 0.14983467757701874,
      "learning_rate": 9.221007991028814e-05,
      "loss": 0.0411,
      "step": 699
    },
    {
      "epoch": 0.5251312828207052,
      "grad_norm": 0.08731064200401306,
      "learning_rate": 9.217494629655094e-05,
      "loss": 0.0227,
      "step": 700
    },
    {
      "epoch": 0.5258814703675919,
      "grad_norm": 0.1175181195139885,
      "learning_rate": 9.213974035693389e-05,
      "loss": 0.0285,
      "step": 701
    },
    {
      "epoch": 0.5266316579144786,
      "grad_norm": 0.12022960186004639,
      "learning_rate": 9.21044621518117e-05,
      "loss": 0.0292,
      "step": 702
    },
    {
      "epoch": 0.5273818454613654,
      "grad_norm": 0.12367639690637589,
      "learning_rate": 9.206911174168301e-05,
      "loss": 0.039,
      "step": 703
    },
    {
      "epoch": 0.5281320330082521,
      "grad_norm": 0.21845479309558868,
      "learning_rate": 9.20336891871703e-05,
      "loss": 0.0486,
      "step": 704
    },
    {
      "epoch": 0.5288822205551388,
      "grad_norm": 0.11741850525140762,
      "learning_rate": 9.199819454901977e-05,
      "loss": 0.0344,
      "step": 705
    },
    {
      "epoch": 0.5296324081020255,
      "grad_norm": 0.24551405012607574,
      "learning_rate": 9.196262788810121e-05,
      "loss": 0.0664,
      "step": 706
    },
    {
      "epoch": 0.5303825956489122,
      "grad_norm": 0.15015265345573425,
      "learning_rate": 9.192698926540795e-05,
      "loss": 0.0361,
      "step": 707
    },
    {
      "epoch": 0.5311327831957989,
      "grad_norm": 0.16593538224697113,
      "learning_rate": 9.189127874205674e-05,
      "loss": 0.0423,
      "step": 708
    },
    {
      "epoch": 0.5318829707426856,
      "grad_norm": 0.11336637288331985,
      "learning_rate": 9.185549637928758e-05,
      "loss": 0.0302,
      "step": 709
    },
    {
      "epoch": 0.5326331582895724,
      "grad_norm": 0.17877517640590668,
      "learning_rate": 9.181964223846371e-05,
      "loss": 0.0448,
      "step": 710
    },
    {
      "epoch": 0.5333833458364591,
      "grad_norm": 0.12126030772924423,
      "learning_rate": 9.178371638107146e-05,
      "loss": 0.0384,
      "step": 711
    },
    {
      "epoch": 0.5341335333833458,
      "grad_norm": 0.12825068831443787,
      "learning_rate": 9.174771886872011e-05,
      "loss": 0.0283,
      "step": 712
    },
    {
      "epoch": 0.5348837209302325,
      "grad_norm": 0.19938157498836517,
      "learning_rate": 9.17116497631419e-05,
      "loss": 0.0415,
      "step": 713
    },
    {
      "epoch": 0.5356339084771192,
      "grad_norm": 0.12208958715200424,
      "learning_rate": 9.167550912619173e-05,
      "loss": 0.034,
      "step": 714
    },
    {
      "epoch": 0.536384096024006,
      "grad_norm": 0.11766844242811203,
      "learning_rate": 9.16392970198473e-05,
      "loss": 0.0261,
      "step": 715
    },
    {
      "epoch": 0.5371342835708928,
      "grad_norm": 0.09077750146389008,
      "learning_rate": 9.160301350620875e-05,
      "loss": 0.0263,
      "step": 716
    },
    {
      "epoch": 0.5378844711177795,
      "grad_norm": 0.08629114925861359,
      "learning_rate": 9.156665864749876e-05,
      "loss": 0.0233,
      "step": 717
    },
    {
      "epoch": 0.5386346586646662,
      "grad_norm": 0.09420029819011688,
      "learning_rate": 9.153023250606234e-05,
      "loss": 0.0274,
      "step": 718
    },
    {
      "epoch": 0.5393848462115529,
      "grad_norm": 0.14515256881713867,
      "learning_rate": 9.14937351443667e-05,
      "loss": 0.0402,
      "step": 719
    },
    {
      "epoch": 0.5401350337584396,
      "grad_norm": 0.14991170167922974,
      "learning_rate": 9.145716662500126e-05,
      "loss": 0.0379,
      "step": 720
    },
    {
      "epoch": 0.5408852213053263,
      "grad_norm": 0.13430874049663544,
      "learning_rate": 9.142052701067741e-05,
      "loss": 0.0499,
      "step": 721
    },
    {
      "epoch": 0.5416354088522131,
      "grad_norm": 0.16690127551555634,
      "learning_rate": 9.13838163642285e-05,
      "loss": 0.0396,
      "step": 722
    },
    {
      "epoch": 0.5423855963990998,
      "grad_norm": 0.1311984360218048,
      "learning_rate": 9.134703474860963e-05,
      "loss": 0.0251,
      "step": 723
    },
    {
      "epoch": 0.5431357839459865,
      "grad_norm": 0.11423755437135696,
      "learning_rate": 9.13101822268977e-05,
      "loss": 0.0339,
      "step": 724
    },
    {
      "epoch": 0.5438859714928732,
      "grad_norm": 0.24795664846897125,
      "learning_rate": 9.127325886229115e-05,
      "loss": 0.0458,
      "step": 725
    },
    {
      "epoch": 0.5446361590397599,
      "grad_norm": 0.09746094048023224,
      "learning_rate": 9.123626471810988e-05,
      "loss": 0.0247,
      "step": 726
    },
    {
      "epoch": 0.5453863465866466,
      "grad_norm": 0.19780874252319336,
      "learning_rate": 9.119919985779521e-05,
      "loss": 0.0467,
      "step": 727
    },
    {
      "epoch": 0.5461365341335334,
      "grad_norm": 0.1802770495414734,
      "learning_rate": 9.116206434490976e-05,
      "loss": 0.0508,
      "step": 728
    },
    {
      "epoch": 0.5468867216804201,
      "grad_norm": 0.16529236733913422,
      "learning_rate": 9.112485824313726e-05,
      "loss": 0.0376,
      "step": 729
    },
    {
      "epoch": 0.5476369092273068,
      "grad_norm": 0.11283475160598755,
      "learning_rate": 9.10875816162825e-05,
      "loss": 0.0307,
      "step": 730
    },
    {
      "epoch": 0.5483870967741935,
      "grad_norm": 0.09951961785554886,
      "learning_rate": 9.105023452827121e-05,
      "loss": 0.0285,
      "step": 731
    },
    {
      "epoch": 0.5491372843210802,
      "grad_norm": 0.09138637781143188,
      "learning_rate": 9.101281704315002e-05,
      "loss": 0.0274,
      "step": 732
    },
    {
      "epoch": 0.5498874718679669,
      "grad_norm": 0.12191305309534073,
      "learning_rate": 9.097532922508619e-05,
      "loss": 0.0398,
      "step": 733
    },
    {
      "epoch": 0.5506376594148538,
      "grad_norm": 0.12989825010299683,
      "learning_rate": 9.093777113836765e-05,
      "loss": 0.0294,
      "step": 734
    },
    {
      "epoch": 0.5513878469617405,
      "grad_norm": 0.13126561045646667,
      "learning_rate": 9.090014284740283e-05,
      "loss": 0.0353,
      "step": 735
    },
    {
      "epoch": 0.5521380345086272,
      "grad_norm": 0.11954871565103531,
      "learning_rate": 9.086244441672052e-05,
      "loss": 0.0298,
      "step": 736
    },
    {
      "epoch": 0.5528882220555139,
      "grad_norm": 0.20153184235095978,
      "learning_rate": 9.082467591096982e-05,
      "loss": 0.0482,
      "step": 737
    },
    {
      "epoch": 0.5536384096024006,
      "grad_norm": 0.1392621099948883,
      "learning_rate": 9.078683739492002e-05,
      "loss": 0.0382,
      "step": 738
    },
    {
      "epoch": 0.5543885971492873,
      "grad_norm": 0.16660861670970917,
      "learning_rate": 9.074892893346043e-05,
      "loss": 0.0419,
      "step": 739
    },
    {
      "epoch": 0.5551387846961741,
      "grad_norm": 0.0964227095246315,
      "learning_rate": 9.071095059160035e-05,
      "loss": 0.0267,
      "step": 740
    },
    {
      "epoch": 0.5558889722430608,
      "grad_norm": 0.0833270475268364,
      "learning_rate": 9.067290243446887e-05,
      "loss": 0.0233,
      "step": 741
    },
    {
      "epoch": 0.5566391597899475,
      "grad_norm": 0.09658671915531158,
      "learning_rate": 9.063478452731484e-05,
      "loss": 0.0246,
      "step": 742
    },
    {
      "epoch": 0.5573893473368342,
      "grad_norm": 0.12930825352668762,
      "learning_rate": 9.059659693550673e-05,
      "loss": 0.0336,
      "step": 743
    },
    {
      "epoch": 0.5581395348837209,
      "grad_norm": 0.09614095836877823,
      "learning_rate": 9.055833972453249e-05,
      "loss": 0.0335,
      "step": 744
    },
    {
      "epoch": 0.5588897224306076,
      "grad_norm": 0.22925931215286255,
      "learning_rate": 9.052001295999947e-05,
      "loss": 0.0664,
      "step": 745
    },
    {
      "epoch": 0.5596399099774944,
      "grad_norm": 0.13005341589450836,
      "learning_rate": 9.048161670763429e-05,
      "loss": 0.0297,
      "step": 746
    },
    {
      "epoch": 0.5603900975243811,
      "grad_norm": 0.10627295076847076,
      "learning_rate": 9.044315103328276e-05,
      "loss": 0.0259,
      "step": 747
    },
    {
      "epoch": 0.5611402850712678,
      "grad_norm": 0.1018269807100296,
      "learning_rate": 9.04046160029097e-05,
      "loss": 0.0275,
      "step": 748
    },
    {
      "epoch": 0.5618904726181545,
      "grad_norm": 0.13973957300186157,
      "learning_rate": 9.036601168259893e-05,
      "loss": 0.0283,
      "step": 749
    },
    {
      "epoch": 0.5626406601650412,
      "grad_norm": 0.17691271007061005,
      "learning_rate": 9.032733813855301e-05,
      "loss": 0.0356,
      "step": 750
    },
    {
      "epoch": 0.5633908477119279,
      "grad_norm": 0.10310542583465576,
      "learning_rate": 9.02885954370933e-05,
      "loss": 0.0272,
      "step": 751
    },
    {
      "epoch": 0.5641410352588146,
      "grad_norm": 0.1504839062690735,
      "learning_rate": 9.02497836446597e-05,
      "loss": 0.0411,
      "step": 752
    },
    {
      "epoch": 0.5648912228057015,
      "grad_norm": 0.13666382431983948,
      "learning_rate": 9.021090282781059e-05,
      "loss": 0.0311,
      "step": 753
    },
    {
      "epoch": 0.5656414103525882,
      "grad_norm": 0.13605177402496338,
      "learning_rate": 9.01719530532228e-05,
      "loss": 0.0389,
      "step": 754
    },
    {
      "epoch": 0.5663915978994749,
      "grad_norm": 0.07885899394750595,
      "learning_rate": 9.01329343876913e-05,
      "loss": 0.021,
      "step": 755
    },
    {
      "epoch": 0.5671417854463616,
      "grad_norm": 0.13754671812057495,
      "learning_rate": 9.009384689812928e-05,
      "loss": 0.0401,
      "step": 756
    },
    {
      "epoch": 0.5678919729932483,
      "grad_norm": 0.09757539629936218,
      "learning_rate": 9.005469065156795e-05,
      "loss": 0.0228,
      "step": 757
    },
    {
      "epoch": 0.568642160540135,
      "grad_norm": 0.17368543148040771,
      "learning_rate": 9.00154657151564e-05,
      "loss": 0.04,
      "step": 758
    },
    {
      "epoch": 0.5693923480870218,
      "grad_norm": 0.11420293897390366,
      "learning_rate": 8.997617215616154e-05,
      "loss": 0.0335,
      "step": 759
    },
    {
      "epoch": 0.5701425356339085,
      "grad_norm": 0.11507721245288849,
      "learning_rate": 8.993681004196797e-05,
      "loss": 0.0282,
      "step": 760
    },
    {
      "epoch": 0.5708927231807952,
      "grad_norm": 0.18800199031829834,
      "learning_rate": 8.989737944007781e-05,
      "loss": 0.0238,
      "step": 761
    },
    {
      "epoch": 0.5716429107276819,
      "grad_norm": 0.174385666847229,
      "learning_rate": 8.985788041811068e-05,
      "loss": 0.0454,
      "step": 762
    },
    {
      "epoch": 0.5723930982745686,
      "grad_norm": 0.14115120470523834,
      "learning_rate": 8.981831304380348e-05,
      "loss": 0.0399,
      "step": 763
    },
    {
      "epoch": 0.5731432858214554,
      "grad_norm": 0.11411836743354797,
      "learning_rate": 8.97786773850104e-05,
      "loss": 0.0284,
      "step": 764
    },
    {
      "epoch": 0.5738934733683421,
      "grad_norm": 0.08721485733985901,
      "learning_rate": 8.973897350970269e-05,
      "loss": 0.0292,
      "step": 765
    },
    {
      "epoch": 0.5746436609152288,
      "grad_norm": 0.17708098888397217,
      "learning_rate": 8.969920148596857e-05,
      "loss": 0.0521,
      "step": 766
    },
    {
      "epoch": 0.5753938484621155,
      "grad_norm": 0.16419649124145508,
      "learning_rate": 8.965936138201314e-05,
      "loss": 0.0354,
      "step": 767
    },
    {
      "epoch": 0.5761440360090022,
      "grad_norm": 0.16116437315940857,
      "learning_rate": 8.961945326615829e-05,
      "loss": 0.0466,
      "step": 768
    },
    {
      "epoch": 0.5768942235558889,
      "grad_norm": 0.09195220470428467,
      "learning_rate": 8.957947720684246e-05,
      "loss": 0.0319,
      "step": 769
    },
    {
      "epoch": 0.5776444111027756,
      "grad_norm": 0.13473393023014069,
      "learning_rate": 8.953943327262066e-05,
      "loss": 0.0383,
      "step": 770
    },
    {
      "epoch": 0.5783945986496624,
      "grad_norm": 0.14092779159545898,
      "learning_rate": 8.949932153216434e-05,
      "loss": 0.0282,
      "step": 771
    },
    {
      "epoch": 0.5791447861965492,
      "grad_norm": 0.07207917422056198,
      "learning_rate": 8.945914205426116e-05,
      "loss": 0.0196,
      "step": 772
    },
    {
      "epoch": 0.5798949737434359,
      "grad_norm": 0.14177407324314117,
      "learning_rate": 8.941889490781494e-05,
      "loss": 0.0343,
      "step": 773
    },
    {
      "epoch": 0.5806451612903226,
      "grad_norm": 0.12721553444862366,
      "learning_rate": 8.937858016184563e-05,
      "loss": 0.0355,
      "step": 774
    },
    {
      "epoch": 0.5813953488372093,
      "grad_norm": 0.11577153205871582,
      "learning_rate": 8.933819788548899e-05,
      "loss": 0.0388,
      "step": 775
    },
    {
      "epoch": 0.582145536384096,
      "grad_norm": 0.06961927562952042,
      "learning_rate": 8.92977481479967e-05,
      "loss": 0.0222,
      "step": 776
    },
    {
      "epoch": 0.5828957239309828,
      "grad_norm": 0.12361857295036316,
      "learning_rate": 8.925723101873603e-05,
      "loss": 0.0346,
      "step": 777
    },
    {
      "epoch": 0.5836459114778695,
      "grad_norm": 0.08191924542188644,
      "learning_rate": 8.92166465671899e-05,
      "loss": 0.027,
      "step": 778
    },
    {
      "epoch": 0.5843960990247562,
      "grad_norm": 0.17663846909999847,
      "learning_rate": 8.917599486295664e-05,
      "loss": 0.0468,
      "step": 779
    },
    {
      "epoch": 0.5851462865716429,
      "grad_norm": 0.09279919415712357,
      "learning_rate": 8.913527597574991e-05,
      "loss": 0.0266,
      "step": 780
    },
    {
      "epoch": 0.5858964741185296,
      "grad_norm": 0.0862743929028511,
      "learning_rate": 8.90944899753986e-05,
      "loss": 0.0243,
      "step": 781
    },
    {
      "epoch": 0.5866466616654163,
      "grad_norm": 0.1053123027086258,
      "learning_rate": 8.905363693184668e-05,
      "loss": 0.0215,
      "step": 782
    },
    {
      "epoch": 0.5873968492123031,
      "grad_norm": 0.15901346504688263,
      "learning_rate": 8.901271691515309e-05,
      "loss": 0.0405,
      "step": 783
    },
    {
      "epoch": 0.5881470367591898,
      "grad_norm": 0.09581103920936584,
      "learning_rate": 8.897172999549165e-05,
      "loss": 0.0254,
      "step": 784
    },
    {
      "epoch": 0.5888972243060765,
      "grad_norm": 0.09868829697370529,
      "learning_rate": 8.893067624315088e-05,
      "loss": 0.0188,
      "step": 785
    },
    {
      "epoch": 0.5896474118529632,
      "grad_norm": 0.16848154366016388,
      "learning_rate": 8.888955572853392e-05,
      "loss": 0.0334,
      "step": 786
    },
    {
      "epoch": 0.5903975993998499,
      "grad_norm": 0.08405663818120956,
      "learning_rate": 8.884836852215841e-05,
      "loss": 0.0233,
      "step": 787
    },
    {
      "epoch": 0.5911477869467366,
      "grad_norm": 0.20354647934436798,
      "learning_rate": 8.880711469465635e-05,
      "loss": 0.0414,
      "step": 788
    },
    {
      "epoch": 0.5918979744936234,
      "grad_norm": 0.10922642797231674,
      "learning_rate": 8.876579431677398e-05,
      "loss": 0.0304,
      "step": 789
    },
    {
      "epoch": 0.5926481620405101,
      "grad_norm": 0.17113949358463287,
      "learning_rate": 8.87244074593717e-05,
      "loss": 0.035,
      "step": 790
    },
    {
      "epoch": 0.5933983495873969,
      "grad_norm": 0.16370461881160736,
      "learning_rate": 8.868295419342389e-05,
      "loss": 0.0338,
      "step": 791
    },
    {
      "epoch": 0.5941485371342836,
      "grad_norm": 0.15194040536880493,
      "learning_rate": 8.86414345900188e-05,
      "loss": 0.0466,
      "step": 792
    },
    {
      "epoch": 0.5948987246811703,
      "grad_norm": 0.13242392241954803,
      "learning_rate": 8.859984872035849e-05,
      "loss": 0.0378,
      "step": 793
    },
    {
      "epoch": 0.595648912228057,
      "grad_norm": 0.10735834389925003,
      "learning_rate": 8.85581966557586e-05,
      "loss": 0.0319,
      "step": 794
    },
    {
      "epoch": 0.5963990997749438,
      "grad_norm": 0.11282511800527573,
      "learning_rate": 8.851647846764835e-05,
      "loss": 0.0244,
      "step": 795
    },
    {
      "epoch": 0.5971492873218305,
      "grad_norm": 0.14782768487930298,
      "learning_rate": 8.847469422757031e-05,
      "loss": 0.0274,
      "step": 796
    },
    {
      "epoch": 0.5978994748687172,
      "grad_norm": 0.09905406087636948,
      "learning_rate": 8.843284400718033e-05,
      "loss": 0.0215,
      "step": 797
    },
    {
      "epoch": 0.5986496624156039,
      "grad_norm": 0.1274043172597885,
      "learning_rate": 8.839092787824743e-05,
      "loss": 0.0331,
      "step": 798
    },
    {
      "epoch": 0.5993998499624906,
      "grad_norm": 0.06478995829820633,
      "learning_rate": 8.834894591265364e-05,
      "loss": 0.0224,
      "step": 799
    },
    {
      "epoch": 0.6001500375093773,
      "grad_norm": 0.09536460041999817,
      "learning_rate": 8.830689818239388e-05,
      "loss": 0.0267,
      "step": 800
    },
    {
      "epoch": 0.6001500375093773,
      "eval_loss": 0.034302037209272385,
      "eval_runtime": 5.1278,
      "eval_samples_per_second": 10.531,
      "eval_steps_per_second": 2.73,
      "step": 800
    },
    {
      "epoch": 0.6009002250562641,
      "grad_norm": 0.09608499705791473,
      "learning_rate": 8.826478475957589e-05,
      "loss": 0.0193,
      "step": 801
    },
    {
      "epoch": 0.6016504126031508,
      "grad_norm": 0.11023902148008347,
      "learning_rate": 8.822260571642005e-05,
      "loss": 0.027,
      "step": 802
    },
    {
      "epoch": 0.6024006001500375,
      "grad_norm": 0.07215134054422379,
      "learning_rate": 8.818036112525924e-05,
      "loss": 0.0268,
      "step": 803
    },
    {
      "epoch": 0.6031507876969242,
      "grad_norm": 0.14665848016738892,
      "learning_rate": 8.813805105853879e-05,
      "loss": 0.046,
      "step": 804
    },
    {
      "epoch": 0.6039009752438109,
      "grad_norm": 0.08387989550828934,
      "learning_rate": 8.809567558881628e-05,
      "loss": 0.0212,
      "step": 805
    },
    {
      "epoch": 0.6046511627906976,
      "grad_norm": 0.15180405974388123,
      "learning_rate": 8.805323478876149e-05,
      "loss": 0.0384,
      "step": 806
    },
    {
      "epoch": 0.6054013503375844,
      "grad_norm": 0.12858478724956512,
      "learning_rate": 8.80107287311562e-05,
      "loss": 0.0353,
      "step": 807
    },
    {
      "epoch": 0.6061515378844711,
      "grad_norm": 0.11187388002872467,
      "learning_rate": 8.796815748889413e-05,
      "loss": 0.0276,
      "step": 808
    },
    {
      "epoch": 0.6069017254313578,
      "grad_norm": 0.07408938556909561,
      "learning_rate": 8.792552113498073e-05,
      "loss": 0.0224,
      "step": 809
    },
    {
      "epoch": 0.6076519129782446,
      "grad_norm": 0.1739557683467865,
      "learning_rate": 8.788281974253318e-05,
      "loss": 0.0476,
      "step": 810
    },
    {
      "epoch": 0.6084021005251313,
      "grad_norm": 0.1207609549164772,
      "learning_rate": 8.784005338478017e-05,
      "loss": 0.0421,
      "step": 811
    },
    {
      "epoch": 0.609152288072018,
      "grad_norm": 0.1739661693572998,
      "learning_rate": 8.779722213506178e-05,
      "loss": 0.0517,
      "step": 812
    },
    {
      "epoch": 0.6099024756189048,
      "grad_norm": 0.1354176253080368,
      "learning_rate": 8.775432606682937e-05,
      "loss": 0.049,
      "step": 813
    },
    {
      "epoch": 0.6106526631657915,
      "grad_norm": 0.12110552936792374,
      "learning_rate": 8.77113652536455e-05,
      "loss": 0.0367,
      "step": 814
    },
    {
      "epoch": 0.6114028507126782,
      "grad_norm": 0.26730018854141235,
      "learning_rate": 8.766833976918371e-05,
      "loss": 0.0467,
      "step": 815
    },
    {
      "epoch": 0.6121530382595649,
      "grad_norm": 0.19855856895446777,
      "learning_rate": 8.76252496872285e-05,
      "loss": 0.04,
      "step": 816
    },
    {
      "epoch": 0.6129032258064516,
      "grad_norm": 0.07605290412902832,
      "learning_rate": 8.758209508167508e-05,
      "loss": 0.0223,
      "step": 817
    },
    {
      "epoch": 0.6136534133533383,
      "grad_norm": 0.11542303115129471,
      "learning_rate": 8.753887602652937e-05,
      "loss": 0.0439,
      "step": 818
    },
    {
      "epoch": 0.614403600900225,
      "grad_norm": 0.1014133095741272,
      "learning_rate": 8.74955925959078e-05,
      "loss": 0.0233,
      "step": 819
    },
    {
      "epoch": 0.6151537884471118,
      "grad_norm": 0.14494650065898895,
      "learning_rate": 8.745224486403718e-05,
      "loss": 0.043,
      "step": 820
    },
    {
      "epoch": 0.6159039759939985,
      "grad_norm": 0.1557653397321701,
      "learning_rate": 8.74088329052546e-05,
      "loss": 0.0375,
      "step": 821
    },
    {
      "epoch": 0.6166541635408852,
      "grad_norm": 0.13315404951572418,
      "learning_rate": 8.73653567940073e-05,
      "loss": 0.0413,
      "step": 822
    },
    {
      "epoch": 0.6174043510877719,
      "grad_norm": 0.10892429202795029,
      "learning_rate": 8.732181660485252e-05,
      "loss": 0.0272,
      "step": 823
    },
    {
      "epoch": 0.6181545386346586,
      "grad_norm": 0.10928978025913239,
      "learning_rate": 8.727821241245742e-05,
      "loss": 0.0236,
      "step": 824
    },
    {
      "epoch": 0.6189047261815454,
      "grad_norm": 0.08895453810691833,
      "learning_rate": 8.723454429159888e-05,
      "loss": 0.0205,
      "step": 825
    },
    {
      "epoch": 0.6196549137284321,
      "grad_norm": 0.122958704829216,
      "learning_rate": 8.719081231716341e-05,
      "loss": 0.0264,
      "step": 826
    },
    {
      "epoch": 0.6204051012753188,
      "grad_norm": 0.09298314154148102,
      "learning_rate": 8.714701656414708e-05,
      "loss": 0.021,
      "step": 827
    },
    {
      "epoch": 0.6211552888222055,
      "grad_norm": 0.17051957547664642,
      "learning_rate": 8.710315710765526e-05,
      "loss": 0.0351,
      "step": 828
    },
    {
      "epoch": 0.6219054763690923,
      "grad_norm": 0.15878413617610931,
      "learning_rate": 8.705923402290261e-05,
      "loss": 0.0339,
      "step": 829
    },
    {
      "epoch": 0.622655663915979,
      "grad_norm": 0.08549269288778305,
      "learning_rate": 8.701524738521291e-05,
      "loss": 0.0239,
      "step": 830
    },
    {
      "epoch": 0.6234058514628658,
      "grad_norm": 0.17661143839359283,
      "learning_rate": 8.697119727001887e-05,
      "loss": 0.0477,
      "step": 831
    },
    {
      "epoch": 0.6241560390097525,
      "grad_norm": 0.13153639435768127,
      "learning_rate": 8.692708375286217e-05,
      "loss": 0.0311,
      "step": 832
    },
    {
      "epoch": 0.6249062265566392,
      "grad_norm": 0.18589378893375397,
      "learning_rate": 8.688290690939307e-05,
      "loss": 0.0434,
      "step": 833
    },
    {
      "epoch": 0.6256564141035259,
      "grad_norm": 0.1077880933880806,
      "learning_rate": 8.683866681537054e-05,
      "loss": 0.0281,
      "step": 834
    },
    {
      "epoch": 0.6264066016504126,
      "grad_norm": 0.062748022377491,
      "learning_rate": 8.679436354666202e-05,
      "loss": 0.016,
      "step": 835
    },
    {
      "epoch": 0.6271567891972993,
      "grad_norm": 0.16130079329013824,
      "learning_rate": 8.67499971792432e-05,
      "loss": 0.0409,
      "step": 836
    },
    {
      "epoch": 0.627906976744186,
      "grad_norm": 0.10744091123342514,
      "learning_rate": 8.670556778919805e-05,
      "loss": 0.0301,
      "step": 837
    },
    {
      "epoch": 0.6286571642910728,
      "grad_norm": 0.36741015315055847,
      "learning_rate": 8.666107545271859e-05,
      "loss": 0.0771,
      "step": 838
    },
    {
      "epoch": 0.6294073518379595,
      "grad_norm": 0.1694938838481903,
      "learning_rate": 8.661652024610482e-05,
      "loss": 0.0511,
      "step": 839
    },
    {
      "epoch": 0.6301575393848462,
      "grad_norm": 0.10456728935241699,
      "learning_rate": 8.657190224576453e-05,
      "loss": 0.0228,
      "step": 840
    },
    {
      "epoch": 0.6309077269317329,
      "grad_norm": 0.086213119328022,
      "learning_rate": 8.652722152821318e-05,
      "loss": 0.0285,
      "step": 841
    },
    {
      "epoch": 0.6316579144786196,
      "grad_norm": 0.08290264755487442,
      "learning_rate": 8.64824781700738e-05,
      "loss": 0.0269,
      "step": 842
    },
    {
      "epoch": 0.6324081020255063,
      "grad_norm": 0.12326061725616455,
      "learning_rate": 8.643767224807685e-05,
      "loss": 0.0391,
      "step": 843
    },
    {
      "epoch": 0.6331582895723931,
      "grad_norm": 0.09560957551002502,
      "learning_rate": 8.639280383906008e-05,
      "loss": 0.0352,
      "step": 844
    },
    {
      "epoch": 0.6339084771192798,
      "grad_norm": 0.23914098739624023,
      "learning_rate": 8.634787301996839e-05,
      "loss": 0.0523,
      "step": 845
    },
    {
      "epoch": 0.6346586646661665,
      "grad_norm": 0.10551799088716507,
      "learning_rate": 8.630287986785368e-05,
      "loss": 0.0322,
      "step": 846
    },
    {
      "epoch": 0.6354088522130532,
      "grad_norm": 0.12217076122760773,
      "learning_rate": 8.625782445987483e-05,
      "loss": 0.0344,
      "step": 847
    },
    {
      "epoch": 0.63615903975994,
      "grad_norm": 0.1267174631357193,
      "learning_rate": 8.621270687329738e-05,
      "loss": 0.0341,
      "step": 848
    },
    {
      "epoch": 0.6369092273068268,
      "grad_norm": 0.12219149619340897,
      "learning_rate": 8.616752718549359e-05,
      "loss": 0.0411,
      "step": 849
    },
    {
      "epoch": 0.6376594148537135,
      "grad_norm": 0.13776132464408875,
      "learning_rate": 8.612228547394215e-05,
      "loss": 0.0464,
      "step": 850
    },
    {
      "epoch": 0.6384096024006002,
      "grad_norm": 0.19186848402023315,
      "learning_rate": 8.607698181622814e-05,
      "loss": 0.0501,
      "step": 851
    },
    {
      "epoch": 0.6391597899474869,
      "grad_norm": 0.14366553723812103,
      "learning_rate": 8.603161629004287e-05,
      "loss": 0.0402,
      "step": 852
    },
    {
      "epoch": 0.6399099774943736,
      "grad_norm": 0.14673426747322083,
      "learning_rate": 8.598618897318375e-05,
      "loss": 0.0432,
      "step": 853
    },
    {
      "epoch": 0.6406601650412603,
      "grad_norm": 0.1311863660812378,
      "learning_rate": 8.594069994355419e-05,
      "loss": 0.0285,
      "step": 854
    },
    {
      "epoch": 0.641410352588147,
      "grad_norm": 0.1472454071044922,
      "learning_rate": 8.589514927916336e-05,
      "loss": 0.0316,
      "step": 855
    },
    {
      "epoch": 0.6421605401350338,
      "grad_norm": 0.09931936115026474,
      "learning_rate": 8.584953705812615e-05,
      "loss": 0.0284,
      "step": 856
    },
    {
      "epoch": 0.6429107276819205,
      "grad_norm": 0.11538329720497131,
      "learning_rate": 8.580386335866308e-05,
      "loss": 0.0325,
      "step": 857
    },
    {
      "epoch": 0.6436609152288072,
      "grad_norm": 0.09810412675142288,
      "learning_rate": 8.575812825909998e-05,
      "loss": 0.0266,
      "step": 858
    },
    {
      "epoch": 0.6444111027756939,
      "grad_norm": 0.11734863370656967,
      "learning_rate": 8.57123318378681e-05,
      "loss": 0.0347,
      "step": 859
    },
    {
      "epoch": 0.6451612903225806,
      "grad_norm": 0.1653948426246643,
      "learning_rate": 8.566647417350378e-05,
      "loss": 0.0318,
      "step": 860
    },
    {
      "epoch": 0.6459114778694673,
      "grad_norm": 0.14309543371200562,
      "learning_rate": 8.562055534464838e-05,
      "loss": 0.0332,
      "step": 861
    },
    {
      "epoch": 0.6466616654163541,
      "grad_norm": 0.1541086882352829,
      "learning_rate": 8.557457543004819e-05,
      "loss": 0.0399,
      "step": 862
    },
    {
      "epoch": 0.6474118529632408,
      "grad_norm": 0.11427025496959686,
      "learning_rate": 8.552853450855422e-05,
      "loss": 0.0332,
      "step": 863
    },
    {
      "epoch": 0.6481620405101275,
      "grad_norm": 0.10530763864517212,
      "learning_rate": 8.548243265912213e-05,
      "loss": 0.0365,
      "step": 864
    },
    {
      "epoch": 0.6489122280570142,
      "grad_norm": 0.10539809614419937,
      "learning_rate": 8.543626996081202e-05,
      "loss": 0.0223,
      "step": 865
    },
    {
      "epoch": 0.6496624156039009,
      "grad_norm": 0.1444813460111618,
      "learning_rate": 8.539004649278841e-05,
      "loss": 0.0325,
      "step": 866
    },
    {
      "epoch": 0.6504126031507877,
      "grad_norm": 0.10531052201986313,
      "learning_rate": 8.534376233432e-05,
      "loss": 0.0231,
      "step": 867
    },
    {
      "epoch": 0.6511627906976745,
      "grad_norm": 0.0951962098479271,
      "learning_rate": 8.529741756477953e-05,
      "loss": 0.0311,
      "step": 868
    },
    {
      "epoch": 0.6519129782445612,
      "grad_norm": 0.09423990547657013,
      "learning_rate": 8.525101226364374e-05,
      "loss": 0.0328,
      "step": 869
    },
    {
      "epoch": 0.6526631657914479,
      "grad_norm": 0.10251491516828537,
      "learning_rate": 8.520454651049313e-05,
      "loss": 0.0225,
      "step": 870
    },
    {
      "epoch": 0.6534133533383346,
      "grad_norm": 0.09833940863609314,
      "learning_rate": 8.51580203850119e-05,
      "loss": 0.023,
      "step": 871
    },
    {
      "epoch": 0.6541635408852213,
      "grad_norm": 0.10047540813684464,
      "learning_rate": 8.511143396698781e-05,
      "loss": 0.0324,
      "step": 872
    },
    {
      "epoch": 0.654913728432108,
      "grad_norm": 0.3961799740791321,
      "learning_rate": 8.506478733631193e-05,
      "loss": 0.06,
      "step": 873
    },
    {
      "epoch": 0.6556639159789948,
      "grad_norm": 0.10680108517408371,
      "learning_rate": 8.501808057297866e-05,
      "loss": 0.0269,
      "step": 874
    },
    {
      "epoch": 0.6564141035258815,
      "grad_norm": 0.10586558282375336,
      "learning_rate": 8.49713137570855e-05,
      "loss": 0.025,
      "step": 875
    },
    {
      "epoch": 0.6571642910727682,
      "grad_norm": 0.10135544836521149,
      "learning_rate": 8.492448696883292e-05,
      "loss": 0.0236,
      "step": 876
    },
    {
      "epoch": 0.6579144786196549,
      "grad_norm": 0.10831738263368607,
      "learning_rate": 8.487760028852427e-05,
      "loss": 0.0235,
      "step": 877
    },
    {
      "epoch": 0.6586646661665416,
      "grad_norm": 0.11996128410100937,
      "learning_rate": 8.483065379656558e-05,
      "loss": 0.0302,
      "step": 878
    },
    {
      "epoch": 0.6594148537134283,
      "grad_norm": 0.11158309131860733,
      "learning_rate": 8.478364757346546e-05,
      "loss": 0.026,
      "step": 879
    },
    {
      "epoch": 0.660165041260315,
      "grad_norm": 0.0804918184876442,
      "learning_rate": 8.473658169983496e-05,
      "loss": 0.0189,
      "step": 880
    },
    {
      "epoch": 0.6609152288072018,
      "grad_norm": 0.13769303262233734,
      "learning_rate": 8.468945625638742e-05,
      "loss": 0.0333,
      "step": 881
    },
    {
      "epoch": 0.6616654163540885,
      "grad_norm": 0.19014306366443634,
      "learning_rate": 8.464227132393831e-05,
      "loss": 0.0488,
      "step": 882
    },
    {
      "epoch": 0.6624156039009752,
      "grad_norm": 0.11594495177268982,
      "learning_rate": 8.459502698340519e-05,
      "loss": 0.0305,
      "step": 883
    },
    {
      "epoch": 0.6631657914478619,
      "grad_norm": 0.11026813834905624,
      "learning_rate": 8.45477233158074e-05,
      "loss": 0.0232,
      "step": 884
    },
    {
      "epoch": 0.6639159789947486,
      "grad_norm": 0.12670838832855225,
      "learning_rate": 8.450036040226612e-05,
      "loss": 0.0301,
      "step": 885
    },
    {
      "epoch": 0.6646661665416355,
      "grad_norm": 0.09368249028921127,
      "learning_rate": 8.445293832400402e-05,
      "loss": 0.0208,
      "step": 886
    },
    {
      "epoch": 0.6654163540885222,
      "grad_norm": 0.10260524600744247,
      "learning_rate": 8.440545716234538e-05,
      "loss": 0.0235,
      "step": 887
    },
    {
      "epoch": 0.6661665416354089,
      "grad_norm": 0.0919303148984909,
      "learning_rate": 8.435791699871564e-05,
      "loss": 0.0212,
      "step": 888
    },
    {
      "epoch": 0.6669167291822956,
      "grad_norm": 0.19766244292259216,
      "learning_rate": 8.431031791464155e-05,
      "loss": 0.0611,
      "step": 889
    },
    {
      "epoch": 0.6676669167291823,
      "grad_norm": 0.08774958550930023,
      "learning_rate": 8.426265999175081e-05,
      "loss": 0.0241,
      "step": 890
    },
    {
      "epoch": 0.668417104276069,
      "grad_norm": 0.16776250302791595,
      "learning_rate": 8.421494331177208e-05,
      "loss": 0.04,
      "step": 891
    },
    {
      "epoch": 0.6691672918229558,
      "grad_norm": 0.10904256254434586,
      "learning_rate": 8.41671679565348e-05,
      "loss": 0.0304,
      "step": 892
    },
    {
      "epoch": 0.6699174793698425,
      "grad_norm": 0.2265564650297165,
      "learning_rate": 8.411933400796896e-05,
      "loss": 0.047,
      "step": 893
    },
    {
      "epoch": 0.6706676669167292,
      "grad_norm": 0.10385122150182724,
      "learning_rate": 8.407144154810509e-05,
      "loss": 0.0331,
      "step": 894
    },
    {
      "epoch": 0.6714178544636159,
      "grad_norm": 0.11064758896827698,
      "learning_rate": 8.402349065907403e-05,
      "loss": 0.0298,
      "step": 895
    },
    {
      "epoch": 0.6721680420105026,
      "grad_norm": 0.17340891063213348,
      "learning_rate": 8.397548142310685e-05,
      "loss": 0.0362,
      "step": 896
    },
    {
      "epoch": 0.6729182295573893,
      "grad_norm": 0.16428762674331665,
      "learning_rate": 8.392741392253465e-05,
      "loss": 0.0454,
      "step": 897
    },
    {
      "epoch": 0.673668417104276,
      "grad_norm": 0.12004006654024124,
      "learning_rate": 8.387928823978846e-05,
      "loss": 0.0302,
      "step": 898
    },
    {
      "epoch": 0.6744186046511628,
      "grad_norm": 0.07734087109565735,
      "learning_rate": 8.383110445739907e-05,
      "loss": 0.0182,
      "step": 899
    },
    {
      "epoch": 0.6751687921980495,
      "grad_norm": 0.06279458105564117,
      "learning_rate": 8.378286265799698e-05,
      "loss": 0.0239,
      "step": 900
    },
    {
      "epoch": 0.6759189797449362,
      "grad_norm": 0.1248641386628151,
      "learning_rate": 8.373456292431206e-05,
      "loss": 0.0394,
      "step": 901
    },
    {
      "epoch": 0.6766691672918229,
      "grad_norm": 0.10201413929462433,
      "learning_rate": 8.368620533917363e-05,
      "loss": 0.0307,
      "step": 902
    },
    {
      "epoch": 0.6774193548387096,
      "grad_norm": 0.09084980189800262,
      "learning_rate": 8.363778998551018e-05,
      "loss": 0.0331,
      "step": 903
    },
    {
      "epoch": 0.6781695423855963,
      "grad_norm": 0.09921452403068542,
      "learning_rate": 8.358931694634928e-05,
      "loss": 0.0287,
      "step": 904
    },
    {
      "epoch": 0.6789197299324832,
      "grad_norm": 0.1051078587770462,
      "learning_rate": 8.35407863048174e-05,
      "loss": 0.0284,
      "step": 905
    },
    {
      "epoch": 0.6796699174793699,
      "grad_norm": 0.09181918948888779,
      "learning_rate": 8.349219814413984e-05,
      "loss": 0.0207,
      "step": 906
    },
    {
      "epoch": 0.6804201050262566,
      "grad_norm": 0.19342802464962006,
      "learning_rate": 8.344355254764049e-05,
      "loss": 0.0405,
      "step": 907
    },
    {
      "epoch": 0.6811702925731433,
      "grad_norm": 0.0752057135105133,
      "learning_rate": 8.339484959874178e-05,
      "loss": 0.0169,
      "step": 908
    },
    {
      "epoch": 0.68192048012003,
      "grad_norm": 0.1359151005744934,
      "learning_rate": 8.334608938096443e-05,
      "loss": 0.0338,
      "step": 909
    },
    {
      "epoch": 0.6826706676669168,
      "grad_norm": 0.15660357475280762,
      "learning_rate": 8.329727197792744e-05,
      "loss": 0.029,
      "step": 910
    },
    {
      "epoch": 0.6834208552138035,
      "grad_norm": 0.06472530961036682,
      "learning_rate": 8.324839747334787e-05,
      "loss": 0.0203,
      "step": 911
    },
    {
      "epoch": 0.6841710427606902,
      "grad_norm": 0.25108665227890015,
      "learning_rate": 8.319946595104065e-05,
      "loss": 0.0393,
      "step": 912
    },
    {
      "epoch": 0.6849212303075769,
      "grad_norm": 0.08752661198377609,
      "learning_rate": 8.315047749491851e-05,
      "loss": 0.0255,
      "step": 913
    },
    {
      "epoch": 0.6856714178544636,
      "grad_norm": 0.0976942628622055,
      "learning_rate": 8.310143218899187e-05,
      "loss": 0.0242,
      "step": 914
    },
    {
      "epoch": 0.6864216054013503,
      "grad_norm": 0.15198934078216553,
      "learning_rate": 8.305233011736857e-05,
      "loss": 0.0419,
      "step": 915
    },
    {
      "epoch": 0.687171792948237,
      "grad_norm": 0.11538293212652206,
      "learning_rate": 8.300317136425385e-05,
      "loss": 0.0273,
      "step": 916
    },
    {
      "epoch": 0.6879219804951238,
      "grad_norm": 0.16572171449661255,
      "learning_rate": 8.295395601395011e-05,
      "loss": 0.031,
      "step": 917
    },
    {
      "epoch": 0.6886721680420105,
      "grad_norm": 0.12102951854467392,
      "learning_rate": 8.290468415085683e-05,
      "loss": 0.0286,
      "step": 918
    },
    {
      "epoch": 0.6894223555888972,
      "grad_norm": 0.24656544625759125,
      "learning_rate": 8.285535585947042e-05,
      "loss": 0.0555,
      "step": 919
    },
    {
      "epoch": 0.6901725431357839,
      "grad_norm": 0.1117316484451294,
      "learning_rate": 8.280597122438404e-05,
      "loss": 0.0255,
      "step": 920
    },
    {
      "epoch": 0.6909227306826706,
      "grad_norm": 0.15054990351200104,
      "learning_rate": 8.275653033028745e-05,
      "loss": 0.0357,
      "step": 921
    },
    {
      "epoch": 0.6916729182295573,
      "grad_norm": 0.1011425033211708,
      "learning_rate": 8.270703326196696e-05,
      "loss": 0.0295,
      "step": 922
    },
    {
      "epoch": 0.6924231057764441,
      "grad_norm": 0.09935387969017029,
      "learning_rate": 8.265748010430513e-05,
      "loss": 0.029,
      "step": 923
    },
    {
      "epoch": 0.6931732933233309,
      "grad_norm": 0.1303011029958725,
      "learning_rate": 8.260787094228076e-05,
      "loss": 0.0336,
      "step": 924
    },
    {
      "epoch": 0.6939234808702176,
      "grad_norm": 0.09792095422744751,
      "learning_rate": 8.255820586096867e-05,
      "loss": 0.0303,
      "step": 925
    },
    {
      "epoch": 0.6946736684171043,
      "grad_norm": 0.11036955565214157,
      "learning_rate": 8.25084849455396e-05,
      "loss": 0.0345,
      "step": 926
    },
    {
      "epoch": 0.695423855963991,
      "grad_norm": 0.1368400603532791,
      "learning_rate": 8.245870828126e-05,
      "loss": 0.0487,
      "step": 927
    },
    {
      "epoch": 0.6961740435108777,
      "grad_norm": 0.13145577907562256,
      "learning_rate": 8.240887595349197e-05,
      "loss": 0.0379,
      "step": 928
    },
    {
      "epoch": 0.6969242310577645,
      "grad_norm": 0.10650297999382019,
      "learning_rate": 8.235898804769303e-05,
      "loss": 0.034,
      "step": 929
    },
    {
      "epoch": 0.6976744186046512,
      "grad_norm": 0.11778818815946579,
      "learning_rate": 8.230904464941604e-05,
      "loss": 0.0358,
      "step": 930
    },
    {
      "epoch": 0.6984246061515379,
      "grad_norm": 0.12128084897994995,
      "learning_rate": 8.225904584430901e-05,
      "loss": 0.0426,
      "step": 931
    },
    {
      "epoch": 0.6991747936984246,
      "grad_norm": 0.1473531872034073,
      "learning_rate": 8.220899171811495e-05,
      "loss": 0.0399,
      "step": 932
    },
    {
      "epoch": 0.6999249812453113,
      "grad_norm": 0.10071301460266113,
      "learning_rate": 8.215888235667176e-05,
      "loss": 0.0311,
      "step": 933
    },
    {
      "epoch": 0.700675168792198,
      "grad_norm": 0.1421838104724884,
      "learning_rate": 8.210871784591207e-05,
      "loss": 0.0399,
      "step": 934
    },
    {
      "epoch": 0.7014253563390848,
      "grad_norm": 0.11759334057569504,
      "learning_rate": 8.205849827186308e-05,
      "loss": 0.0273,
      "step": 935
    },
    {
      "epoch": 0.7021755438859715,
      "grad_norm": 0.09371329843997955,
      "learning_rate": 8.200822372064641e-05,
      "loss": 0.0282,
      "step": 936
    },
    {
      "epoch": 0.7029257314328582,
      "grad_norm": 0.08694262057542801,
      "learning_rate": 8.195789427847796e-05,
      "loss": 0.0212,
      "step": 937
    },
    {
      "epoch": 0.7036759189797449,
      "grad_norm": 0.08702065050601959,
      "learning_rate": 8.190751003166778e-05,
      "loss": 0.0205,
      "step": 938
    },
    {
      "epoch": 0.7044261065266316,
      "grad_norm": 0.11808331310749054,
      "learning_rate": 8.185707106661986e-05,
      "loss": 0.0356,
      "step": 939
    },
    {
      "epoch": 0.7051762940735183,
      "grad_norm": 0.12836730480194092,
      "learning_rate": 8.18065774698321e-05,
      "loss": 0.0312,
      "step": 940
    },
    {
      "epoch": 0.705926481620405,
      "grad_norm": 0.14114151895046234,
      "learning_rate": 8.175602932789601e-05,
      "loss": 0.0335,
      "step": 941
    },
    {
      "epoch": 0.7066766691672918,
      "grad_norm": 0.17943109571933746,
      "learning_rate": 8.17054267274967e-05,
      "loss": 0.0359,
      "step": 942
    },
    {
      "epoch": 0.7074268567141786,
      "grad_norm": 0.12193148583173752,
      "learning_rate": 8.165476975541264e-05,
      "loss": 0.0343,
      "step": 943
    },
    {
      "epoch": 0.7081770442610653,
      "grad_norm": 0.1203569769859314,
      "learning_rate": 8.160405849851556e-05,
      "loss": 0.0263,
      "step": 944
    },
    {
      "epoch": 0.708927231807952,
      "grad_norm": 0.08798082917928696,
      "learning_rate": 8.155329304377025e-05,
      "loss": 0.0249,
      "step": 945
    },
    {
      "epoch": 0.7096774193548387,
      "grad_norm": 0.12763750553131104,
      "learning_rate": 8.150247347823448e-05,
      "loss": 0.0363,
      "step": 946
    },
    {
      "epoch": 0.7104276069017255,
      "grad_norm": 0.12717604637145996,
      "learning_rate": 8.145159988905879e-05,
      "loss": 0.0406,
      "step": 947
    },
    {
      "epoch": 0.7111777944486122,
      "grad_norm": 0.10531704872846603,
      "learning_rate": 8.140067236348638e-05,
      "loss": 0.0323,
      "step": 948
    },
    {
      "epoch": 0.7119279819954989,
      "grad_norm": 0.15488344430923462,
      "learning_rate": 8.134969098885294e-05,
      "loss": 0.0403,
      "step": 949
    },
    {
      "epoch": 0.7126781695423856,
      "grad_norm": 0.12809032201766968,
      "learning_rate": 8.129865585258653e-05,
      "loss": 0.0329,
      "step": 950
    },
    {
      "epoch": 0.7134283570892723,
      "grad_norm": 0.09889086335897446,
      "learning_rate": 8.124756704220735e-05,
      "loss": 0.0277,
      "step": 951
    },
    {
      "epoch": 0.714178544636159,
      "grad_norm": 0.1394570767879486,
      "learning_rate": 8.11964246453277e-05,
      "loss": 0.0397,
      "step": 952
    },
    {
      "epoch": 0.7149287321830458,
      "grad_norm": 0.1902426928281784,
      "learning_rate": 8.114522874965174e-05,
      "loss": 0.0428,
      "step": 953
    },
    {
      "epoch": 0.7156789197299325,
      "grad_norm": 0.15730929374694824,
      "learning_rate": 8.10939794429754e-05,
      "loss": 0.0356,
      "step": 954
    },
    {
      "epoch": 0.7164291072768192,
      "grad_norm": 0.16836601495742798,
      "learning_rate": 8.10426768131862e-05,
      "loss": 0.0433,
      "step": 955
    },
    {
      "epoch": 0.7171792948237059,
      "grad_norm": 0.09496788680553436,
      "learning_rate": 8.099132094826308e-05,
      "loss": 0.0266,
      "step": 956
    },
    {
      "epoch": 0.7179294823705926,
      "grad_norm": 0.11788193881511688,
      "learning_rate": 8.093991193627631e-05,
      "loss": 0.0386,
      "step": 957
    },
    {
      "epoch": 0.7186796699174793,
      "grad_norm": 0.17791517078876495,
      "learning_rate": 8.088844986538727e-05,
      "loss": 0.0376,
      "step": 958
    },
    {
      "epoch": 0.719429857464366,
      "grad_norm": 0.13959939777851105,
      "learning_rate": 8.083693482384836e-05,
      "loss": 0.0319,
      "step": 959
    },
    {
      "epoch": 0.7201800450112528,
      "grad_norm": 0.21248754858970642,
      "learning_rate": 8.078536690000278e-05,
      "loss": 0.0425,
      "step": 960
    },
    {
      "epoch": 0.7209302325581395,
      "grad_norm": 0.131417915225029,
      "learning_rate": 8.073374618228445e-05,
      "loss": 0.0343,
      "step": 961
    },
    {
      "epoch": 0.7216804201050263,
      "grad_norm": 0.10697544366121292,
      "learning_rate": 8.068207275921782e-05,
      "loss": 0.0337,
      "step": 962
    },
    {
      "epoch": 0.722430607651913,
      "grad_norm": 0.09917831420898438,
      "learning_rate": 8.063034671941774e-05,
      "loss": 0.0225,
      "step": 963
    },
    {
      "epoch": 0.7231807951987997,
      "grad_norm": 0.14584000408649445,
      "learning_rate": 8.057856815158924e-05,
      "loss": 0.0366,
      "step": 964
    },
    {
      "epoch": 0.7239309827456865,
      "grad_norm": 0.08406969904899597,
      "learning_rate": 8.05267371445275e-05,
      "loss": 0.025,
      "step": 965
    },
    {
      "epoch": 0.7246811702925732,
      "grad_norm": 0.13167530298233032,
      "learning_rate": 8.047485378711756e-05,
      "loss": 0.0236,
      "step": 966
    },
    {
      "epoch": 0.7254313578394599,
      "grad_norm": 0.09242197871208191,
      "learning_rate": 8.042291816833429e-05,
      "loss": 0.0261,
      "step": 967
    },
    {
      "epoch": 0.7261815453863466,
      "grad_norm": 0.16659840941429138,
      "learning_rate": 8.037093037724216e-05,
      "loss": 0.0366,
      "step": 968
    },
    {
      "epoch": 0.7269317329332333,
      "grad_norm": 0.12860220670700073,
      "learning_rate": 8.031889050299511e-05,
      "loss": 0.0319,
      "step": 969
    },
    {
      "epoch": 0.72768192048012,
      "grad_norm": 0.1668856143951416,
      "learning_rate": 8.02667986348364e-05,
      "loss": 0.0578,
      "step": 970
    },
    {
      "epoch": 0.7284321080270068,
      "grad_norm": 0.11582063883543015,
      "learning_rate": 8.021465486209846e-05,
      "loss": 0.0374,
      "step": 971
    },
    {
      "epoch": 0.7291822955738935,
      "grad_norm": 0.10971240699291229,
      "learning_rate": 8.016245927420272e-05,
      "loss": 0.0355,
      "step": 972
    },
    {
      "epoch": 0.7299324831207802,
      "grad_norm": 0.11834264546632767,
      "learning_rate": 8.011021196065946e-05,
      "loss": 0.041,
      "step": 973
    },
    {
      "epoch": 0.7306826706676669,
      "grad_norm": 0.14551803469657898,
      "learning_rate": 8.005791301106769e-05,
      "loss": 0.037,
      "step": 974
    },
    {
      "epoch": 0.7314328582145536,
      "grad_norm": 0.10361959785223007,
      "learning_rate": 8.000556251511498e-05,
      "loss": 0.033,
      "step": 975
    },
    {
      "epoch": 0.7321830457614403,
      "grad_norm": 0.08322405815124512,
      "learning_rate": 7.995316056257723e-05,
      "loss": 0.0223,
      "step": 976
    },
    {
      "epoch": 0.732933233308327,
      "grad_norm": 0.10396233201026917,
      "learning_rate": 7.990070724331866e-05,
      "loss": 0.0228,
      "step": 977
    },
    {
      "epoch": 0.7336834208552138,
      "grad_norm": 0.1658412218093872,
      "learning_rate": 7.984820264729156e-05,
      "loss": 0.0441,
      "step": 978
    },
    {
      "epoch": 0.7344336084021005,
      "grad_norm": 0.09091618657112122,
      "learning_rate": 7.979564686453612e-05,
      "loss": 0.0258,
      "step": 979
    },
    {
      "epoch": 0.7351837959489872,
      "grad_norm": 0.10041514039039612,
      "learning_rate": 7.974303998518031e-05,
      "loss": 0.0229,
      "step": 980
    },
    {
      "epoch": 0.735933983495874,
      "grad_norm": 0.15248219668865204,
      "learning_rate": 7.96903820994398e-05,
      "loss": 0.0278,
      "step": 981
    },
    {
      "epoch": 0.7366841710427607,
      "grad_norm": 0.1488843411207199,
      "learning_rate": 7.963767329761762e-05,
      "loss": 0.0473,
      "step": 982
    },
    {
      "epoch": 0.7374343585896475,
      "grad_norm": 0.1008438840508461,
      "learning_rate": 7.958491367010423e-05,
      "loss": 0.027,
      "step": 983
    },
    {
      "epoch": 0.7381845461365342,
      "grad_norm": 0.10297491401433945,
      "learning_rate": 7.953210330737718e-05,
      "loss": 0.0319,
      "step": 984
    },
    {
      "epoch": 0.7389347336834209,
      "grad_norm": 0.11224748194217682,
      "learning_rate": 7.947924230000102e-05,
      "loss": 0.0254,
      "step": 985
    },
    {
      "epoch": 0.7396849212303076,
      "grad_norm": 0.1465216726064682,
      "learning_rate": 7.942633073862718e-05,
      "loss": 0.0352,
      "step": 986
    },
    {
      "epoch": 0.7404351087771943,
      "grad_norm": 0.18327566981315613,
      "learning_rate": 7.937336871399379e-05,
      "loss": 0.0414,
      "step": 987
    },
    {
      "epoch": 0.741185296324081,
      "grad_norm": 0.17351044714450836,
      "learning_rate": 7.932035631692549e-05,
      "loss": 0.0432,
      "step": 988
    },
    {
      "epoch": 0.7419354838709677,
      "grad_norm": 0.2046956866979599,
      "learning_rate": 7.926729363833335e-05,
      "loss": 0.0374,
      "step": 989
    },
    {
      "epoch": 0.7426856714178545,
      "grad_norm": 0.18247826397418976,
      "learning_rate": 7.921418076921461e-05,
      "loss": 0.0433,
      "step": 990
    },
    {
      "epoch": 0.7434358589647412,
      "grad_norm": 0.11379354447126389,
      "learning_rate": 7.916101780065263e-05,
      "loss": 0.0276,
      "step": 991
    },
    {
      "epoch": 0.7441860465116279,
      "grad_norm": 0.10715629160404205,
      "learning_rate": 7.910780482381665e-05,
      "loss": 0.0303,
      "step": 992
    },
    {
      "epoch": 0.7449362340585146,
      "grad_norm": 0.10846918076276779,
      "learning_rate": 7.905454192996169e-05,
      "loss": 0.029,
      "step": 993
    },
    {
      "epoch": 0.7456864216054013,
      "grad_norm": 0.1614159196615219,
      "learning_rate": 7.900122921042837e-05,
      "loss": 0.0332,
      "step": 994
    },
    {
      "epoch": 0.746436609152288,
      "grad_norm": 0.12448745965957642,
      "learning_rate": 7.894786675664277e-05,
      "loss": 0.0313,
      "step": 995
    },
    {
      "epoch": 0.7471867966991748,
      "grad_norm": 0.12879522144794464,
      "learning_rate": 7.88944546601162e-05,
      "loss": 0.0327,
      "step": 996
    },
    {
      "epoch": 0.7479369842460615,
      "grad_norm": 0.08149383217096329,
      "learning_rate": 7.884099301244519e-05,
      "loss": 0.0193,
      "step": 997
    },
    {
      "epoch": 0.7486871717929482,
      "grad_norm": 0.09928504377603531,
      "learning_rate": 7.878748190531118e-05,
      "loss": 0.0201,
      "step": 998
    },
    {
      "epoch": 0.7494373593398349,
      "grad_norm": 0.09547089040279388,
      "learning_rate": 7.873392143048047e-05,
      "loss": 0.0263,
      "step": 999
    },
    {
      "epoch": 0.7501875468867217,
      "grad_norm": 0.15654493868350983,
      "learning_rate": 7.868031167980397e-05,
      "loss": 0.0499,
      "step": 1000
    },
    {
      "epoch": 0.7501875468867217,
      "eval_loss": 0.033354200422763824,
      "eval_runtime": 5.1415,
      "eval_samples_per_second": 10.503,
      "eval_steps_per_second": 2.723,
      "step": 1000
    },
    {
      "epoch": 0.7509377344336085,
      "grad_norm": 0.15055902302265167,
      "learning_rate": 7.862665274521712e-05,
      "loss": 0.0369,
      "step": 1001
    },
    {
      "epoch": 0.7516879219804952,
      "grad_norm": 0.15537412464618683,
      "learning_rate": 7.857294471873975e-05,
      "loss": 0.0342,
      "step": 1002
    },
    {
      "epoch": 0.7524381095273819,
      "grad_norm": 0.20609122514724731,
      "learning_rate": 7.851918769247582e-05,
      "loss": 0.041,
      "step": 1003
    },
    {
      "epoch": 0.7531882970742686,
      "grad_norm": 0.09634378552436829,
      "learning_rate": 7.846538175861332e-05,
      "loss": 0.023,
      "step": 1004
    },
    {
      "epoch": 0.7539384846211553,
      "grad_norm": 0.147333025932312,
      "learning_rate": 7.841152700942413e-05,
      "loss": 0.0266,
      "step": 1005
    },
    {
      "epoch": 0.754688672168042,
      "grad_norm": 0.13298511505126953,
      "learning_rate": 7.835762353726386e-05,
      "loss": 0.0347,
      "step": 1006
    },
    {
      "epoch": 0.7554388597149287,
      "grad_norm": 0.15836963057518005,
      "learning_rate": 7.830367143457165e-05,
      "loss": 0.0396,
      "step": 1007
    },
    {
      "epoch": 0.7561890472618155,
      "grad_norm": 0.3826274275779724,
      "learning_rate": 7.824967079387002e-05,
      "loss": 0.0618,
      "step": 1008
    },
    {
      "epoch": 0.7569392348087022,
      "grad_norm": 0.18430115282535553,
      "learning_rate": 7.81956217077648e-05,
      "loss": 0.0463,
      "step": 1009
    },
    {
      "epoch": 0.7576894223555889,
      "grad_norm": 0.12182758003473282,
      "learning_rate": 7.814152426894478e-05,
      "loss": 0.0296,
      "step": 1010
    },
    {
      "epoch": 0.7584396099024756,
      "grad_norm": 0.1450658142566681,
      "learning_rate": 7.808737857018182e-05,
      "loss": 0.0414,
      "step": 1011
    },
    {
      "epoch": 0.7591897974493623,
      "grad_norm": 0.15949197113513947,
      "learning_rate": 7.803318470433042e-05,
      "loss": 0.0319,
      "step": 1012
    },
    {
      "epoch": 0.759939984996249,
      "grad_norm": 0.1380879431962967,
      "learning_rate": 7.797894276432772e-05,
      "loss": 0.0359,
      "step": 1013
    },
    {
      "epoch": 0.7606901725431358,
      "grad_norm": 0.10468295961618423,
      "learning_rate": 7.792465284319332e-05,
      "loss": 0.0292,
      "step": 1014
    },
    {
      "epoch": 0.7614403600900225,
      "grad_norm": 0.11396298557519913,
      "learning_rate": 7.787031503402907e-05,
      "loss": 0.0277,
      "step": 1015
    },
    {
      "epoch": 0.7621905476369092,
      "grad_norm": 0.154616579413414,
      "learning_rate": 7.781592943001899e-05,
      "loss": 0.0413,
      "step": 1016
    },
    {
      "epoch": 0.7629407351837959,
      "grad_norm": 0.10276743024587631,
      "learning_rate": 7.776149612442899e-05,
      "loss": 0.0276,
      "step": 1017
    },
    {
      "epoch": 0.7636909227306826,
      "grad_norm": 0.15479224920272827,
      "learning_rate": 7.770701521060688e-05,
      "loss": 0.0345,
      "step": 1018
    },
    {
      "epoch": 0.7644411102775694,
      "grad_norm": 0.11939147859811783,
      "learning_rate": 7.765248678198203e-05,
      "loss": 0.0259,
      "step": 1019
    },
    {
      "epoch": 0.7651912978244562,
      "grad_norm": 0.09478700160980225,
      "learning_rate": 7.759791093206534e-05,
      "loss": 0.0259,
      "step": 1020
    },
    {
      "epoch": 0.7659414853713429,
      "grad_norm": 0.1022457480430603,
      "learning_rate": 7.754328775444903e-05,
      "loss": 0.0284,
      "step": 1021
    },
    {
      "epoch": 0.7666916729182296,
      "grad_norm": 0.1470392495393753,
      "learning_rate": 7.748861734280643e-05,
      "loss": 0.0369,
      "step": 1022
    },
    {
      "epoch": 0.7674418604651163,
      "grad_norm": 0.10798605531454086,
      "learning_rate": 7.743389979089196e-05,
      "loss": 0.0243,
      "step": 1023
    },
    {
      "epoch": 0.768192048012003,
      "grad_norm": 0.09053485095500946,
      "learning_rate": 7.737913519254079e-05,
      "loss": 0.0252,
      "step": 1024
    },
    {
      "epoch": 0.7689422355588897,
      "grad_norm": 0.20014923810958862,
      "learning_rate": 7.732432364166884e-05,
      "loss": 0.0538,
      "step": 1025
    },
    {
      "epoch": 0.7696924231057765,
      "grad_norm": 0.2505134642124176,
      "learning_rate": 7.726946523227251e-05,
      "loss": 0.0463,
      "step": 1026
    },
    {
      "epoch": 0.7704426106526632,
      "grad_norm": 0.09509759396314621,
      "learning_rate": 7.721456005842861e-05,
      "loss": 0.0261,
      "step": 1027
    },
    {
      "epoch": 0.7711927981995499,
      "grad_norm": 0.09843521565198898,
      "learning_rate": 7.715960821429404e-05,
      "loss": 0.0269,
      "step": 1028
    },
    {
      "epoch": 0.7719429857464366,
      "grad_norm": 0.13212156295776367,
      "learning_rate": 7.710460979410585e-05,
      "loss": 0.0334,
      "step": 1029
    },
    {
      "epoch": 0.7726931732933233,
      "grad_norm": 0.18599221110343933,
      "learning_rate": 7.704956489218091e-05,
      "loss": 0.0477,
      "step": 1030
    },
    {
      "epoch": 0.77344336084021,
      "grad_norm": 0.15956397354602814,
      "learning_rate": 7.699447360291576e-05,
      "loss": 0.0424,
      "step": 1031
    },
    {
      "epoch": 0.7741935483870968,
      "grad_norm": 0.10117616504430771,
      "learning_rate": 7.69393360207866e-05,
      "loss": 0.0224,
      "step": 1032
    },
    {
      "epoch": 0.7749437359339835,
      "grad_norm": 0.09441590309143066,
      "learning_rate": 7.688415224034893e-05,
      "loss": 0.0273,
      "step": 1033
    },
    {
      "epoch": 0.7756939234808702,
      "grad_norm": 0.08132372796535492,
      "learning_rate": 7.682892235623749e-05,
      "loss": 0.0216,
      "step": 1034
    },
    {
      "epoch": 0.7764441110277569,
      "grad_norm": 0.08580277860164642,
      "learning_rate": 7.67736464631661e-05,
      "loss": 0.0218,
      "step": 1035
    },
    {
      "epoch": 0.7771942985746436,
      "grad_norm": 0.14287903904914856,
      "learning_rate": 7.671832465592746e-05,
      "loss": 0.0366,
      "step": 1036
    },
    {
      "epoch": 0.7779444861215303,
      "grad_norm": 0.11992555856704712,
      "learning_rate": 7.666295702939305e-05,
      "loss": 0.0336,
      "step": 1037
    },
    {
      "epoch": 0.7786946736684172,
      "grad_norm": 0.12070818245410919,
      "learning_rate": 7.660754367851286e-05,
      "loss": 0.0308,
      "step": 1038
    },
    {
      "epoch": 0.7794448612153039,
      "grad_norm": 0.17960987985134125,
      "learning_rate": 7.655208469831536e-05,
      "loss": 0.0467,
      "step": 1039
    },
    {
      "epoch": 0.7801950487621906,
      "grad_norm": 0.1322227567434311,
      "learning_rate": 7.649658018390725e-05,
      "loss": 0.0258,
      "step": 1040
    },
    {
      "epoch": 0.7809452363090773,
      "grad_norm": 0.07779379934072495,
      "learning_rate": 7.644103023047327e-05,
      "loss": 0.0215,
      "step": 1041
    },
    {
      "epoch": 0.781695423855964,
      "grad_norm": 0.11337679624557495,
      "learning_rate": 7.638543493327613e-05,
      "loss": 0.0267,
      "step": 1042
    },
    {
      "epoch": 0.7824456114028507,
      "grad_norm": 0.1161460429430008,
      "learning_rate": 7.63297943876563e-05,
      "loss": 0.0345,
      "step": 1043
    },
    {
      "epoch": 0.7831957989497375,
      "grad_norm": 0.12373264878988266,
      "learning_rate": 7.627410868903184e-05,
      "loss": 0.0338,
      "step": 1044
    },
    {
      "epoch": 0.7839459864966242,
      "grad_norm": 0.14728686213493347,
      "learning_rate": 7.621837793289824e-05,
      "loss": 0.0487,
      "step": 1045
    },
    {
      "epoch": 0.7846961740435109,
      "grad_norm": 0.16242580115795135,
      "learning_rate": 7.616260221482825e-05,
      "loss": 0.0378,
      "step": 1046
    },
    {
      "epoch": 0.7854463615903976,
      "grad_norm": 0.17540179193019867,
      "learning_rate": 7.610678163047174e-05,
      "loss": 0.0487,
      "step": 1047
    },
    {
      "epoch": 0.7861965491372843,
      "grad_norm": 0.0956827774643898,
      "learning_rate": 7.60509162755555e-05,
      "loss": 0.0257,
      "step": 1048
    },
    {
      "epoch": 0.786946736684171,
      "grad_norm": 0.14103733003139496,
      "learning_rate": 7.599500624588314e-05,
      "loss": 0.0357,
      "step": 1049
    },
    {
      "epoch": 0.7876969242310577,
      "grad_norm": 0.1205405592918396,
      "learning_rate": 7.593905163733484e-05,
      "loss": 0.0335,
      "step": 1050
    },
    {
      "epoch": 0.7884471117779445,
      "grad_norm": 0.1581084281206131,
      "learning_rate": 7.588305254586724e-05,
      "loss": 0.0392,
      "step": 1051
    },
    {
      "epoch": 0.7891972993248312,
      "grad_norm": 0.08614162355661392,
      "learning_rate": 7.582700906751325e-05,
      "loss": 0.0246,
      "step": 1052
    },
    {
      "epoch": 0.7899474868717179,
      "grad_norm": 0.15688768029212952,
      "learning_rate": 7.577092129838197e-05,
      "loss": 0.0421,
      "step": 1053
    },
    {
      "epoch": 0.7906976744186046,
      "grad_norm": 0.08980410546064377,
      "learning_rate": 7.571478933465836e-05,
      "loss": 0.0237,
      "step": 1054
    },
    {
      "epoch": 0.7914478619654913,
      "grad_norm": 0.11821522563695908,
      "learning_rate": 7.565861327260322e-05,
      "loss": 0.0315,
      "step": 1055
    },
    {
      "epoch": 0.792198049512378,
      "grad_norm": 0.2338874340057373,
      "learning_rate": 7.560239320855296e-05,
      "loss": 0.066,
      "step": 1056
    },
    {
      "epoch": 0.7929482370592649,
      "grad_norm": 0.2845582365989685,
      "learning_rate": 7.554612923891946e-05,
      "loss": 0.0546,
      "step": 1057
    },
    {
      "epoch": 0.7936984246061516,
      "grad_norm": 0.15993359684944153,
      "learning_rate": 7.548982146018988e-05,
      "loss": 0.0538,
      "step": 1058
    },
    {
      "epoch": 0.7944486121530383,
      "grad_norm": 0.12397005409002304,
      "learning_rate": 7.543346996892654e-05,
      "loss": 0.0415,
      "step": 1059
    },
    {
      "epoch": 0.795198799699925,
      "grad_norm": 0.11393854767084122,
      "learning_rate": 7.537707486176667e-05,
      "loss": 0.0327,
      "step": 1060
    },
    {
      "epoch": 0.7959489872468117,
      "grad_norm": 0.0845160260796547,
      "learning_rate": 7.532063623542231e-05,
      "loss": 0.031,
      "step": 1061
    },
    {
      "epoch": 0.7966991747936985,
      "grad_norm": 0.11460570245981216,
      "learning_rate": 7.52641541866802e-05,
      "loss": 0.0232,
      "step": 1062
    },
    {
      "epoch": 0.7974493623405852,
      "grad_norm": 0.13715198636054993,
      "learning_rate": 7.520762881240147e-05,
      "loss": 0.0486,
      "step": 1063
    },
    {
      "epoch": 0.7981995498874719,
      "grad_norm": 0.14128798246383667,
      "learning_rate": 7.515106020952156e-05,
      "loss": 0.044,
      "step": 1064
    },
    {
      "epoch": 0.7989497374343586,
      "grad_norm": 0.09283804148435593,
      "learning_rate": 7.509444847505005e-05,
      "loss": 0.0308,
      "step": 1065
    },
    {
      "epoch": 0.7996999249812453,
      "grad_norm": 0.12887823581695557,
      "learning_rate": 7.503779370607049e-05,
      "loss": 0.0282,
      "step": 1066
    },
    {
      "epoch": 0.800450112528132,
      "grad_norm": 0.10709967464208603,
      "learning_rate": 7.498109599974024e-05,
      "loss": 0.0328,
      "step": 1067
    },
    {
      "epoch": 0.8012003000750187,
      "grad_norm": 0.09748120605945587,
      "learning_rate": 7.49243554532903e-05,
      "loss": 0.0306,
      "step": 1068
    },
    {
      "epoch": 0.8019504876219055,
      "grad_norm": 0.13216429948806763,
      "learning_rate": 7.486757216402509e-05,
      "loss": 0.0276,
      "step": 1069
    },
    {
      "epoch": 0.8027006751687922,
      "grad_norm": 0.10359996557235718,
      "learning_rate": 7.481074622932236e-05,
      "loss": 0.0302,
      "step": 1070
    },
    {
      "epoch": 0.8034508627156789,
      "grad_norm": 0.10987622290849686,
      "learning_rate": 7.475387774663302e-05,
      "loss": 0.0258,
      "step": 1071
    },
    {
      "epoch": 0.8042010502625656,
      "grad_norm": 0.12450256943702698,
      "learning_rate": 7.469696681348088e-05,
      "loss": 0.0281,
      "step": 1072
    },
    {
      "epoch": 0.8049512378094523,
      "grad_norm": 0.2062714844942093,
      "learning_rate": 7.464001352746263e-05,
      "loss": 0.052,
      "step": 1073
    },
    {
      "epoch": 0.805701425356339,
      "grad_norm": 0.12377651035785675,
      "learning_rate": 7.45830179862475e-05,
      "loss": 0.0331,
      "step": 1074
    },
    {
      "epoch": 0.8064516129032258,
      "grad_norm": 0.09847312420606613,
      "learning_rate": 7.452598028757729e-05,
      "loss": 0.0373,
      "step": 1075
    },
    {
      "epoch": 0.8072018004501126,
      "grad_norm": 0.1826242357492447,
      "learning_rate": 7.446890052926598e-05,
      "loss": 0.0441,
      "step": 1076
    },
    {
      "epoch": 0.8079519879969993,
      "grad_norm": 0.0950358659029007,
      "learning_rate": 7.441177880919976e-05,
      "loss": 0.0283,
      "step": 1077
    },
    {
      "epoch": 0.808702175543886,
      "grad_norm": 0.13997478783130646,
      "learning_rate": 7.435461522533674e-05,
      "loss": 0.034,
      "step": 1078
    },
    {
      "epoch": 0.8094523630907727,
      "grad_norm": 0.12771092355251312,
      "learning_rate": 7.429740987570686e-05,
      "loss": 0.0349,
      "step": 1079
    },
    {
      "epoch": 0.8102025506376594,
      "grad_norm": 0.10692954063415527,
      "learning_rate": 7.424016285841165e-05,
      "loss": 0.0281,
      "step": 1080
    },
    {
      "epoch": 0.8109527381845462,
      "grad_norm": 0.11578011512756348,
      "learning_rate": 7.41828742716241e-05,
      "loss": 0.0303,
      "step": 1081
    },
    {
      "epoch": 0.8117029257314329,
      "grad_norm": 0.17760562896728516,
      "learning_rate": 7.41255442135885e-05,
      "loss": 0.0412,
      "step": 1082
    },
    {
      "epoch": 0.8124531132783196,
      "grad_norm": 0.236124649643898,
      "learning_rate": 7.406817278262027e-05,
      "loss": 0.0483,
      "step": 1083
    },
    {
      "epoch": 0.8132033008252063,
      "grad_norm": 0.08512605726718903,
      "learning_rate": 7.401076007710575e-05,
      "loss": 0.0261,
      "step": 1084
    },
    {
      "epoch": 0.813953488372093,
      "grad_norm": 0.10752329975366592,
      "learning_rate": 7.395330619550207e-05,
      "loss": 0.0291,
      "step": 1085
    },
    {
      "epoch": 0.8147036759189797,
      "grad_norm": 0.17289288341999054,
      "learning_rate": 7.3895811236337e-05,
      "loss": 0.0485,
      "step": 1086
    },
    {
      "epoch": 0.8154538634658665,
      "grad_norm": 0.143681600689888,
      "learning_rate": 7.38382752982087e-05,
      "loss": 0.0384,
      "step": 1087
    },
    {
      "epoch": 0.8162040510127532,
      "grad_norm": 0.15613281726837158,
      "learning_rate": 7.378069847978568e-05,
      "loss": 0.0471,
      "step": 1088
    },
    {
      "epoch": 0.8169542385596399,
      "grad_norm": 0.16713473200798035,
      "learning_rate": 7.372308087980647e-05,
      "loss": 0.0347,
      "step": 1089
    },
    {
      "epoch": 0.8177044261065266,
      "grad_norm": 0.1506505310535431,
      "learning_rate": 7.366542259707962e-05,
      "loss": 0.035,
      "step": 1090
    },
    {
      "epoch": 0.8184546136534133,
      "grad_norm": 0.14795953035354614,
      "learning_rate": 7.360772373048338e-05,
      "loss": 0.0426,
      "step": 1091
    },
    {
      "epoch": 0.8192048012003,
      "grad_norm": 0.095690056681633,
      "learning_rate": 7.354998437896565e-05,
      "loss": 0.0278,
      "step": 1092
    },
    {
      "epoch": 0.8199549887471868,
      "grad_norm": 0.13107137382030487,
      "learning_rate": 7.349220464154371e-05,
      "loss": 0.0433,
      "step": 1093
    },
    {
      "epoch": 0.8207051762940735,
      "grad_norm": 0.08680702745914459,
      "learning_rate": 7.343438461730411e-05,
      "loss": 0.0194,
      "step": 1094
    },
    {
      "epoch": 0.8214553638409603,
      "grad_norm": 0.12979038059711456,
      "learning_rate": 7.337652440540252e-05,
      "loss": 0.042,
      "step": 1095
    },
    {
      "epoch": 0.822205551387847,
      "grad_norm": 0.14768902957439423,
      "learning_rate": 7.331862410506353e-05,
      "loss": 0.038,
      "step": 1096
    },
    {
      "epoch": 0.8229557389347337,
      "grad_norm": 0.14361713826656342,
      "learning_rate": 7.32606838155804e-05,
      "loss": 0.029,
      "step": 1097
    },
    {
      "epoch": 0.8237059264816204,
      "grad_norm": 0.09816596657037735,
      "learning_rate": 7.320270363631505e-05,
      "loss": 0.0312,
      "step": 1098
    },
    {
      "epoch": 0.8244561140285072,
      "grad_norm": 0.13732968270778656,
      "learning_rate": 7.314468366669777e-05,
      "loss": 0.0355,
      "step": 1099
    },
    {
      "epoch": 0.8252063015753939,
      "grad_norm": 0.11866525560617447,
      "learning_rate": 7.30866240062271e-05,
      "loss": 0.0293,
      "step": 1100
    },
    {
      "epoch": 0.8259564891222806,
      "grad_norm": 0.1313275694847107,
      "learning_rate": 7.302852475446963e-05,
      "loss": 0.0384,
      "step": 1101
    },
    {
      "epoch": 0.8267066766691673,
      "grad_norm": 0.12741775810718536,
      "learning_rate": 7.297038601105988e-05,
      "loss": 0.042,
      "step": 1102
    },
    {
      "epoch": 0.827456864216054,
      "grad_norm": 0.12329462915658951,
      "learning_rate": 7.291220787570005e-05,
      "loss": 0.0292,
      "step": 1103
    },
    {
      "epoch": 0.8282070517629407,
      "grad_norm": 0.1349187195301056,
      "learning_rate": 7.28539904481599e-05,
      "loss": 0.0427,
      "step": 1104
    },
    {
      "epoch": 0.8289572393098275,
      "grad_norm": 0.10525055974721909,
      "learning_rate": 7.279573382827662e-05,
      "loss": 0.0241,
      "step": 1105
    },
    {
      "epoch": 0.8297074268567142,
      "grad_norm": 0.08475113660097122,
      "learning_rate": 7.273743811595454e-05,
      "loss": 0.0231,
      "step": 1106
    },
    {
      "epoch": 0.8304576144036009,
      "grad_norm": 0.13441896438598633,
      "learning_rate": 7.267910341116512e-05,
      "loss": 0.0374,
      "step": 1107
    },
    {
      "epoch": 0.8312078019504876,
      "grad_norm": 0.12029675394296646,
      "learning_rate": 7.262072981394656e-05,
      "loss": 0.0267,
      "step": 1108
    },
    {
      "epoch": 0.8319579894973743,
      "grad_norm": 0.16057652235031128,
      "learning_rate": 7.256231742440389e-05,
      "loss": 0.0457,
      "step": 1109
    },
    {
      "epoch": 0.832708177044261,
      "grad_norm": 0.17000728845596313,
      "learning_rate": 7.25038663427086e-05,
      "loss": 0.0313,
      "step": 1110
    },
    {
      "epoch": 0.8334583645911477,
      "grad_norm": 0.18660447001457214,
      "learning_rate": 7.24453766690985e-05,
      "loss": 0.0315,
      "step": 1111
    },
    {
      "epoch": 0.8342085521380345,
      "grad_norm": 0.1456262469291687,
      "learning_rate": 7.238684850387765e-05,
      "loss": 0.0291,
      "step": 1112
    },
    {
      "epoch": 0.8349587396849212,
      "grad_norm": 0.142116978764534,
      "learning_rate": 7.232828194741611e-05,
      "loss": 0.0404,
      "step": 1113
    },
    {
      "epoch": 0.8357089272318079,
      "grad_norm": 0.17252390086650848,
      "learning_rate": 7.226967710014971e-05,
      "loss": 0.0371,
      "step": 1114
    },
    {
      "epoch": 0.8364591147786947,
      "grad_norm": 0.12147200107574463,
      "learning_rate": 7.221103406258003e-05,
      "loss": 0.024,
      "step": 1115
    },
    {
      "epoch": 0.8372093023255814,
      "grad_norm": 0.09407100081443787,
      "learning_rate": 7.215235293527409e-05,
      "loss": 0.0267,
      "step": 1116
    },
    {
      "epoch": 0.8379594898724682,
      "grad_norm": 0.1135459691286087,
      "learning_rate": 7.209363381886423e-05,
      "loss": 0.0363,
      "step": 1117
    },
    {
      "epoch": 0.8387096774193549,
      "grad_norm": 0.08867119997739792,
      "learning_rate": 7.203487681404798e-05,
      "loss": 0.0192,
      "step": 1118
    },
    {
      "epoch": 0.8394598649662416,
      "grad_norm": 0.13186821341514587,
      "learning_rate": 7.19760820215878e-05,
      "loss": 0.0371,
      "step": 1119
    },
    {
      "epoch": 0.8402100525131283,
      "grad_norm": 0.14221522212028503,
      "learning_rate": 7.191724954231098e-05,
      "loss": 0.0341,
      "step": 1120
    },
    {
      "epoch": 0.840960240060015,
      "grad_norm": 0.10038655251264572,
      "learning_rate": 7.185837947710943e-05,
      "loss": 0.0271,
      "step": 1121
    },
    {
      "epoch": 0.8417104276069017,
      "grad_norm": 0.09764457494020462,
      "learning_rate": 7.17994719269395e-05,
      "loss": 0.0276,
      "step": 1122
    },
    {
      "epoch": 0.8424606151537885,
      "grad_norm": 0.11716485768556595,
      "learning_rate": 7.174052699282183e-05,
      "loss": 0.0355,
      "step": 1123
    },
    {
      "epoch": 0.8432108027006752,
      "grad_norm": 0.517447829246521,
      "learning_rate": 7.168154477584123e-05,
      "loss": 0.0535,
      "step": 1124
    },
    {
      "epoch": 0.8439609902475619,
      "grad_norm": 0.1132436990737915,
      "learning_rate": 7.162252537714633e-05,
      "loss": 0.0324,
      "step": 1125
    },
    {
      "epoch": 0.8447111777944486,
      "grad_norm": 0.0971536934375763,
      "learning_rate": 7.156346889794962e-05,
      "loss": 0.0354,
      "step": 1126
    },
    {
      "epoch": 0.8454613653413353,
      "grad_norm": 0.18410785496234894,
      "learning_rate": 7.150437543952715e-05,
      "loss": 0.0564,
      "step": 1127
    },
    {
      "epoch": 0.846211552888222,
      "grad_norm": 0.07700716704130173,
      "learning_rate": 7.144524510321837e-05,
      "loss": 0.0267,
      "step": 1128
    },
    {
      "epoch": 0.8469617404351087,
      "grad_norm": 0.14188776910305023,
      "learning_rate": 7.138607799042598e-05,
      "loss": 0.0381,
      "step": 1129
    },
    {
      "epoch": 0.8477119279819955,
      "grad_norm": 0.13262468576431274,
      "learning_rate": 7.132687420261576e-05,
      "loss": 0.0405,
      "step": 1130
    },
    {
      "epoch": 0.8484621155288822,
      "grad_norm": 0.15925203263759613,
      "learning_rate": 7.126763384131638e-05,
      "loss": 0.035,
      "step": 1131
    },
    {
      "epoch": 0.8492123030757689,
      "grad_norm": 0.09985039383172989,
      "learning_rate": 7.120835700811923e-05,
      "loss": 0.0285,
      "step": 1132
    },
    {
      "epoch": 0.8499624906226556,
      "grad_norm": 0.09418608248233795,
      "learning_rate": 7.114904380467823e-05,
      "loss": 0.0273,
      "step": 1133
    },
    {
      "epoch": 0.8507126781695424,
      "grad_norm": 0.07992705702781677,
      "learning_rate": 7.108969433270968e-05,
      "loss": 0.0226,
      "step": 1134
    },
    {
      "epoch": 0.8514628657164292,
      "grad_norm": 0.11864115297794342,
      "learning_rate": 7.10303086939921e-05,
      "loss": 0.0299,
      "step": 1135
    },
    {
      "epoch": 0.8522130532633159,
      "grad_norm": 0.107135109603405,
      "learning_rate": 7.097088699036598e-05,
      "loss": 0.041,
      "step": 1136
    },
    {
      "epoch": 0.8529632408102026,
      "grad_norm": 0.12338218092918396,
      "learning_rate": 7.091142932373371e-05,
      "loss": 0.0331,
      "step": 1137
    },
    {
      "epoch": 0.8537134283570893,
      "grad_norm": 0.24427056312561035,
      "learning_rate": 7.085193579605935e-05,
      "loss": 0.0508,
      "step": 1138
    },
    {
      "epoch": 0.854463615903976,
      "grad_norm": 0.08631362020969391,
      "learning_rate": 7.079240650936843e-05,
      "loss": 0.0284,
      "step": 1139
    },
    {
      "epoch": 0.8552138034508627,
      "grad_norm": 0.12822337448596954,
      "learning_rate": 7.07328415657478e-05,
      "loss": 0.0295,
      "step": 1140
    },
    {
      "epoch": 0.8559639909977494,
      "grad_norm": 0.13125036656856537,
      "learning_rate": 7.067324106734548e-05,
      "loss": 0.029,
      "step": 1141
    },
    {
      "epoch": 0.8567141785446362,
      "grad_norm": 0.1482788324356079,
      "learning_rate": 7.061360511637045e-05,
      "loss": 0.0271,
      "step": 1142
    },
    {
      "epoch": 0.8574643660915229,
      "grad_norm": 0.7413152456283569,
      "learning_rate": 7.055393381509253e-05,
      "loss": 0.0391,
      "step": 1143
    },
    {
      "epoch": 0.8582145536384096,
      "grad_norm": 0.10552649945020676,
      "learning_rate": 7.049422726584206e-05,
      "loss": 0.0316,
      "step": 1144
    },
    {
      "epoch": 0.8589647411852963,
      "grad_norm": 0.12541364133358002,
      "learning_rate": 7.043448557100995e-05,
      "loss": 0.0448,
      "step": 1145
    },
    {
      "epoch": 0.859714928732183,
      "grad_norm": 0.0947171300649643,
      "learning_rate": 7.037470883304731e-05,
      "loss": 0.0212,
      "step": 1146
    },
    {
      "epoch": 0.8604651162790697,
      "grad_norm": 0.11750858277082443,
      "learning_rate": 7.031489715446535e-05,
      "loss": 0.0274,
      "step": 1147
    },
    {
      "epoch": 0.8612153038259565,
      "grad_norm": 0.17968003451824188,
      "learning_rate": 7.02550506378352e-05,
      "loss": 0.0411,
      "step": 1148
    },
    {
      "epoch": 0.8619654913728432,
      "grad_norm": 0.1670011729001999,
      "learning_rate": 7.019516938578777e-05,
      "loss": 0.0287,
      "step": 1149
    },
    {
      "epoch": 0.8627156789197299,
      "grad_norm": 0.12383407354354858,
      "learning_rate": 7.013525350101348e-05,
      "loss": 0.028,
      "step": 1150
    },
    {
      "epoch": 0.8634658664666166,
      "grad_norm": 0.13687686622142792,
      "learning_rate": 7.00753030862622e-05,
      "loss": 0.0394,
      "step": 1151
    },
    {
      "epoch": 0.8642160540135033,
      "grad_norm": 0.1233229711651802,
      "learning_rate": 7.001531824434299e-05,
      "loss": 0.0326,
      "step": 1152
    },
    {
      "epoch": 0.8649662415603901,
      "grad_norm": 0.13290193676948547,
      "learning_rate": 6.995529907812391e-05,
      "loss": 0.039,
      "step": 1153
    },
    {
      "epoch": 0.8657164291072769,
      "grad_norm": 0.09817846119403839,
      "learning_rate": 6.989524569053196e-05,
      "loss": 0.0312,
      "step": 1154
    },
    {
      "epoch": 0.8664666166541636,
      "grad_norm": 0.10229942947626114,
      "learning_rate": 6.983515818455275e-05,
      "loss": 0.0261,
      "step": 1155
    },
    {
      "epoch": 0.8672168042010503,
      "grad_norm": 0.1152266263961792,
      "learning_rate": 6.977503666323048e-05,
      "loss": 0.0427,
      "step": 1156
    },
    {
      "epoch": 0.867966991747937,
      "grad_norm": 0.14483705163002014,
      "learning_rate": 6.971488122966758e-05,
      "loss": 0.0435,
      "step": 1157
    },
    {
      "epoch": 0.8687171792948237,
      "grad_norm": 0.1095462441444397,
      "learning_rate": 6.965469198702475e-05,
      "loss": 0.0274,
      "step": 1158
    },
    {
      "epoch": 0.8694673668417104,
      "grad_norm": 0.11495321244001389,
      "learning_rate": 6.95944690385206e-05,
      "loss": 0.0265,
      "step": 1159
    },
    {
      "epoch": 0.8702175543885972,
      "grad_norm": 0.1372273713350296,
      "learning_rate": 6.953421248743154e-05,
      "loss": 0.0423,
      "step": 1160
    },
    {
      "epoch": 0.8709677419354839,
      "grad_norm": 0.11597606539726257,
      "learning_rate": 6.947392243709163e-05,
      "loss": 0.0286,
      "step": 1161
    },
    {
      "epoch": 0.8717179294823706,
      "grad_norm": 0.13315455615520477,
      "learning_rate": 6.941359899089238e-05,
      "loss": 0.0301,
      "step": 1162
    },
    {
      "epoch": 0.8724681170292573,
      "grad_norm": 0.1332310438156128,
      "learning_rate": 6.935324225228254e-05,
      "loss": 0.0325,
      "step": 1163
    },
    {
      "epoch": 0.873218304576144,
      "grad_norm": 0.13525554537773132,
      "learning_rate": 6.929285232476797e-05,
      "loss": 0.0379,
      "step": 1164
    },
    {
      "epoch": 0.8739684921230307,
      "grad_norm": 0.09915309399366379,
      "learning_rate": 6.923242931191148e-05,
      "loss": 0.0292,
      "step": 1165
    },
    {
      "epoch": 0.8747186796699175,
      "grad_norm": 0.15619905292987823,
      "learning_rate": 6.917197331733257e-05,
      "loss": 0.0391,
      "step": 1166
    },
    {
      "epoch": 0.8754688672168042,
      "grad_norm": 0.13476809859275818,
      "learning_rate": 6.911148444470731e-05,
      "loss": 0.0227,
      "step": 1167
    },
    {
      "epoch": 0.8762190547636909,
      "grad_norm": 0.12116993218660355,
      "learning_rate": 6.905096279776819e-05,
      "loss": 0.0332,
      "step": 1168
    },
    {
      "epoch": 0.8769692423105776,
      "grad_norm": 0.07770241796970367,
      "learning_rate": 6.899040848030384e-05,
      "loss": 0.0199,
      "step": 1169
    },
    {
      "epoch": 0.8777194298574643,
      "grad_norm": 0.09852062910795212,
      "learning_rate": 6.892982159615895e-05,
      "loss": 0.0329,
      "step": 1170
    },
    {
      "epoch": 0.878469617404351,
      "grad_norm": 0.16133050620555878,
      "learning_rate": 6.886920224923408e-05,
      "loss": 0.0374,
      "step": 1171
    },
    {
      "epoch": 0.8792198049512379,
      "grad_norm": 0.11031442135572433,
      "learning_rate": 6.880855054348543e-05,
      "loss": 0.0286,
      "step": 1172
    },
    {
      "epoch": 0.8799699924981246,
      "grad_norm": 0.13426519930362701,
      "learning_rate": 6.874786658292472e-05,
      "loss": 0.0315,
      "step": 1173
    },
    {
      "epoch": 0.8807201800450113,
      "grad_norm": 0.14885659515857697,
      "learning_rate": 6.868715047161896e-05,
      "loss": 0.0389,
      "step": 1174
    },
    {
      "epoch": 0.881470367591898,
      "grad_norm": 0.17592477798461914,
      "learning_rate": 6.862640231369029e-05,
      "loss": 0.0367,
      "step": 1175
    },
    {
      "epoch": 0.8822205551387847,
      "grad_norm": 0.11507947742938995,
      "learning_rate": 6.856562221331581e-05,
      "loss": 0.0276,
      "step": 1176
    },
    {
      "epoch": 0.8829707426856714,
      "grad_norm": 0.10695502161979675,
      "learning_rate": 6.850481027472743e-05,
      "loss": 0.0297,
      "step": 1177
    },
    {
      "epoch": 0.8837209302325582,
      "grad_norm": 0.08131774514913559,
      "learning_rate": 6.844396660221164e-05,
      "loss": 0.0195,
      "step": 1178
    },
    {
      "epoch": 0.8844711177794449,
      "grad_norm": 0.11771385371685028,
      "learning_rate": 6.838309130010933e-05,
      "loss": 0.0236,
      "step": 1179
    },
    {
      "epoch": 0.8852213053263316,
      "grad_norm": 0.09990225732326508,
      "learning_rate": 6.832218447281566e-05,
      "loss": 0.0249,
      "step": 1180
    },
    {
      "epoch": 0.8859714928732183,
      "grad_norm": 0.12525203824043274,
      "learning_rate": 6.826124622477981e-05,
      "loss": 0.0318,
      "step": 1181
    },
    {
      "epoch": 0.886721680420105,
      "grad_norm": 0.19846653938293457,
      "learning_rate": 6.820027666050493e-05,
      "loss": 0.0335,
      "step": 1182
    },
    {
      "epoch": 0.8874718679669917,
      "grad_norm": 0.17230094969272614,
      "learning_rate": 6.813927588454778e-05,
      "loss": 0.0324,
      "step": 1183
    },
    {
      "epoch": 0.8882220555138785,
      "grad_norm": 0.11220695078372955,
      "learning_rate": 6.80782440015187e-05,
      "loss": 0.0291,
      "step": 1184
    },
    {
      "epoch": 0.8889722430607652,
      "grad_norm": 0.1424178034067154,
      "learning_rate": 6.801718111608133e-05,
      "loss": 0.0398,
      "step": 1185
    },
    {
      "epoch": 0.8897224306076519,
      "grad_norm": 0.1714315265417099,
      "learning_rate": 6.795608733295254e-05,
      "loss": 0.0308,
      "step": 1186
    },
    {
      "epoch": 0.8904726181545386,
      "grad_norm": 0.1779463142156601,
      "learning_rate": 6.789496275690215e-05,
      "loss": 0.0317,
      "step": 1187
    },
    {
      "epoch": 0.8912228057014253,
      "grad_norm": 0.12434078007936478,
      "learning_rate": 6.783380749275277e-05,
      "loss": 0.0288,
      "step": 1188
    },
    {
      "epoch": 0.891972993248312,
      "grad_norm": 0.09934748709201813,
      "learning_rate": 6.777262164537966e-05,
      "loss": 0.0287,
      "step": 1189
    },
    {
      "epoch": 0.8927231807951987,
      "grad_norm": 0.09271322935819626,
      "learning_rate": 6.771140531971054e-05,
      "loss": 0.0217,
      "step": 1190
    },
    {
      "epoch": 0.8934733683420856,
      "grad_norm": 0.1522030234336853,
      "learning_rate": 6.765015862072536e-05,
      "loss": 0.0347,
      "step": 1191
    },
    {
      "epoch": 0.8942235558889723,
      "grad_norm": 0.07218828052282333,
      "learning_rate": 6.758888165345619e-05,
      "loss": 0.0219,
      "step": 1192
    },
    {
      "epoch": 0.894973743435859,
      "grad_norm": 0.1766599416732788,
      "learning_rate": 6.7527574522987e-05,
      "loss": 0.0377,
      "step": 1193
    },
    {
      "epoch": 0.8957239309827457,
      "grad_norm": 0.09117019921541214,
      "learning_rate": 6.746623733445346e-05,
      "loss": 0.0208,
      "step": 1194
    },
    {
      "epoch": 0.8964741185296324,
      "grad_norm": 0.10044237971305847,
      "learning_rate": 6.740487019304282e-05,
      "loss": 0.0222,
      "step": 1195
    },
    {
      "epoch": 0.8972243060765192,
      "grad_norm": 0.11803122609853745,
      "learning_rate": 6.734347320399369e-05,
      "loss": 0.0208,
      "step": 1196
    },
    {
      "epoch": 0.8979744936234059,
      "grad_norm": 0.11625832319259644,
      "learning_rate": 6.728204647259586e-05,
      "loss": 0.0434,
      "step": 1197
    },
    {
      "epoch": 0.8987246811702926,
      "grad_norm": 0.07271845638751984,
      "learning_rate": 6.722059010419013e-05,
      "loss": 0.0222,
      "step": 1198
    },
    {
      "epoch": 0.8994748687171793,
      "grad_norm": 0.15432986617088318,
      "learning_rate": 6.715910420416809e-05,
      "loss": 0.0373,
      "step": 1199
    },
    {
      "epoch": 0.900225056264066,
      "grad_norm": 0.09750841557979584,
      "learning_rate": 6.709758887797205e-05,
      "loss": 0.0258,
      "step": 1200
    },
    {
      "epoch": 0.900225056264066,
      "eval_loss": 0.03287737816572189,
      "eval_runtime": 5.1474,
      "eval_samples_per_second": 10.491,
      "eval_steps_per_second": 2.72,
      "step": 1200
    },
    {
      "epoch": 0.9009752438109527,
      "grad_norm": 0.12649789452552795,
      "learning_rate": 6.703604423109468e-05,
      "loss": 0.0351,
      "step": 1201
    },
    {
      "epoch": 0.9017254313578394,
      "grad_norm": 0.11160416901111603,
      "learning_rate": 6.697447036907904e-05,
      "loss": 0.0233,
      "step": 1202
    },
    {
      "epoch": 0.9024756189047262,
      "grad_norm": 0.055738791823387146,
      "learning_rate": 6.691286739751824e-05,
      "loss": 0.0164,
      "step": 1203
    },
    {
      "epoch": 0.9032258064516129,
      "grad_norm": 0.15643253922462463,
      "learning_rate": 6.685123542205526e-05,
      "loss": 0.0255,
      "step": 1204
    },
    {
      "epoch": 0.9039759939984996,
      "grad_norm": 0.17560407519340515,
      "learning_rate": 6.678957454838292e-05,
      "loss": 0.0479,
      "step": 1205
    },
    {
      "epoch": 0.9047261815453863,
      "grad_norm": 0.12479657679796219,
      "learning_rate": 6.672788488224352e-05,
      "loss": 0.0309,
      "step": 1206
    },
    {
      "epoch": 0.905476369092273,
      "grad_norm": 0.11823344230651855,
      "learning_rate": 6.666616652942878e-05,
      "loss": 0.0282,
      "step": 1207
    },
    {
      "epoch": 0.9062265566391597,
      "grad_norm": 0.09630381315946579,
      "learning_rate": 6.660441959577958e-05,
      "loss": 0.0258,
      "step": 1208
    },
    {
      "epoch": 0.9069767441860465,
      "grad_norm": 0.14388345181941986,
      "learning_rate": 6.654264418718584e-05,
      "loss": 0.0392,
      "step": 1209
    },
    {
      "epoch": 0.9077269317329333,
      "grad_norm": 0.11177119612693787,
      "learning_rate": 6.64808404095863e-05,
      "loss": 0.0259,
      "step": 1210
    },
    {
      "epoch": 0.90847711927982,
      "grad_norm": 0.1320035457611084,
      "learning_rate": 6.641900836896835e-05,
      "loss": 0.0302,
      "step": 1211
    },
    {
      "epoch": 0.9092273068267067,
      "grad_norm": 0.10894173383712769,
      "learning_rate": 6.635714817136785e-05,
      "loss": 0.0222,
      "step": 1212
    },
    {
      "epoch": 0.9099774943735934,
      "grad_norm": 0.08480621874332428,
      "learning_rate": 6.629525992286898e-05,
      "loss": 0.02,
      "step": 1213
    },
    {
      "epoch": 0.9107276819204801,
      "grad_norm": 0.11812885105609894,
      "learning_rate": 6.623334372960393e-05,
      "loss": 0.031,
      "step": 1214
    },
    {
      "epoch": 0.9114778694673669,
      "grad_norm": 0.15104812383651733,
      "learning_rate": 6.617139969775292e-05,
      "loss": 0.0399,
      "step": 1215
    },
    {
      "epoch": 0.9122280570142536,
      "grad_norm": 0.1338520050048828,
      "learning_rate": 6.610942793354387e-05,
      "loss": 0.0343,
      "step": 1216
    },
    {
      "epoch": 0.9129782445611403,
      "grad_norm": 0.10010179877281189,
      "learning_rate": 6.604742854325222e-05,
      "loss": 0.027,
      "step": 1217
    },
    {
      "epoch": 0.913728432108027,
      "grad_norm": 0.1490030139684677,
      "learning_rate": 6.598540163320084e-05,
      "loss": 0.0243,
      "step": 1218
    },
    {
      "epoch": 0.9144786196549137,
      "grad_norm": 0.13927358388900757,
      "learning_rate": 6.592334730975975e-05,
      "loss": 0.038,
      "step": 1219
    },
    {
      "epoch": 0.9152288072018004,
      "grad_norm": 0.1642918735742569,
      "learning_rate": 6.586126567934605e-05,
      "loss": 0.0475,
      "step": 1220
    },
    {
      "epoch": 0.9159789947486872,
      "grad_norm": 0.12470587342977524,
      "learning_rate": 6.57991568484236e-05,
      "loss": 0.0282,
      "step": 1221
    },
    {
      "epoch": 0.9167291822955739,
      "grad_norm": 0.11485706269741058,
      "learning_rate": 6.573702092350292e-05,
      "loss": 0.0327,
      "step": 1222
    },
    {
      "epoch": 0.9174793698424606,
      "grad_norm": 0.15273365378379822,
      "learning_rate": 6.567485801114099e-05,
      "loss": 0.0312,
      "step": 1223
    },
    {
      "epoch": 0.9182295573893473,
      "grad_norm": 0.0842658132314682,
      "learning_rate": 6.561266821794111e-05,
      "loss": 0.0205,
      "step": 1224
    },
    {
      "epoch": 0.918979744936234,
      "grad_norm": 0.1895236372947693,
      "learning_rate": 6.555045165055263e-05,
      "loss": 0.0356,
      "step": 1225
    },
    {
      "epoch": 0.9197299324831207,
      "grad_norm": 0.13262400031089783,
      "learning_rate": 6.548820841567086e-05,
      "loss": 0.0285,
      "step": 1226
    },
    {
      "epoch": 0.9204801200300075,
      "grad_norm": 0.10285253077745438,
      "learning_rate": 6.54259386200368e-05,
      "loss": 0.0256,
      "step": 1227
    },
    {
      "epoch": 0.9212303075768942,
      "grad_norm": 0.11173094063997269,
      "learning_rate": 6.536364237043703e-05,
      "loss": 0.0263,
      "step": 1228
    },
    {
      "epoch": 0.921980495123781,
      "grad_norm": 0.1017502024769783,
      "learning_rate": 6.530131977370348e-05,
      "loss": 0.025,
      "step": 1229
    },
    {
      "epoch": 0.9227306826706677,
      "grad_norm": 0.11662982404232025,
      "learning_rate": 6.523897093671326e-05,
      "loss": 0.0265,
      "step": 1230
    },
    {
      "epoch": 0.9234808702175544,
      "grad_norm": 0.10500746220350266,
      "learning_rate": 6.51765959663885e-05,
      "loss": 0.0287,
      "step": 1231
    },
    {
      "epoch": 0.9242310577644411,
      "grad_norm": 0.17836572229862213,
      "learning_rate": 6.511419496969612e-05,
      "loss": 0.042,
      "step": 1232
    },
    {
      "epoch": 0.9249812453113279,
      "grad_norm": 0.1660510152578354,
      "learning_rate": 6.505176805364767e-05,
      "loss": 0.0501,
      "step": 1233
    },
    {
      "epoch": 0.9257314328582146,
      "grad_norm": 0.08452211320400238,
      "learning_rate": 6.498931532529921e-05,
      "loss": 0.0257,
      "step": 1234
    },
    {
      "epoch": 0.9264816204051013,
      "grad_norm": 0.1757364720106125,
      "learning_rate": 6.492683689175098e-05,
      "loss": 0.0301,
      "step": 1235
    },
    {
      "epoch": 0.927231807951988,
      "grad_norm": 0.1669311672449112,
      "learning_rate": 6.486433286014734e-05,
      "loss": 0.0374,
      "step": 1236
    },
    {
      "epoch": 0.9279819954988747,
      "grad_norm": 0.1517716944217682,
      "learning_rate": 6.480180333767658e-05,
      "loss": 0.0473,
      "step": 1237
    },
    {
      "epoch": 0.9287321830457614,
      "grad_norm": 0.11780434101819992,
      "learning_rate": 6.473924843157065e-05,
      "loss": 0.0236,
      "step": 1238
    },
    {
      "epoch": 0.9294823705926482,
      "grad_norm": 0.08752031624317169,
      "learning_rate": 6.467666824910505e-05,
      "loss": 0.026,
      "step": 1239
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 0.11586204916238785,
      "learning_rate": 6.461406289759862e-05,
      "loss": 0.0274,
      "step": 1240
    },
    {
      "epoch": 0.9309827456864216,
      "grad_norm": 0.10578847676515579,
      "learning_rate": 6.455143248441342e-05,
      "loss": 0.0289,
      "step": 1241
    },
    {
      "epoch": 0.9317329332333083,
      "grad_norm": 0.12548696994781494,
      "learning_rate": 6.44887771169544e-05,
      "loss": 0.0234,
      "step": 1242
    },
    {
      "epoch": 0.932483120780195,
      "grad_norm": 0.11294113844633102,
      "learning_rate": 6.442609690266937e-05,
      "loss": 0.0391,
      "step": 1243
    },
    {
      "epoch": 0.9332333083270817,
      "grad_norm": 0.10007303953170776,
      "learning_rate": 6.436339194904872e-05,
      "loss": 0.0253,
      "step": 1244
    },
    {
      "epoch": 0.9339834958739685,
      "grad_norm": 0.10004784166812897,
      "learning_rate": 6.430066236362524e-05,
      "loss": 0.0321,
      "step": 1245
    },
    {
      "epoch": 0.9347336834208552,
      "grad_norm": 0.15509480237960815,
      "learning_rate": 6.423790825397404e-05,
      "loss": 0.0314,
      "step": 1246
    },
    {
      "epoch": 0.9354838709677419,
      "grad_norm": 0.1072937548160553,
      "learning_rate": 6.417512972771219e-05,
      "loss": 0.0278,
      "step": 1247
    },
    {
      "epoch": 0.9362340585146287,
      "grad_norm": 0.2849551737308502,
      "learning_rate": 6.411232689249873e-05,
      "loss": 0.0447,
      "step": 1248
    },
    {
      "epoch": 0.9369842460615154,
      "grad_norm": 0.10778176784515381,
      "learning_rate": 6.40494998560343e-05,
      "loss": 0.0247,
      "step": 1249
    },
    {
      "epoch": 0.9377344336084021,
      "grad_norm": 0.15809431672096252,
      "learning_rate": 6.39866487260611e-05,
      "loss": 0.0276,
      "step": 1250
    },
    {
      "epoch": 0.9384846211552889,
      "grad_norm": 0.08884350955486298,
      "learning_rate": 6.392377361036262e-05,
      "loss": 0.0189,
      "step": 1251
    },
    {
      "epoch": 0.9392348087021756,
      "grad_norm": 0.07642669975757599,
      "learning_rate": 6.386087461676351e-05,
      "loss": 0.0201,
      "step": 1252
    },
    {
      "epoch": 0.9399849962490623,
      "grad_norm": 0.10072652995586395,
      "learning_rate": 6.379795185312933e-05,
      "loss": 0.0202,
      "step": 1253
    },
    {
      "epoch": 0.940735183795949,
      "grad_norm": 0.132140651345253,
      "learning_rate": 6.373500542736643e-05,
      "loss": 0.029,
      "step": 1254
    },
    {
      "epoch": 0.9414853713428357,
      "grad_norm": 0.09724639356136322,
      "learning_rate": 6.367203544742171e-05,
      "loss": 0.0252,
      "step": 1255
    },
    {
      "epoch": 0.9422355588897224,
      "grad_norm": 0.20375050604343414,
      "learning_rate": 6.360904202128252e-05,
      "loss": 0.0373,
      "step": 1256
    },
    {
      "epoch": 0.9429857464366092,
      "grad_norm": 0.14257559180259705,
      "learning_rate": 6.354602525697638e-05,
      "loss": 0.033,
      "step": 1257
    },
    {
      "epoch": 0.9437359339834959,
      "grad_norm": 0.12351656705141068,
      "learning_rate": 6.348298526257082e-05,
      "loss": 0.0328,
      "step": 1258
    },
    {
      "epoch": 0.9444861215303826,
      "grad_norm": 0.12928931415081024,
      "learning_rate": 6.341992214617323e-05,
      "loss": 0.0264,
      "step": 1259
    },
    {
      "epoch": 0.9452363090772693,
      "grad_norm": 0.13477769494056702,
      "learning_rate": 6.335683601593062e-05,
      "loss": 0.0402,
      "step": 1260
    },
    {
      "epoch": 0.945986496624156,
      "grad_norm": 0.1282130926847458,
      "learning_rate": 6.329372698002954e-05,
      "loss": 0.0261,
      "step": 1261
    },
    {
      "epoch": 0.9467366841710427,
      "grad_norm": 0.3520214259624481,
      "learning_rate": 6.323059514669571e-05,
      "loss": 0.0611,
      "step": 1262
    },
    {
      "epoch": 0.9474868717179294,
      "grad_norm": 0.1838792860507965,
      "learning_rate": 6.316744062419409e-05,
      "loss": 0.0302,
      "step": 1263
    },
    {
      "epoch": 0.9482370592648162,
      "grad_norm": 0.21650949120521545,
      "learning_rate": 6.310426352082838e-05,
      "loss": 0.0434,
      "step": 1264
    },
    {
      "epoch": 0.9489872468117029,
      "grad_norm": 0.10584396123886108,
      "learning_rate": 6.304106394494116e-05,
      "loss": 0.0273,
      "step": 1265
    },
    {
      "epoch": 0.9497374343585896,
      "grad_norm": 0.16731998324394226,
      "learning_rate": 6.297784200491343e-05,
      "loss": 0.0417,
      "step": 1266
    },
    {
      "epoch": 0.9504876219054764,
      "grad_norm": 0.10184057801961899,
      "learning_rate": 6.291459780916463e-05,
      "loss": 0.0207,
      "step": 1267
    },
    {
      "epoch": 0.9512378094523631,
      "grad_norm": 0.14981421828269958,
      "learning_rate": 6.285133146615228e-05,
      "loss": 0.0292,
      "step": 1268
    },
    {
      "epoch": 0.9519879969992499,
      "grad_norm": 0.12305141985416412,
      "learning_rate": 6.278804308437198e-05,
      "loss": 0.0255,
      "step": 1269
    },
    {
      "epoch": 0.9527381845461366,
      "grad_norm": 0.1167244091629982,
      "learning_rate": 6.272473277235703e-05,
      "loss": 0.03,
      "step": 1270
    },
    {
      "epoch": 0.9534883720930233,
      "grad_norm": 0.12276820093393326,
      "learning_rate": 6.266140063867843e-05,
      "loss": 0.0307,
      "step": 1271
    },
    {
      "epoch": 0.95423855963991,
      "grad_norm": 0.16493280231952667,
      "learning_rate": 6.25980467919445e-05,
      "loss": 0.039,
      "step": 1272
    },
    {
      "epoch": 0.9549887471867967,
      "grad_norm": 0.11039330810308456,
      "learning_rate": 6.253467134080088e-05,
      "loss": 0.0265,
      "step": 1273
    },
    {
      "epoch": 0.9557389347336834,
      "grad_norm": 0.08698292821645737,
      "learning_rate": 6.247127439393023e-05,
      "loss": 0.0236,
      "step": 1274
    },
    {
      "epoch": 0.9564891222805701,
      "grad_norm": 0.13140971958637238,
      "learning_rate": 6.240785606005206e-05,
      "loss": 0.0295,
      "step": 1275
    },
    {
      "epoch": 0.9572393098274569,
      "grad_norm": 0.21670088171958923,
      "learning_rate": 6.234441644792256e-05,
      "loss": 0.0256,
      "step": 1276
    },
    {
      "epoch": 0.9579894973743436,
      "grad_norm": 0.1584072709083557,
      "learning_rate": 6.228095566633443e-05,
      "loss": 0.0393,
      "step": 1277
    },
    {
      "epoch": 0.9587396849212303,
      "grad_norm": 0.07527758181095123,
      "learning_rate": 6.221747382411667e-05,
      "loss": 0.0181,
      "step": 1278
    },
    {
      "epoch": 0.959489872468117,
      "grad_norm": 0.121820367872715,
      "learning_rate": 6.215397103013436e-05,
      "loss": 0.0328,
      "step": 1279
    },
    {
      "epoch": 0.9602400600150037,
      "grad_norm": 0.09028357267379761,
      "learning_rate": 6.209044739328858e-05,
      "loss": 0.0247,
      "step": 1280
    },
    {
      "epoch": 0.9609902475618904,
      "grad_norm": 0.13452784717082977,
      "learning_rate": 6.202690302251606e-05,
      "loss": 0.0375,
      "step": 1281
    },
    {
      "epoch": 0.9617404351087772,
      "grad_norm": 0.10062450915575027,
      "learning_rate": 6.196333802678918e-05,
      "loss": 0.029,
      "step": 1282
    },
    {
      "epoch": 0.9624906226556639,
      "grad_norm": 0.2688952088356018,
      "learning_rate": 6.189975251511562e-05,
      "loss": 0.0416,
      "step": 1283
    },
    {
      "epoch": 0.9632408102025506,
      "grad_norm": 0.1521732211112976,
      "learning_rate": 6.18361465965383e-05,
      "loss": 0.0431,
      "step": 1284
    },
    {
      "epoch": 0.9639909977494373,
      "grad_norm": 0.11925928294658661,
      "learning_rate": 6.177252038013509e-05,
      "loss": 0.0282,
      "step": 1285
    },
    {
      "epoch": 0.9647411852963241,
      "grad_norm": 0.07919792085886002,
      "learning_rate": 6.170887397501868e-05,
      "loss": 0.0203,
      "step": 1286
    },
    {
      "epoch": 0.9654913728432108,
      "grad_norm": 0.16996318101882935,
      "learning_rate": 6.16452074903364e-05,
      "loss": 0.0437,
      "step": 1287
    },
    {
      "epoch": 0.9662415603900976,
      "grad_norm": 0.17046809196472168,
      "learning_rate": 6.158152103527e-05,
      "loss": 0.046,
      "step": 1288
    },
    {
      "epoch": 0.9669917479369843,
      "grad_norm": 0.14924894273281097,
      "learning_rate": 6.151781471903548e-05,
      "loss": 0.0272,
      "step": 1289
    },
    {
      "epoch": 0.967741935483871,
      "grad_norm": 0.16980625689029694,
      "learning_rate": 6.14540886508829e-05,
      "loss": 0.051,
      "step": 1290
    },
    {
      "epoch": 0.9684921230307577,
      "grad_norm": 0.12601691484451294,
      "learning_rate": 6.139034294009617e-05,
      "loss": 0.0333,
      "step": 1291
    },
    {
      "epoch": 0.9692423105776444,
      "grad_norm": 0.10268208384513855,
      "learning_rate": 6.132657769599293e-05,
      "loss": 0.0283,
      "step": 1292
    },
    {
      "epoch": 0.9699924981245311,
      "grad_norm": 0.16088128089904785,
      "learning_rate": 6.126279302792429e-05,
      "loss": 0.0515,
      "step": 1293
    },
    {
      "epoch": 0.9707426856714179,
      "grad_norm": 0.07889743149280548,
      "learning_rate": 6.119898904527468e-05,
      "loss": 0.0165,
      "step": 1294
    },
    {
      "epoch": 0.9714928732183046,
      "grad_norm": 0.11629171669483185,
      "learning_rate": 6.113516585746164e-05,
      "loss": 0.0484,
      "step": 1295
    },
    {
      "epoch": 0.9722430607651913,
      "grad_norm": 0.11696446686983109,
      "learning_rate": 6.107132357393563e-05,
      "loss": 0.0304,
      "step": 1296
    },
    {
      "epoch": 0.972993248312078,
      "grad_norm": 0.1155184879899025,
      "learning_rate": 6.100746230417993e-05,
      "loss": 0.0218,
      "step": 1297
    },
    {
      "epoch": 0.9737434358589647,
      "grad_norm": 0.13009968400001526,
      "learning_rate": 6.0943582157710285e-05,
      "loss": 0.0274,
      "step": 1298
    },
    {
      "epoch": 0.9744936234058514,
      "grad_norm": 0.11266110092401505,
      "learning_rate": 6.0879683244074894e-05,
      "loss": 0.0321,
      "step": 1299
    },
    {
      "epoch": 0.9752438109527382,
      "grad_norm": 0.0928129181265831,
      "learning_rate": 6.0815765672854065e-05,
      "loss": 0.0265,
      "step": 1300
    },
    {
      "epoch": 0.9759939984996249,
      "grad_norm": 0.10671856254339218,
      "learning_rate": 6.0751829553660155e-05,
      "loss": 0.0249,
      "step": 1301
    },
    {
      "epoch": 0.9767441860465116,
      "grad_norm": 0.09111996740102768,
      "learning_rate": 6.06878749961373e-05,
      "loss": 0.0299,
      "step": 1302
    },
    {
      "epoch": 0.9774943735933983,
      "grad_norm": 0.19780229032039642,
      "learning_rate": 6.0623902109961295e-05,
      "loss": 0.0385,
      "step": 1303
    },
    {
      "epoch": 0.978244561140285,
      "grad_norm": 0.13554957509040833,
      "learning_rate": 6.055991100483932e-05,
      "loss": 0.0384,
      "step": 1304
    },
    {
      "epoch": 0.9789947486871718,
      "grad_norm": 0.13489709794521332,
      "learning_rate": 6.0495901790509836e-05,
      "loss": 0.0292,
      "step": 1305
    },
    {
      "epoch": 0.9797449362340586,
      "grad_norm": 0.11630948632955551,
      "learning_rate": 6.043187457674231e-05,
      "loss": 0.025,
      "step": 1306
    },
    {
      "epoch": 0.9804951237809453,
      "grad_norm": 0.14980089664459229,
      "learning_rate": 6.0367829473337136e-05,
      "loss": 0.0324,
      "step": 1307
    },
    {
      "epoch": 0.981245311327832,
      "grad_norm": 0.09936217218637466,
      "learning_rate": 6.0303766590125365e-05,
      "loss": 0.0231,
      "step": 1308
    },
    {
      "epoch": 0.9819954988747187,
      "grad_norm": 0.10359448194503784,
      "learning_rate": 6.02396860369685e-05,
      "loss": 0.0348,
      "step": 1309
    },
    {
      "epoch": 0.9827456864216054,
      "grad_norm": 0.17159534990787506,
      "learning_rate": 6.0175587923758416e-05,
      "loss": 0.0413,
      "step": 1310
    },
    {
      "epoch": 0.9834958739684921,
      "grad_norm": 0.15349924564361572,
      "learning_rate": 6.0111472360417044e-05,
      "loss": 0.0321,
      "step": 1311
    },
    {
      "epoch": 0.9842460615153789,
      "grad_norm": 0.12869088351726532,
      "learning_rate": 6.004733945689628e-05,
      "loss": 0.0318,
      "step": 1312
    },
    {
      "epoch": 0.9849962490622656,
      "grad_norm": 0.1338404268026352,
      "learning_rate": 5.998318932317771e-05,
      "loss": 0.0393,
      "step": 1313
    },
    {
      "epoch": 0.9857464366091523,
      "grad_norm": 0.12612248957157135,
      "learning_rate": 5.991902206927252e-05,
      "loss": 0.0264,
      "step": 1314
    },
    {
      "epoch": 0.986496624156039,
      "grad_norm": 0.081502266228199,
      "learning_rate": 5.985483780522122e-05,
      "loss": 0.0223,
      "step": 1315
    },
    {
      "epoch": 0.9872468117029257,
      "grad_norm": 0.14809326827526093,
      "learning_rate": 5.9790636641093523e-05,
      "loss": 0.0258,
      "step": 1316
    },
    {
      "epoch": 0.9879969992498124,
      "grad_norm": 0.12299007922410965,
      "learning_rate": 5.972641868698805e-05,
      "loss": 0.0258,
      "step": 1317
    },
    {
      "epoch": 0.9887471867966992,
      "grad_norm": 0.20244580507278442,
      "learning_rate": 5.966218405303234e-05,
      "loss": 0.0429,
      "step": 1318
    },
    {
      "epoch": 0.9894973743435859,
      "grad_norm": 0.09541887789964676,
      "learning_rate": 5.959793284938242e-05,
      "loss": 0.0285,
      "step": 1319
    },
    {
      "epoch": 0.9902475618904726,
      "grad_norm": 0.1416272521018982,
      "learning_rate": 5.953366518622279e-05,
      "loss": 0.0379,
      "step": 1320
    },
    {
      "epoch": 0.9909977494373593,
      "grad_norm": 0.10651285201311111,
      "learning_rate": 5.946938117376616e-05,
      "loss": 0.0292,
      "step": 1321
    },
    {
      "epoch": 0.991747936984246,
      "grad_norm": 0.12336979806423187,
      "learning_rate": 5.940508092225328e-05,
      "loss": 0.0266,
      "step": 1322
    },
    {
      "epoch": 0.9924981245311327,
      "grad_norm": 0.1195278912782669,
      "learning_rate": 5.9340764541952755e-05,
      "loss": 0.0298,
      "step": 1323
    },
    {
      "epoch": 0.9932483120780196,
      "grad_norm": 0.08796902745962143,
      "learning_rate": 5.9276432143160835e-05,
      "loss": 0.0184,
      "step": 1324
    },
    {
      "epoch": 0.9939984996249063,
      "grad_norm": 0.10521414875984192,
      "learning_rate": 5.921208383620126e-05,
      "loss": 0.0213,
      "step": 1325
    },
    {
      "epoch": 0.994748687171793,
      "grad_norm": 0.2021062821149826,
      "learning_rate": 5.9147719731425034e-05,
      "loss": 0.0414,
      "step": 1326
    },
    {
      "epoch": 0.9954988747186797,
      "grad_norm": 0.09227265417575836,
      "learning_rate": 5.908333993921027e-05,
      "loss": 0.0245,
      "step": 1327
    },
    {
      "epoch": 0.9962490622655664,
      "grad_norm": 0.09092900902032852,
      "learning_rate": 5.901894456996196e-05,
      "loss": 0.0254,
      "step": 1328
    },
    {
      "epoch": 0.9969992498124531,
      "grad_norm": 0.18778061866760254,
      "learning_rate": 5.895453373411182e-05,
      "loss": 0.0473,
      "step": 1329
    },
    {
      "epoch": 0.9977494373593399,
      "grad_norm": 0.20227041840553284,
      "learning_rate": 5.889010754211809e-05,
      "loss": 0.0473,
      "step": 1330
    },
    {
      "epoch": 0.9984996249062266,
      "grad_norm": 0.13871866464614868,
      "learning_rate": 5.882566610446534e-05,
      "loss": 0.0239,
      "step": 1331
    },
    {
      "epoch": 0.9992498124531133,
      "grad_norm": 0.11242667585611343,
      "learning_rate": 5.8761209531664306e-05,
      "loss": 0.0278,
      "step": 1332
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.16068680584430695,
      "learning_rate": 5.869673793425168e-05,
      "loss": 0.0362,
      "step": 1333
    },
    {
      "epoch": 1.0007501875468867,
      "grad_norm": 0.09198249876499176,
      "learning_rate": 5.863225142278985e-05,
      "loss": 0.027,
      "step": 1334
    },
    {
      "epoch": 1.0015003750937734,
      "grad_norm": 0.09522329270839691,
      "learning_rate": 5.856775010786687e-05,
      "loss": 0.0186,
      "step": 1335
    },
    {
      "epoch": 1.0022505626406601,
      "grad_norm": 0.10609661787748337,
      "learning_rate": 5.850323410009614e-05,
      "loss": 0.0189,
      "step": 1336
    },
    {
      "epoch": 1.0030007501875469,
      "grad_norm": 0.1119598001241684,
      "learning_rate": 5.8438703510116256e-05,
      "loss": 0.0216,
      "step": 1337
    },
    {
      "epoch": 1.0037509377344336,
      "grad_norm": 0.21976079046726227,
      "learning_rate": 5.8374158448590823e-05,
      "loss": 0.0358,
      "step": 1338
    },
    {
      "epoch": 1.0045011252813203,
      "grad_norm": 0.06545736640691757,
      "learning_rate": 5.830959902620826e-05,
      "loss": 0.0168,
      "step": 1339
    },
    {
      "epoch": 1.005251312828207,
      "grad_norm": 0.11048869788646698,
      "learning_rate": 5.824502535368164e-05,
      "loss": 0.0216,
      "step": 1340
    },
    {
      "epoch": 1.0060015003750937,
      "grad_norm": 0.0982142835855484,
      "learning_rate": 5.818043754174843e-05,
      "loss": 0.0216,
      "step": 1341
    },
    {
      "epoch": 1.0067516879219804,
      "grad_norm": 0.13148756325244904,
      "learning_rate": 5.81158357011704e-05,
      "loss": 0.0278,
      "step": 1342
    },
    {
      "epoch": 1.0075018754688672,
      "grad_norm": 0.05980212241411209,
      "learning_rate": 5.80512199427333e-05,
      "loss": 0.0121,
      "step": 1343
    },
    {
      "epoch": 1.0082520630157539,
      "grad_norm": 0.09239038079977036,
      "learning_rate": 5.798659037724683e-05,
      "loss": 0.0203,
      "step": 1344
    },
    {
      "epoch": 1.0090022505626406,
      "grad_norm": 0.09601394832134247,
      "learning_rate": 5.792194711554429e-05,
      "loss": 0.0205,
      "step": 1345
    },
    {
      "epoch": 1.0097524381095273,
      "grad_norm": 0.08131718635559082,
      "learning_rate": 5.7857290268482555e-05,
      "loss": 0.0216,
      "step": 1346
    },
    {
      "epoch": 1.010502625656414,
      "grad_norm": 0.06746546179056168,
      "learning_rate": 5.779261994694173e-05,
      "loss": 0.0093,
      "step": 1347
    },
    {
      "epoch": 1.0112528132033007,
      "grad_norm": 0.14873553812503815,
      "learning_rate": 5.772793626182506e-05,
      "loss": 0.0256,
      "step": 1348
    },
    {
      "epoch": 1.0120030007501875,
      "grad_norm": 0.09325051307678223,
      "learning_rate": 5.766323932405866e-05,
      "loss": 0.0187,
      "step": 1349
    },
    {
      "epoch": 1.0127531882970742,
      "grad_norm": 0.16514313220977783,
      "learning_rate": 5.7598529244591436e-05,
      "loss": 0.0274,
      "step": 1350
    },
    {
      "epoch": 1.0135033758439609,
      "grad_norm": 0.139047771692276,
      "learning_rate": 5.7533806134394806e-05,
      "loss": 0.0194,
      "step": 1351
    },
    {
      "epoch": 1.0142535633908478,
      "grad_norm": 0.09351517260074615,
      "learning_rate": 5.746907010446252e-05,
      "loss": 0.0155,
      "step": 1352
    },
    {
      "epoch": 1.0150037509377345,
      "grad_norm": 0.09682505577802658,
      "learning_rate": 5.740432126581049e-05,
      "loss": 0.0171,
      "step": 1353
    },
    {
      "epoch": 1.0157539384846213,
      "grad_norm": 0.14819428324699402,
      "learning_rate": 5.73395597294766e-05,
      "loss": 0.0103,
      "step": 1354
    },
    {
      "epoch": 1.016504126031508,
      "grad_norm": 0.10570473223924637,
      "learning_rate": 5.727478560652053e-05,
      "loss": 0.0219,
      "step": 1355
    },
    {
      "epoch": 1.0172543135783947,
      "grad_norm": 0.1157071441411972,
      "learning_rate": 5.7209999008023496e-05,
      "loss": 0.025,
      "step": 1356
    },
    {
      "epoch": 1.0180045011252814,
      "grad_norm": 0.15539030730724335,
      "learning_rate": 5.7145200045088156e-05,
      "loss": 0.0244,
      "step": 1357
    },
    {
      "epoch": 1.0187546886721681,
      "grad_norm": 0.06445297598838806,
      "learning_rate": 5.7080388828838324e-05,
      "loss": 0.0138,
      "step": 1358
    },
    {
      "epoch": 1.0195048762190548,
      "grad_norm": 0.14610378444194794,
      "learning_rate": 5.701556547041888e-05,
      "loss": 0.026,
      "step": 1359
    },
    {
      "epoch": 1.0202550637659416,
      "grad_norm": 0.1265845000743866,
      "learning_rate": 5.695073008099547e-05,
      "loss": 0.0182,
      "step": 1360
    },
    {
      "epoch": 1.0210052513128283,
      "grad_norm": 0.16218756139278412,
      "learning_rate": 5.688588277175444e-05,
      "loss": 0.0211,
      "step": 1361
    },
    {
      "epoch": 1.021755438859715,
      "grad_norm": 0.11030188202857971,
      "learning_rate": 5.6821023653902517e-05,
      "loss": 0.0159,
      "step": 1362
    },
    {
      "epoch": 1.0225056264066017,
      "grad_norm": 0.0763898640871048,
      "learning_rate": 5.675615283866671e-05,
      "loss": 0.013,
      "step": 1363
    },
    {
      "epoch": 1.0232558139534884,
      "grad_norm": 0.15313951671123505,
      "learning_rate": 5.669127043729406e-05,
      "loss": 0.0282,
      "step": 1364
    },
    {
      "epoch": 1.0240060015003751,
      "grad_norm": 0.07773459702730179,
      "learning_rate": 5.662637656105152e-05,
      "loss": 0.0177,
      "step": 1365
    },
    {
      "epoch": 1.0247561890472618,
      "grad_norm": 0.1399940848350525,
      "learning_rate": 5.6561471321225676e-05,
      "loss": 0.0252,
      "step": 1366
    },
    {
      "epoch": 1.0255063765941486,
      "grad_norm": 0.14890314638614655,
      "learning_rate": 5.649655482912265e-05,
      "loss": 0.0334,
      "step": 1367
    },
    {
      "epoch": 1.0262565641410353,
      "grad_norm": 0.10625971853733063,
      "learning_rate": 5.6431627196067816e-05,
      "loss": 0.0205,
      "step": 1368
    },
    {
      "epoch": 1.027006751687922,
      "grad_norm": 0.059316691011190414,
      "learning_rate": 5.636668853340567e-05,
      "loss": 0.0101,
      "step": 1369
    },
    {
      "epoch": 1.0277569392348087,
      "grad_norm": 0.05441593378782272,
      "learning_rate": 5.6301738952499636e-05,
      "loss": 0.0118,
      "step": 1370
    },
    {
      "epoch": 1.0285071267816954,
      "grad_norm": 0.12986020743846893,
      "learning_rate": 5.623677856473183e-05,
      "loss": 0.0306,
      "step": 1371
    },
    {
      "epoch": 1.0292573143285821,
      "grad_norm": 0.08668938279151917,
      "learning_rate": 5.617180748150295e-05,
      "loss": 0.017,
      "step": 1372
    },
    {
      "epoch": 1.0300075018754689,
      "grad_norm": 0.055281803011894226,
      "learning_rate": 5.6106825814231953e-05,
      "loss": 0.014,
      "step": 1373
    },
    {
      "epoch": 1.0307576894223556,
      "grad_norm": 0.14348354935646057,
      "learning_rate": 5.604183367435606e-05,
      "loss": 0.0191,
      "step": 1374
    },
    {
      "epoch": 1.0315078769692423,
      "grad_norm": 0.15625552833080292,
      "learning_rate": 5.597683117333036e-05,
      "loss": 0.0277,
      "step": 1375
    },
    {
      "epoch": 1.032258064516129,
      "grad_norm": 0.0986480638384819,
      "learning_rate": 5.591181842262776e-05,
      "loss": 0.0175,
      "step": 1376
    },
    {
      "epoch": 1.0330082520630157,
      "grad_norm": 0.16133823990821838,
      "learning_rate": 5.584679553373869e-05,
      "loss": 0.0175,
      "step": 1377
    },
    {
      "epoch": 1.0337584396099024,
      "grad_norm": 0.08799955993890762,
      "learning_rate": 5.578176261817104e-05,
      "loss": 0.0159,
      "step": 1378
    },
    {
      "epoch": 1.0345086271567892,
      "grad_norm": 0.09160026162862778,
      "learning_rate": 5.571671978744983e-05,
      "loss": 0.0143,
      "step": 1379
    },
    {
      "epoch": 1.0352588147036759,
      "grad_norm": 0.10948408395051956,
      "learning_rate": 5.565166715311711e-05,
      "loss": 0.0188,
      "step": 1380
    },
    {
      "epoch": 1.0360090022505626,
      "grad_norm": 0.05880796164274216,
      "learning_rate": 5.558660482673177e-05,
      "loss": 0.0119,
      "step": 1381
    },
    {
      "epoch": 1.0367591897974493,
      "grad_norm": 0.1001565083861351,
      "learning_rate": 5.552153291986927e-05,
      "loss": 0.0227,
      "step": 1382
    },
    {
      "epoch": 1.037509377344336,
      "grad_norm": 0.06972013413906097,
      "learning_rate": 5.5456451544121523e-05,
      "loss": 0.0175,
      "step": 1383
    },
    {
      "epoch": 1.0382595648912227,
      "grad_norm": 0.11640258133411407,
      "learning_rate": 5.5391360811096684e-05,
      "loss": 0.0156,
      "step": 1384
    },
    {
      "epoch": 1.0390097524381094,
      "grad_norm": 0.0860258936882019,
      "learning_rate": 5.5326260832418955e-05,
      "loss": 0.0173,
      "step": 1385
    },
    {
      "epoch": 1.0397599399849962,
      "grad_norm": 0.11858560144901276,
      "learning_rate": 5.526115171972838e-05,
      "loss": 0.0224,
      "step": 1386
    },
    {
      "epoch": 1.0405101275318829,
      "grad_norm": 0.10665401071310043,
      "learning_rate": 5.5196033584680675e-05,
      "loss": 0.0222,
      "step": 1387
    },
    {
      "epoch": 1.0412603150787696,
      "grad_norm": 0.08180026710033417,
      "learning_rate": 5.5130906538947034e-05,
      "loss": 0.0151,
      "step": 1388
    },
    {
      "epoch": 1.0420105026256563,
      "grad_norm": 0.10901845246553421,
      "learning_rate": 5.506577069421395e-05,
      "loss": 0.0143,
      "step": 1389
    },
    {
      "epoch": 1.042760690172543,
      "grad_norm": 0.14608041942119598,
      "learning_rate": 5.5000626162182944e-05,
      "loss": 0.0217,
      "step": 1390
    },
    {
      "epoch": 1.04351087771943,
      "grad_norm": 0.1770605593919754,
      "learning_rate": 5.49354730545705e-05,
      "loss": 0.0321,
      "step": 1391
    },
    {
      "epoch": 1.0442610652663167,
      "grad_norm": 0.11851367354393005,
      "learning_rate": 5.487031148310775e-05,
      "loss": 0.0147,
      "step": 1392
    },
    {
      "epoch": 1.0450112528132034,
      "grad_norm": 0.10560742020606995,
      "learning_rate": 5.480514155954042e-05,
      "loss": 0.0228,
      "step": 1393
    },
    {
      "epoch": 1.04576144036009,
      "grad_norm": 0.10256641358137131,
      "learning_rate": 5.4739963395628456e-05,
      "loss": 0.0153,
      "step": 1394
    },
    {
      "epoch": 1.0465116279069768,
      "grad_norm": 0.08817125111818314,
      "learning_rate": 5.4674777103146045e-05,
      "loss": 0.0142,
      "step": 1395
    },
    {
      "epoch": 1.0472618154538635,
      "grad_norm": 0.12297326326370239,
      "learning_rate": 5.460958279388122e-05,
      "loss": 0.0196,
      "step": 1396
    },
    {
      "epoch": 1.0480120030007503,
      "grad_norm": 0.07521190494298935,
      "learning_rate": 5.4544380579635824e-05,
      "loss": 0.0167,
      "step": 1397
    },
    {
      "epoch": 1.048762190547637,
      "grad_norm": 0.09659700840711594,
      "learning_rate": 5.447917057222523e-05,
      "loss": 0.0198,
      "step": 1398
    },
    {
      "epoch": 1.0495123780945237,
      "grad_norm": 0.09551558643579483,
      "learning_rate": 5.441395288347818e-05,
      "loss": 0.0156,
      "step": 1399
    },
    {
      "epoch": 1.0502625656414104,
      "grad_norm": 0.08930996060371399,
      "learning_rate": 5.434872762523658e-05,
      "loss": 0.0175,
      "step": 1400
    },
    {
      "epoch": 1.0502625656414104,
      "eval_loss": 0.03195345029234886,
      "eval_runtime": 5.1212,
      "eval_samples_per_second": 10.544,
      "eval_steps_per_second": 2.734,
      "step": 1400
    },
    {
      "epoch": 1.0510127531882971,
      "grad_norm": 0.08074672520160675,
      "learning_rate": 5.4283494909355314e-05,
      "loss": 0.014,
      "step": 1401
    },
    {
      "epoch": 1.0517629407351838,
      "grad_norm": 0.17933864891529083,
      "learning_rate": 5.42182548477021e-05,
      "loss": 0.0193,
      "step": 1402
    },
    {
      "epoch": 1.0525131282820706,
      "grad_norm": 0.11653754115104675,
      "learning_rate": 5.41530075521572e-05,
      "loss": 0.0163,
      "step": 1403
    },
    {
      "epoch": 1.0532633158289573,
      "grad_norm": 0.087211973965168,
      "learning_rate": 5.4087753134613294e-05,
      "loss": 0.0156,
      "step": 1404
    },
    {
      "epoch": 1.054013503375844,
      "grad_norm": 0.1614168882369995,
      "learning_rate": 5.40224917069753e-05,
      "loss": 0.0166,
      "step": 1405
    },
    {
      "epoch": 1.0547636909227307,
      "grad_norm": 0.10352689027786255,
      "learning_rate": 5.3957223381160126e-05,
      "loss": 0.0173,
      "step": 1406
    },
    {
      "epoch": 1.0555138784696174,
      "grad_norm": 0.1243826299905777,
      "learning_rate": 5.389194826909653e-05,
      "loss": 0.0165,
      "step": 1407
    },
    {
      "epoch": 1.0562640660165041,
      "grad_norm": 0.31279873847961426,
      "learning_rate": 5.382666648272489e-05,
      "loss": 0.0503,
      "step": 1408
    },
    {
      "epoch": 1.0570142535633908,
      "grad_norm": 0.10250795632600784,
      "learning_rate": 5.3761378133997044e-05,
      "loss": 0.0199,
      "step": 1409
    },
    {
      "epoch": 1.0577644411102776,
      "grad_norm": 0.10815533995628357,
      "learning_rate": 5.3696083334876105e-05,
      "loss": 0.0217,
      "step": 1410
    },
    {
      "epoch": 1.0585146286571643,
      "grad_norm": 0.09339367598295212,
      "learning_rate": 5.363078219733619e-05,
      "loss": 0.015,
      "step": 1411
    },
    {
      "epoch": 1.059264816204051,
      "grad_norm": 0.12286090105772018,
      "learning_rate": 5.3565474833362353e-05,
      "loss": 0.0152,
      "step": 1412
    },
    {
      "epoch": 1.0600150037509377,
      "grad_norm": 0.08419165015220642,
      "learning_rate": 5.3500161354950274e-05,
      "loss": 0.0136,
      "step": 1413
    },
    {
      "epoch": 1.0607651912978244,
      "grad_norm": 0.1648758202791214,
      "learning_rate": 5.3434841874106124e-05,
      "loss": 0.0245,
      "step": 1414
    },
    {
      "epoch": 1.0615153788447111,
      "grad_norm": 0.08925841748714447,
      "learning_rate": 5.3369516502846396e-05,
      "loss": 0.0206,
      "step": 1415
    },
    {
      "epoch": 1.0622655663915979,
      "grad_norm": 0.1235809400677681,
      "learning_rate": 5.330418535319768e-05,
      "loss": 0.0239,
      "step": 1416
    },
    {
      "epoch": 1.0630157539384846,
      "grad_norm": 0.09026822447776794,
      "learning_rate": 5.323884853719645e-05,
      "loss": 0.0128,
      "step": 1417
    },
    {
      "epoch": 1.0637659414853713,
      "grad_norm": 0.07726217061281204,
      "learning_rate": 5.31735061668889e-05,
      "loss": 0.0139,
      "step": 1418
    },
    {
      "epoch": 1.064516129032258,
      "grad_norm": 0.12053043395280838,
      "learning_rate": 5.3108158354330795e-05,
      "loss": 0.0218,
      "step": 1419
    },
    {
      "epoch": 1.0652663165791447,
      "grad_norm": 0.08617470413446426,
      "learning_rate": 5.304280521158716e-05,
      "loss": 0.0118,
      "step": 1420
    },
    {
      "epoch": 1.0660165041260314,
      "grad_norm": 0.17755377292633057,
      "learning_rate": 5.2977446850732236e-05,
      "loss": 0.0183,
      "step": 1421
    },
    {
      "epoch": 1.0667666916729182,
      "grad_norm": 0.12063173204660416,
      "learning_rate": 5.291208338384913e-05,
      "loss": 0.0224,
      "step": 1422
    },
    {
      "epoch": 1.0675168792198049,
      "grad_norm": 0.1364491879940033,
      "learning_rate": 5.2846714923029795e-05,
      "loss": 0.0197,
      "step": 1423
    },
    {
      "epoch": 1.0682670667666916,
      "grad_norm": 0.132155641913414,
      "learning_rate": 5.278134158037469e-05,
      "loss": 0.0181,
      "step": 1424
    },
    {
      "epoch": 1.0690172543135783,
      "grad_norm": 0.0874357596039772,
      "learning_rate": 5.2715963467992656e-05,
      "loss": 0.0146,
      "step": 1425
    },
    {
      "epoch": 1.069767441860465,
      "grad_norm": 0.17030027508735657,
      "learning_rate": 5.265058069800072e-05,
      "loss": 0.0358,
      "step": 1426
    },
    {
      "epoch": 1.0705176294073517,
      "grad_norm": 0.10025857388973236,
      "learning_rate": 5.258519338252389e-05,
      "loss": 0.0221,
      "step": 1427
    },
    {
      "epoch": 1.0712678169542387,
      "grad_norm": 0.10002993047237396,
      "learning_rate": 5.251980163369499e-05,
      "loss": 0.0142,
      "step": 1428
    },
    {
      "epoch": 1.0720180045011252,
      "grad_norm": 0.0852247104048729,
      "learning_rate": 5.24544055636544e-05,
      "loss": 0.0102,
      "step": 1429
    },
    {
      "epoch": 1.072768192048012,
      "grad_norm": 0.08871036022901535,
      "learning_rate": 5.2389005284549954e-05,
      "loss": 0.0136,
      "step": 1430
    },
    {
      "epoch": 1.0735183795948988,
      "grad_norm": 0.10242506861686707,
      "learning_rate": 5.232360090853671e-05,
      "loss": 0.022,
      "step": 1431
    },
    {
      "epoch": 1.0742685671417855,
      "grad_norm": 0.11138740926980972,
      "learning_rate": 5.225819254777671e-05,
      "loss": 0.0167,
      "step": 1432
    },
    {
      "epoch": 1.0750187546886723,
      "grad_norm": 0.1044057235121727,
      "learning_rate": 5.219278031443886e-05,
      "loss": 0.0125,
      "step": 1433
    },
    {
      "epoch": 1.075768942235559,
      "grad_norm": 0.1448916643857956,
      "learning_rate": 5.21273643206987e-05,
      "loss": 0.0182,
      "step": 1434
    },
    {
      "epoch": 1.0765191297824457,
      "grad_norm": 0.11542975157499313,
      "learning_rate": 5.206194467873822e-05,
      "loss": 0.0178,
      "step": 1435
    },
    {
      "epoch": 1.0772693173293324,
      "grad_norm": 0.10346329212188721,
      "learning_rate": 5.1996521500745645e-05,
      "loss": 0.0229,
      "step": 1436
    },
    {
      "epoch": 1.0780195048762191,
      "grad_norm": 0.1779703050851822,
      "learning_rate": 5.19310948989153e-05,
      "loss": 0.0219,
      "step": 1437
    },
    {
      "epoch": 1.0787696924231058,
      "grad_norm": 0.11344017088413239,
      "learning_rate": 5.186566498544737e-05,
      "loss": 0.0179,
      "step": 1438
    },
    {
      "epoch": 1.0795198799699925,
      "grad_norm": 0.08599342405796051,
      "learning_rate": 5.18002318725477e-05,
      "loss": 0.0121,
      "step": 1439
    },
    {
      "epoch": 1.0802700675168793,
      "grad_norm": 0.13371822237968445,
      "learning_rate": 5.173479567242765e-05,
      "loss": 0.018,
      "step": 1440
    },
    {
      "epoch": 1.081020255063766,
      "grad_norm": 0.16816803812980652,
      "learning_rate": 5.1669356497303835e-05,
      "loss": 0.0294,
      "step": 1441
    },
    {
      "epoch": 1.0817704426106527,
      "grad_norm": 0.11909490823745728,
      "learning_rate": 5.1603914459398016e-05,
      "loss": 0.0217,
      "step": 1442
    },
    {
      "epoch": 1.0825206301575394,
      "grad_norm": 0.1320752054452896,
      "learning_rate": 5.153846967093684e-05,
      "loss": 0.0255,
      "step": 1443
    },
    {
      "epoch": 1.0832708177044261,
      "grad_norm": 0.10966454446315765,
      "learning_rate": 5.1473022244151684e-05,
      "loss": 0.0141,
      "step": 1444
    },
    {
      "epoch": 1.0840210052513128,
      "grad_norm": 0.11884035170078278,
      "learning_rate": 5.140757229127842e-05,
      "loss": 0.016,
      "step": 1445
    },
    {
      "epoch": 1.0847711927981996,
      "grad_norm": 0.12369532883167267,
      "learning_rate": 5.1342119924557275e-05,
      "loss": 0.0189,
      "step": 1446
    },
    {
      "epoch": 1.0855213803450863,
      "grad_norm": 0.18551528453826904,
      "learning_rate": 5.127666525623264e-05,
      "loss": 0.0324,
      "step": 1447
    },
    {
      "epoch": 1.086271567891973,
      "grad_norm": 0.16022326052188873,
      "learning_rate": 5.121120839855279e-05,
      "loss": 0.0217,
      "step": 1448
    },
    {
      "epoch": 1.0870217554388597,
      "grad_norm": 0.19785533845424652,
      "learning_rate": 5.114574946376982e-05,
      "loss": 0.0234,
      "step": 1449
    },
    {
      "epoch": 1.0877719429857464,
      "grad_norm": 0.09503129124641418,
      "learning_rate": 5.1080288564139325e-05,
      "loss": 0.0131,
      "step": 1450
    },
    {
      "epoch": 1.0885221305326331,
      "grad_norm": 0.27920621633529663,
      "learning_rate": 5.101482581192033e-05,
      "loss": 0.0254,
      "step": 1451
    },
    {
      "epoch": 1.0892723180795199,
      "grad_norm": 0.11794818937778473,
      "learning_rate": 5.0949361319374996e-05,
      "loss": 0.0232,
      "step": 1452
    },
    {
      "epoch": 1.0900225056264066,
      "grad_norm": 0.1276724487543106,
      "learning_rate": 5.0883895198768494e-05,
      "loss": 0.0217,
      "step": 1453
    },
    {
      "epoch": 1.0907726931732933,
      "grad_norm": 0.08574303984642029,
      "learning_rate": 5.0818427562368764e-05,
      "loss": 0.0126,
      "step": 1454
    },
    {
      "epoch": 1.09152288072018,
      "grad_norm": 0.09915383160114288,
      "learning_rate": 5.0752958522446356e-05,
      "loss": 0.0235,
      "step": 1455
    },
    {
      "epoch": 1.0922730682670667,
      "grad_norm": 0.11475180089473724,
      "learning_rate": 5.0687488191274215e-05,
      "loss": 0.0229,
      "step": 1456
    },
    {
      "epoch": 1.0930232558139534,
      "grad_norm": 0.17217311263084412,
      "learning_rate": 5.0622016681127526e-05,
      "loss": 0.0183,
      "step": 1457
    },
    {
      "epoch": 1.0937734433608401,
      "grad_norm": 0.09383442997932434,
      "learning_rate": 5.055654410428349e-05,
      "loss": 0.0169,
      "step": 1458
    },
    {
      "epoch": 1.0945236309077269,
      "grad_norm": 0.10978903621435165,
      "learning_rate": 5.0491070573021116e-05,
      "loss": 0.0179,
      "step": 1459
    },
    {
      "epoch": 1.0952738184546136,
      "grad_norm": 0.13406389951705933,
      "learning_rate": 5.0425596199621064e-05,
      "loss": 0.0268,
      "step": 1460
    },
    {
      "epoch": 1.0960240060015003,
      "grad_norm": 0.09866870194673538,
      "learning_rate": 5.036012109636543e-05,
      "loss": 0.015,
      "step": 1461
    },
    {
      "epoch": 1.096774193548387,
      "grad_norm": 0.08531273901462555,
      "learning_rate": 5.0294645375537594e-05,
      "loss": 0.0162,
      "step": 1462
    },
    {
      "epoch": 1.0975243810952737,
      "grad_norm": 0.12082972377538681,
      "learning_rate": 5.022916914942195e-05,
      "loss": 0.0169,
      "step": 1463
    },
    {
      "epoch": 1.0982745686421604,
      "grad_norm": 0.1270955353975296,
      "learning_rate": 5.0163692530303774e-05,
      "loss": 0.0183,
      "step": 1464
    },
    {
      "epoch": 1.0990247561890472,
      "grad_norm": 0.17291484773159027,
      "learning_rate": 5.009821563046903e-05,
      "loss": 0.0209,
      "step": 1465
    },
    {
      "epoch": 1.0997749437359339,
      "grad_norm": 0.13547436892986298,
      "learning_rate": 5.003273856220415e-05,
      "loss": 0.022,
      "step": 1466
    },
    {
      "epoch": 1.1005251312828208,
      "grad_norm": 0.10618306696414948,
      "learning_rate": 4.996726143779586e-05,
      "loss": 0.0242,
      "step": 1467
    },
    {
      "epoch": 1.1012753188297075,
      "grad_norm": 0.1009022668004036,
      "learning_rate": 4.990178436953099e-05,
      "loss": 0.0156,
      "step": 1468
    },
    {
      "epoch": 1.1020255063765942,
      "grad_norm": 0.13409952819347382,
      "learning_rate": 4.9836307469696244e-05,
      "loss": 0.0177,
      "step": 1469
    },
    {
      "epoch": 1.102775693923481,
      "grad_norm": 0.07526332139968872,
      "learning_rate": 4.9770830850578075e-05,
      "loss": 0.0109,
      "step": 1470
    },
    {
      "epoch": 1.1035258814703677,
      "grad_norm": 0.09260010719299316,
      "learning_rate": 4.9705354624462424e-05,
      "loss": 0.0161,
      "step": 1471
    },
    {
      "epoch": 1.1042760690172544,
      "grad_norm": 0.11654193699359894,
      "learning_rate": 4.963987890363458e-05,
      "loss": 0.0219,
      "step": 1472
    },
    {
      "epoch": 1.105026256564141,
      "grad_norm": 0.08214106410741806,
      "learning_rate": 4.957440380037896e-05,
      "loss": 0.0173,
      "step": 1473
    },
    {
      "epoch": 1.1057764441110278,
      "grad_norm": 0.13250161707401276,
      "learning_rate": 4.9508929426978896e-05,
      "loss": 0.0202,
      "step": 1474
    },
    {
      "epoch": 1.1065266316579145,
      "grad_norm": 0.061449550092220306,
      "learning_rate": 4.944345589571651e-05,
      "loss": 0.0101,
      "step": 1475
    },
    {
      "epoch": 1.1072768192048013,
      "grad_norm": 0.13920247554779053,
      "learning_rate": 4.937798331887248e-05,
      "loss": 0.0229,
      "step": 1476
    },
    {
      "epoch": 1.108027006751688,
      "grad_norm": 0.0696333572268486,
      "learning_rate": 4.931251180872579e-05,
      "loss": 0.0143,
      "step": 1477
    },
    {
      "epoch": 1.1087771942985747,
      "grad_norm": 0.2159297913312912,
      "learning_rate": 4.9247041477553656e-05,
      "loss": 0.035,
      "step": 1478
    },
    {
      "epoch": 1.1095273818454614,
      "grad_norm": 0.17921189963817596,
      "learning_rate": 4.9181572437631255e-05,
      "loss": 0.0281,
      "step": 1479
    },
    {
      "epoch": 1.1102775693923481,
      "grad_norm": 0.11478715389966965,
      "learning_rate": 4.911610480123151e-05,
      "loss": 0.0205,
      "step": 1480
    },
    {
      "epoch": 1.1110277569392348,
      "grad_norm": 0.15886938571929932,
      "learning_rate": 4.905063868062501e-05,
      "loss": 0.0239,
      "step": 1481
    },
    {
      "epoch": 1.1117779444861215,
      "grad_norm": 0.13714122772216797,
      "learning_rate": 4.898517418807968e-05,
      "loss": 0.0223,
      "step": 1482
    },
    {
      "epoch": 1.1125281320330083,
      "grad_norm": 0.10357560217380524,
      "learning_rate": 4.891971143586069e-05,
      "loss": 0.0209,
      "step": 1483
    },
    {
      "epoch": 1.113278319579895,
      "grad_norm": 0.09297052770853043,
      "learning_rate": 4.88542505362302e-05,
      "loss": 0.0175,
      "step": 1484
    },
    {
      "epoch": 1.1140285071267817,
      "grad_norm": 0.12628571689128876,
      "learning_rate": 4.878879160144723e-05,
      "loss": 0.0198,
      "step": 1485
    },
    {
      "epoch": 1.1147786946736684,
      "grad_norm": 0.10400816053152084,
      "learning_rate": 4.872333474376739e-05,
      "loss": 0.0223,
      "step": 1486
    },
    {
      "epoch": 1.1155288822205551,
      "grad_norm": 0.0815277174115181,
      "learning_rate": 4.865788007544274e-05,
      "loss": 0.0126,
      "step": 1487
    },
    {
      "epoch": 1.1162790697674418,
      "grad_norm": 0.07159218937158585,
      "learning_rate": 4.859242770872158e-05,
      "loss": 0.0096,
      "step": 1488
    },
    {
      "epoch": 1.1170292573143286,
      "grad_norm": 0.1045546755194664,
      "learning_rate": 4.852697775584833e-05,
      "loss": 0.0158,
      "step": 1489
    },
    {
      "epoch": 1.1177794448612153,
      "grad_norm": 0.11663933098316193,
      "learning_rate": 4.846153032906316e-05,
      "loss": 0.0209,
      "step": 1490
    },
    {
      "epoch": 1.118529632408102,
      "grad_norm": 0.14321015775203705,
      "learning_rate": 4.8396085540601995e-05,
      "loss": 0.0244,
      "step": 1491
    },
    {
      "epoch": 1.1192798199549887,
      "grad_norm": 0.08914124220609665,
      "learning_rate": 4.833064350269617e-05,
      "loss": 0.018,
      "step": 1492
    },
    {
      "epoch": 1.1200300075018754,
      "grad_norm": 0.09482667595148087,
      "learning_rate": 4.826520432757236e-05,
      "loss": 0.0205,
      "step": 1493
    },
    {
      "epoch": 1.1207801950487621,
      "grad_norm": 0.11351495236158371,
      "learning_rate": 4.8199768127452314e-05,
      "loss": 0.0231,
      "step": 1494
    },
    {
      "epoch": 1.1215303825956489,
      "grad_norm": 0.17555253207683563,
      "learning_rate": 4.813433501455264e-05,
      "loss": 0.0224,
      "step": 1495
    },
    {
      "epoch": 1.1222805701425356,
      "grad_norm": 0.16453514993190765,
      "learning_rate": 4.806890510108471e-05,
      "loss": 0.0253,
      "step": 1496
    },
    {
      "epoch": 1.1230307576894223,
      "grad_norm": 0.1721792072057724,
      "learning_rate": 4.800347849925437e-05,
      "loss": 0.0266,
      "step": 1497
    },
    {
      "epoch": 1.123780945236309,
      "grad_norm": 0.13679584860801697,
      "learning_rate": 4.793805532126181e-05,
      "loss": 0.0189,
      "step": 1498
    },
    {
      "epoch": 1.1245311327831957,
      "grad_norm": 0.11366323381662369,
      "learning_rate": 4.787263567930132e-05,
      "loss": 0.0101,
      "step": 1499
    },
    {
      "epoch": 1.1252813203300824,
      "grad_norm": 0.17250248789787292,
      "learning_rate": 4.780721968556115e-05,
      "loss": 0.0304,
      "step": 1500
    },
    {
      "epoch": 1.1260315078769692,
      "grad_norm": 0.09911566227674484,
      "learning_rate": 4.774180745222331e-05,
      "loss": 0.0169,
      "step": 1501
    },
    {
      "epoch": 1.1267816954238559,
      "grad_norm": 0.08456264436244965,
      "learning_rate": 4.7676399091463296e-05,
      "loss": 0.0142,
      "step": 1502
    },
    {
      "epoch": 1.1275318829707426,
      "grad_norm": 0.14514417946338654,
      "learning_rate": 4.7610994715450044e-05,
      "loss": 0.0175,
      "step": 1503
    },
    {
      "epoch": 1.1282820705176295,
      "grad_norm": 0.1287779062986374,
      "learning_rate": 4.754559443634561e-05,
      "loss": 0.0159,
      "step": 1504
    },
    {
      "epoch": 1.129032258064516,
      "grad_norm": 0.13349425792694092,
      "learning_rate": 4.748019836630503e-05,
      "loss": 0.0299,
      "step": 1505
    },
    {
      "epoch": 1.129782445611403,
      "grad_norm": 0.11371131986379623,
      "learning_rate": 4.7414806617476124e-05,
      "loss": 0.0211,
      "step": 1506
    },
    {
      "epoch": 1.1305326331582897,
      "grad_norm": 0.13470502197742462,
      "learning_rate": 4.7349419301999294e-05,
      "loss": 0.0221,
      "step": 1507
    },
    {
      "epoch": 1.1312828207051764,
      "grad_norm": 0.10426650941371918,
      "learning_rate": 4.7284036532007356e-05,
      "loss": 0.0197,
      "step": 1508
    },
    {
      "epoch": 1.132033008252063,
      "grad_norm": 0.11969224363565445,
      "learning_rate": 4.721865841962533e-05,
      "loss": 0.0212,
      "step": 1509
    },
    {
      "epoch": 1.1327831957989498,
      "grad_norm": 0.08736421167850494,
      "learning_rate": 4.715328507697021e-05,
      "loss": 0.0165,
      "step": 1510
    },
    {
      "epoch": 1.1335333833458365,
      "grad_norm": 0.12223305553197861,
      "learning_rate": 4.7087916616150886e-05,
      "loss": 0.0209,
      "step": 1511
    },
    {
      "epoch": 1.1342835708927232,
      "grad_norm": 0.1267385184764862,
      "learning_rate": 4.702255314926779e-05,
      "loss": 0.0205,
      "step": 1512
    },
    {
      "epoch": 1.13503375843961,
      "grad_norm": 0.08564191311597824,
      "learning_rate": 4.695719478841286e-05,
      "loss": 0.0175,
      "step": 1513
    },
    {
      "epoch": 1.1357839459864967,
      "grad_norm": 0.12604007124900818,
      "learning_rate": 4.6891841645669224e-05,
      "loss": 0.0168,
      "step": 1514
    },
    {
      "epoch": 1.1365341335333834,
      "grad_norm": 0.08834186941385269,
      "learning_rate": 4.6826493833111104e-05,
      "loss": 0.012,
      "step": 1515
    },
    {
      "epoch": 1.13728432108027,
      "grad_norm": 0.1784517616033554,
      "learning_rate": 4.676115146280356e-05,
      "loss": 0.027,
      "step": 1516
    },
    {
      "epoch": 1.1380345086271568,
      "grad_norm": 0.07656119018793106,
      "learning_rate": 4.669581464680233e-05,
      "loss": 0.0157,
      "step": 1517
    },
    {
      "epoch": 1.1387846961740435,
      "grad_norm": 0.09516746550798416,
      "learning_rate": 4.66304834971536e-05,
      "loss": 0.0144,
      "step": 1518
    },
    {
      "epoch": 1.1395348837209303,
      "grad_norm": 0.08436454832553864,
      "learning_rate": 4.656515812589389e-05,
      "loss": 0.0111,
      "step": 1519
    },
    {
      "epoch": 1.140285071267817,
      "grad_norm": 0.06468910723924637,
      "learning_rate": 4.6499838645049744e-05,
      "loss": 0.0107,
      "step": 1520
    },
    {
      "epoch": 1.1410352588147037,
      "grad_norm": 0.11399751156568527,
      "learning_rate": 4.643452516663766e-05,
      "loss": 0.0216,
      "step": 1521
    },
    {
      "epoch": 1.1417854463615904,
      "grad_norm": 0.13065651059150696,
      "learning_rate": 4.636921780266381e-05,
      "loss": 0.022,
      "step": 1522
    },
    {
      "epoch": 1.1425356339084771,
      "grad_norm": 0.07837221771478653,
      "learning_rate": 4.63039166651239e-05,
      "loss": 0.0149,
      "step": 1523
    },
    {
      "epoch": 1.1432858214553638,
      "grad_norm": 0.10143603384494781,
      "learning_rate": 4.623862186600297e-05,
      "loss": 0.0136,
      "step": 1524
    },
    {
      "epoch": 1.1440360090022506,
      "grad_norm": 0.09773246943950653,
      "learning_rate": 4.617333351727513e-05,
      "loss": 0.0174,
      "step": 1525
    },
    {
      "epoch": 1.1447861965491373,
      "grad_norm": 0.11636144667863846,
      "learning_rate": 4.61080517309035e-05,
      "loss": 0.024,
      "step": 1526
    },
    {
      "epoch": 1.145536384096024,
      "grad_norm": 0.10740485787391663,
      "learning_rate": 4.604277661883989e-05,
      "loss": 0.0186,
      "step": 1527
    },
    {
      "epoch": 1.1462865716429107,
      "grad_norm": 0.085996612906456,
      "learning_rate": 4.5977508293024726e-05,
      "loss": 0.0108,
      "step": 1528
    },
    {
      "epoch": 1.1470367591897974,
      "grad_norm": 0.1368994265794754,
      "learning_rate": 4.591224686538672e-05,
      "loss": 0.0194,
      "step": 1529
    },
    {
      "epoch": 1.1477869467366841,
      "grad_norm": 0.17650572955608368,
      "learning_rate": 4.584699244784281e-05,
      "loss": 0.0253,
      "step": 1530
    },
    {
      "epoch": 1.1485371342835708,
      "grad_norm": 0.19057302176952362,
      "learning_rate": 4.578174515229789e-05,
      "loss": 0.0376,
      "step": 1531
    },
    {
      "epoch": 1.1492873218304576,
      "grad_norm": 0.07737153023481369,
      "learning_rate": 4.5716505090644684e-05,
      "loss": 0.015,
      "step": 1532
    },
    {
      "epoch": 1.1500375093773443,
      "grad_norm": 0.1411726176738739,
      "learning_rate": 4.5651272374763423e-05,
      "loss": 0.0221,
      "step": 1533
    },
    {
      "epoch": 1.150787696924231,
      "grad_norm": 0.10764733701944351,
      "learning_rate": 4.558604711652183e-05,
      "loss": 0.0175,
      "step": 1534
    },
    {
      "epoch": 1.1515378844711177,
      "grad_norm": 0.11171084642410278,
      "learning_rate": 4.552082942777478e-05,
      "loss": 0.0182,
      "step": 1535
    },
    {
      "epoch": 1.1522880720180044,
      "grad_norm": 0.12677061557769775,
      "learning_rate": 4.545561942036418e-05,
      "loss": 0.0183,
      "step": 1536
    },
    {
      "epoch": 1.1530382595648911,
      "grad_norm": 0.12518805265426636,
      "learning_rate": 4.5390417206118784e-05,
      "loss": 0.0187,
      "step": 1537
    },
    {
      "epoch": 1.1537884471117779,
      "grad_norm": 0.11542828381061554,
      "learning_rate": 4.5325222896853966e-05,
      "loss": 0.0224,
      "step": 1538
    },
    {
      "epoch": 1.1545386346586646,
      "grad_norm": 0.14305250346660614,
      "learning_rate": 4.5260036604371556e-05,
      "loss": 0.0222,
      "step": 1539
    },
    {
      "epoch": 1.1552888222055513,
      "grad_norm": 0.12537597119808197,
      "learning_rate": 4.51948584404596e-05,
      "loss": 0.0173,
      "step": 1540
    },
    {
      "epoch": 1.1560390097524382,
      "grad_norm": 0.0862974002957344,
      "learning_rate": 4.5129688516892264e-05,
      "loss": 0.0157,
      "step": 1541
    },
    {
      "epoch": 1.1567891972993247,
      "grad_norm": 0.1084955483675003,
      "learning_rate": 4.506452694542953e-05,
      "loss": 0.0198,
      "step": 1542
    },
    {
      "epoch": 1.1575393848462117,
      "grad_norm": 0.10898890346288681,
      "learning_rate": 4.499937383781708e-05,
      "loss": 0.0166,
      "step": 1543
    },
    {
      "epoch": 1.1582895723930982,
      "grad_norm": 0.07213619351387024,
      "learning_rate": 4.493422930578605e-05,
      "loss": 0.0117,
      "step": 1544
    },
    {
      "epoch": 1.159039759939985,
      "grad_norm": 0.11787864565849304,
      "learning_rate": 4.486909346105296e-05,
      "loss": 0.0235,
      "step": 1545
    },
    {
      "epoch": 1.1597899474868718,
      "grad_norm": 0.12363395094871521,
      "learning_rate": 4.480396641531932e-05,
      "loss": 0.0221,
      "step": 1546
    },
    {
      "epoch": 1.1605401350337585,
      "grad_norm": 0.10760311782360077,
      "learning_rate": 4.4738848280271626e-05,
      "loss": 0.0176,
      "step": 1547
    },
    {
      "epoch": 1.1612903225806452,
      "grad_norm": 0.08016765117645264,
      "learning_rate": 4.467373916758105e-05,
      "loss": 0.0098,
      "step": 1548
    },
    {
      "epoch": 1.162040510127532,
      "grad_norm": 0.19417689740657806,
      "learning_rate": 4.460863918890333e-05,
      "loss": 0.0238,
      "step": 1549
    },
    {
      "epoch": 1.1627906976744187,
      "grad_norm": 0.13339464366436005,
      "learning_rate": 4.454354845587849e-05,
      "loss": 0.0126,
      "step": 1550
    },
    {
      "epoch": 1.1635408852213054,
      "grad_norm": 0.09546097368001938,
      "learning_rate": 4.4478467080130734e-05,
      "loss": 0.0176,
      "step": 1551
    },
    {
      "epoch": 1.164291072768192,
      "grad_norm": 0.12648361921310425,
      "learning_rate": 4.4413395173268243e-05,
      "loss": 0.0207,
      "step": 1552
    },
    {
      "epoch": 1.1650412603150788,
      "grad_norm": 0.14633196592330933,
      "learning_rate": 4.43483328468829e-05,
      "loss": 0.0188,
      "step": 1553
    },
    {
      "epoch": 1.1657914478619655,
      "grad_norm": 0.12665541470050812,
      "learning_rate": 4.4283280212550194e-05,
      "loss": 0.0186,
      "step": 1554
    },
    {
      "epoch": 1.1665416354088523,
      "grad_norm": 0.08633586019277573,
      "learning_rate": 4.421823738182898e-05,
      "loss": 0.0118,
      "step": 1555
    },
    {
      "epoch": 1.167291822955739,
      "grad_norm": 0.09715209156274796,
      "learning_rate": 4.4153204466261334e-05,
      "loss": 0.0156,
      "step": 1556
    },
    {
      "epoch": 1.1680420105026257,
      "grad_norm": 0.09635809808969498,
      "learning_rate": 4.408818157737227e-05,
      "loss": 0.0135,
      "step": 1557
    },
    {
      "epoch": 1.1687921980495124,
      "grad_norm": 0.14372672140598297,
      "learning_rate": 4.402316882666964e-05,
      "loss": 0.0222,
      "step": 1558
    },
    {
      "epoch": 1.1695423855963991,
      "grad_norm": 0.12557637691497803,
      "learning_rate": 4.395816632564393e-05,
      "loss": 0.0183,
      "step": 1559
    },
    {
      "epoch": 1.1702925731432858,
      "grad_norm": 0.13715486228466034,
      "learning_rate": 4.3893174185768045e-05,
      "loss": 0.0242,
      "step": 1560
    },
    {
      "epoch": 1.1710427606901725,
      "grad_norm": 0.14162661135196686,
      "learning_rate": 4.382819251849707e-05,
      "loss": 0.0181,
      "step": 1561
    },
    {
      "epoch": 1.1717929482370593,
      "grad_norm": 0.08794243633747101,
      "learning_rate": 4.376322143526818e-05,
      "loss": 0.0149,
      "step": 1562
    },
    {
      "epoch": 1.172543135783946,
      "grad_norm": 0.13961324095726013,
      "learning_rate": 4.3698261047500376e-05,
      "loss": 0.0206,
      "step": 1563
    },
    {
      "epoch": 1.1732933233308327,
      "grad_norm": 0.1918809711933136,
      "learning_rate": 4.3633311466594345e-05,
      "loss": 0.019,
      "step": 1564
    },
    {
      "epoch": 1.1740435108777194,
      "grad_norm": 0.12562909722328186,
      "learning_rate": 4.3568372803932195e-05,
      "loss": 0.0167,
      "step": 1565
    },
    {
      "epoch": 1.1747936984246061,
      "grad_norm": 0.18379457294940948,
      "learning_rate": 4.3503445170877354e-05,
      "loss": 0.0223,
      "step": 1566
    },
    {
      "epoch": 1.1755438859714928,
      "grad_norm": 0.09681471437215805,
      "learning_rate": 4.343852867877433e-05,
      "loss": 0.0178,
      "step": 1567
    },
    {
      "epoch": 1.1762940735183796,
      "grad_norm": 0.17194296419620514,
      "learning_rate": 4.3373623438948496e-05,
      "loss": 0.0291,
      "step": 1568
    },
    {
      "epoch": 1.1770442610652663,
      "grad_norm": 0.08721692860126495,
      "learning_rate": 4.330872956270596e-05,
      "loss": 0.0164,
      "step": 1569
    },
    {
      "epoch": 1.177794448612153,
      "grad_norm": 0.13167378306388855,
      "learning_rate": 4.324384716133332e-05,
      "loss": 0.0189,
      "step": 1570
    },
    {
      "epoch": 1.1785446361590397,
      "grad_norm": 0.14486682415008545,
      "learning_rate": 4.317897634609751e-05,
      "loss": 0.0174,
      "step": 1571
    },
    {
      "epoch": 1.1792948237059264,
      "grad_norm": 0.2041822224855423,
      "learning_rate": 4.3114117228245565e-05,
      "loss": 0.0306,
      "step": 1572
    },
    {
      "epoch": 1.1800450112528131,
      "grad_norm": 0.10399215668439865,
      "learning_rate": 4.304926991900453e-05,
      "loss": 0.0162,
      "step": 1573
    },
    {
      "epoch": 1.1807951987996999,
      "grad_norm": 0.11682548373937607,
      "learning_rate": 4.298443452958113e-05,
      "loss": 0.0169,
      "step": 1574
    },
    {
      "epoch": 1.1815453863465866,
      "grad_norm": 0.10943344980478287,
      "learning_rate": 4.291961117116168e-05,
      "loss": 0.0154,
      "step": 1575
    },
    {
      "epoch": 1.1822955738934733,
      "grad_norm": 0.07967270910739899,
      "learning_rate": 4.285479995491185e-05,
      "loss": 0.0108,
      "step": 1576
    },
    {
      "epoch": 1.18304576144036,
      "grad_norm": 0.11803645640611649,
      "learning_rate": 4.279000099197651e-05,
      "loss": 0.0173,
      "step": 1577
    },
    {
      "epoch": 1.1837959489872467,
      "grad_norm": 0.06087150797247887,
      "learning_rate": 4.272521439347947e-05,
      "loss": 0.0098,
      "step": 1578
    },
    {
      "epoch": 1.1845461365341334,
      "grad_norm": 0.08632830530405045,
      "learning_rate": 4.26604402705234e-05,
      "loss": 0.0122,
      "step": 1579
    },
    {
      "epoch": 1.1852963240810204,
      "grad_norm": 0.09882350265979767,
      "learning_rate": 4.259567873418952e-05,
      "loss": 0.0162,
      "step": 1580
    },
    {
      "epoch": 1.1860465116279069,
      "grad_norm": 0.1528325378894806,
      "learning_rate": 4.25309298955375e-05,
      "loss": 0.0219,
      "step": 1581
    },
    {
      "epoch": 1.1867966991747938,
      "grad_norm": 0.10851482301950455,
      "learning_rate": 4.246619386560521e-05,
      "loss": 0.011,
      "step": 1582
    },
    {
      "epoch": 1.1875468867216805,
      "grad_norm": 0.05488813295960426,
      "learning_rate": 4.240147075540858e-05,
      "loss": 0.0105,
      "step": 1583
    },
    {
      "epoch": 1.1882970742685672,
      "grad_norm": 0.07339299470186234,
      "learning_rate": 4.233676067594137e-05,
      "loss": 0.0129,
      "step": 1584
    },
    {
      "epoch": 1.189047261815454,
      "grad_norm": 0.13282492756843567,
      "learning_rate": 4.227206373817497e-05,
      "loss": 0.0246,
      "step": 1585
    },
    {
      "epoch": 1.1897974493623407,
      "grad_norm": 0.1161782294511795,
      "learning_rate": 4.220738005305827e-05,
      "loss": 0.0171,
      "step": 1586
    },
    {
      "epoch": 1.1905476369092274,
      "grad_norm": 0.11853188276290894,
      "learning_rate": 4.214270973151745e-05,
      "loss": 0.0259,
      "step": 1587
    },
    {
      "epoch": 1.191297824456114,
      "grad_norm": 0.06727686524391174,
      "learning_rate": 4.207805288445571e-05,
      "loss": 0.0107,
      "step": 1588
    },
    {
      "epoch": 1.1920480120030008,
      "grad_norm": 0.14418961107730865,
      "learning_rate": 4.201340962275318e-05,
      "loss": 0.0303,
      "step": 1589
    },
    {
      "epoch": 1.1927981995498875,
      "grad_norm": 0.06462909281253815,
      "learning_rate": 4.194878005726671e-05,
      "loss": 0.0101,
      "step": 1590
    },
    {
      "epoch": 1.1935483870967742,
      "grad_norm": 0.11350913345813751,
      "learning_rate": 4.1884164298829615e-05,
      "loss": 0.0128,
      "step": 1591
    },
    {
      "epoch": 1.194298574643661,
      "grad_norm": 0.11162211000919342,
      "learning_rate": 4.181956245825158e-05,
      "loss": 0.014,
      "step": 1592
    },
    {
      "epoch": 1.1950487621905477,
      "grad_norm": 0.13950911164283752,
      "learning_rate": 4.1754974646318365e-05,
      "loss": 0.019,
      "step": 1593
    },
    {
      "epoch": 1.1957989497374344,
      "grad_norm": 0.17123188078403473,
      "learning_rate": 4.1690400973791756e-05,
      "loss": 0.0229,
      "step": 1594
    },
    {
      "epoch": 1.196549137284321,
      "grad_norm": 0.09660036861896515,
      "learning_rate": 4.1625841551409195e-05,
      "loss": 0.0164,
      "step": 1595
    },
    {
      "epoch": 1.1972993248312078,
      "grad_norm": 0.19667930901050568,
      "learning_rate": 4.156129648988376e-05,
      "loss": 0.0205,
      "step": 1596
    },
    {
      "epoch": 1.1980495123780945,
      "grad_norm": 0.047219861298799515,
      "learning_rate": 4.149676589990388e-05,
      "loss": 0.008,
      "step": 1597
    },
    {
      "epoch": 1.1987996999249813,
      "grad_norm": 0.17346802353858948,
      "learning_rate": 4.143224989213315e-05,
      "loss": 0.0238,
      "step": 1598
    },
    {
      "epoch": 1.199549887471868,
      "grad_norm": 0.11115290224552155,
      "learning_rate": 4.136774857721017e-05,
      "loss": 0.016,
      "step": 1599
    },
    {
      "epoch": 1.2003000750187547,
      "grad_norm": 0.13619063794612885,
      "learning_rate": 4.130326206574834e-05,
      "loss": 0.0164,
      "step": 1600
    },
    {
      "epoch": 1.2003000750187547,
      "eval_loss": 0.03162379190325737,
      "eval_runtime": 5.1406,
      "eval_samples_per_second": 10.505,
      "eval_steps_per_second": 2.723,
      "step": 1600
    },
    {
      "epoch": 1.2010502625656414,
      "grad_norm": 0.10646115243434906,
      "learning_rate": 4.1238790468335685e-05,
      "loss": 0.0122,
      "step": 1601
    },
    {
      "epoch": 1.2018004501125281,
      "grad_norm": 0.11590476334095001,
      "learning_rate": 4.117433389553466e-05,
      "loss": 0.017,
      "step": 1602
    },
    {
      "epoch": 1.2025506376594148,
      "grad_norm": 0.20600537955760956,
      "learning_rate": 4.1109892457881924e-05,
      "loss": 0.0302,
      "step": 1603
    },
    {
      "epoch": 1.2033008252063015,
      "grad_norm": 0.10678470134735107,
      "learning_rate": 4.1045466265888195e-05,
      "loss": 0.0148,
      "step": 1604
    },
    {
      "epoch": 1.2040510127531883,
      "grad_norm": 0.12323981523513794,
      "learning_rate": 4.0981055430038055e-05,
      "loss": 0.0178,
      "step": 1605
    },
    {
      "epoch": 1.204801200300075,
      "grad_norm": 0.09657885879278183,
      "learning_rate": 4.091666006078974e-05,
      "loss": 0.0119,
      "step": 1606
    },
    {
      "epoch": 1.2055513878469617,
      "grad_norm": 0.1022537425160408,
      "learning_rate": 4.085228026857498e-05,
      "loss": 0.012,
      "step": 1607
    },
    {
      "epoch": 1.2063015753938484,
      "grad_norm": 0.1730087846517563,
      "learning_rate": 4.0787916163798743e-05,
      "loss": 0.0217,
      "step": 1608
    },
    {
      "epoch": 1.2070517629407351,
      "grad_norm": 0.1198115199804306,
      "learning_rate": 4.0723567856839184e-05,
      "loss": 0.0172,
      "step": 1609
    },
    {
      "epoch": 1.2078019504876218,
      "grad_norm": 0.13980574905872345,
      "learning_rate": 4.0659235458047264e-05,
      "loss": 0.0192,
      "step": 1610
    },
    {
      "epoch": 1.2085521380345086,
      "grad_norm": 0.08618589490652084,
      "learning_rate": 4.0594919077746734e-05,
      "loss": 0.0109,
      "step": 1611
    },
    {
      "epoch": 1.2093023255813953,
      "grad_norm": 0.14428244531154633,
      "learning_rate": 4.053061882623386e-05,
      "loss": 0.0215,
      "step": 1612
    },
    {
      "epoch": 1.210052513128282,
      "grad_norm": 0.10157791525125504,
      "learning_rate": 4.0466334813777216e-05,
      "loss": 0.0175,
      "step": 1613
    },
    {
      "epoch": 1.2108027006751687,
      "grad_norm": 0.0878187045454979,
      "learning_rate": 4.040206715061758e-05,
      "loss": 0.0162,
      "step": 1614
    },
    {
      "epoch": 1.2115528882220554,
      "grad_norm": 0.18254134058952332,
      "learning_rate": 4.033781594696767e-05,
      "loss": 0.0277,
      "step": 1615
    },
    {
      "epoch": 1.2123030757689421,
      "grad_norm": 0.10520658642053604,
      "learning_rate": 4.027358131301194e-05,
      "loss": 0.0191,
      "step": 1616
    },
    {
      "epoch": 1.213053263315829,
      "grad_norm": 0.2057482898235321,
      "learning_rate": 4.0209363358906495e-05,
      "loss": 0.015,
      "step": 1617
    },
    {
      "epoch": 1.2138034508627156,
      "grad_norm": 0.14618700742721558,
      "learning_rate": 4.014516219477878e-05,
      "loss": 0.0178,
      "step": 1618
    },
    {
      "epoch": 1.2145536384096025,
      "grad_norm": 0.08962513506412506,
      "learning_rate": 4.008097793072749e-05,
      "loss": 0.0177,
      "step": 1619
    },
    {
      "epoch": 1.215303825956489,
      "grad_norm": 0.10952772945165634,
      "learning_rate": 4.00168106768223e-05,
      "loss": 0.0196,
      "step": 1620
    },
    {
      "epoch": 1.216054013503376,
      "grad_norm": 0.09128102660179138,
      "learning_rate": 3.9952660543103734e-05,
      "loss": 0.0151,
      "step": 1621
    },
    {
      "epoch": 1.2168042010502627,
      "grad_norm": 0.16668932139873505,
      "learning_rate": 3.988852763958297e-05,
      "loss": 0.0185,
      "step": 1622
    },
    {
      "epoch": 1.2175543885971494,
      "grad_norm": 0.13736192882061005,
      "learning_rate": 3.9824412076241595e-05,
      "loss": 0.0163,
      "step": 1623
    },
    {
      "epoch": 1.218304576144036,
      "grad_norm": 0.1512220799922943,
      "learning_rate": 3.9760313963031516e-05,
      "loss": 0.0238,
      "step": 1624
    },
    {
      "epoch": 1.2190547636909228,
      "grad_norm": 0.08548334985971451,
      "learning_rate": 3.9696233409874654e-05,
      "loss": 0.016,
      "step": 1625
    },
    {
      "epoch": 1.2198049512378095,
      "grad_norm": 0.10499925911426544,
      "learning_rate": 3.963217052666287e-05,
      "loss": 0.0154,
      "step": 1626
    },
    {
      "epoch": 1.2205551387846962,
      "grad_norm": 0.08598527312278748,
      "learning_rate": 3.956812542325769e-05,
      "loss": 0.0145,
      "step": 1627
    },
    {
      "epoch": 1.221305326331583,
      "grad_norm": 0.22342023253440857,
      "learning_rate": 3.950409820949018e-05,
      "loss": 0.0288,
      "step": 1628
    },
    {
      "epoch": 1.2220555138784697,
      "grad_norm": 0.20031727850437164,
      "learning_rate": 3.9440088995160676e-05,
      "loss": 0.0262,
      "step": 1629
    },
    {
      "epoch": 1.2228057014253564,
      "grad_norm": 0.11276500672101974,
      "learning_rate": 3.937609789003871e-05,
      "loss": 0.0149,
      "step": 1630
    },
    {
      "epoch": 1.223555888972243,
      "grad_norm": 0.05976317077875137,
      "learning_rate": 3.93121250038627e-05,
      "loss": 0.0109,
      "step": 1631
    },
    {
      "epoch": 1.2243060765191298,
      "grad_norm": 0.1105174720287323,
      "learning_rate": 3.924817044633985e-05,
      "loss": 0.0146,
      "step": 1632
    },
    {
      "epoch": 1.2250562640660165,
      "grad_norm": 0.16175299882888794,
      "learning_rate": 3.9184234327145954e-05,
      "loss": 0.0199,
      "step": 1633
    },
    {
      "epoch": 1.2258064516129032,
      "grad_norm": 0.14765483140945435,
      "learning_rate": 3.912031675592512e-05,
      "loss": 0.0169,
      "step": 1634
    },
    {
      "epoch": 1.22655663915979,
      "grad_norm": 0.10759598761796951,
      "learning_rate": 3.905641784228972e-05,
      "loss": 0.0164,
      "step": 1635
    },
    {
      "epoch": 1.2273068267066767,
      "grad_norm": 0.0886867418885231,
      "learning_rate": 3.899253769582008e-05,
      "loss": 0.0134,
      "step": 1636
    },
    {
      "epoch": 1.2280570142535634,
      "grad_norm": 0.14302989840507507,
      "learning_rate": 3.8928676426064376e-05,
      "loss": 0.016,
      "step": 1637
    },
    {
      "epoch": 1.22880720180045,
      "grad_norm": 0.15930666029453278,
      "learning_rate": 3.886483414253838e-05,
      "loss": 0.0137,
      "step": 1638
    },
    {
      "epoch": 1.2295573893473368,
      "grad_norm": 0.10902878642082214,
      "learning_rate": 3.880101095472535e-05,
      "loss": 0.0149,
      "step": 1639
    },
    {
      "epoch": 1.2303075768942235,
      "grad_norm": 0.11319921165704727,
      "learning_rate": 3.873720697207572e-05,
      "loss": 0.0202,
      "step": 1640
    },
    {
      "epoch": 1.2310577644411103,
      "grad_norm": 0.08670421689748764,
      "learning_rate": 3.867342230400707e-05,
      "loss": 0.0115,
      "step": 1641
    },
    {
      "epoch": 1.231807951987997,
      "grad_norm": 0.07328299432992935,
      "learning_rate": 3.860965705990383e-05,
      "loss": 0.0113,
      "step": 1642
    },
    {
      "epoch": 1.2325581395348837,
      "grad_norm": 0.10198184847831726,
      "learning_rate": 3.8545911349117114e-05,
      "loss": 0.0153,
      "step": 1643
    },
    {
      "epoch": 1.2333083270817704,
      "grad_norm": 0.1404053270816803,
      "learning_rate": 3.848218528096452e-05,
      "loss": 0.0196,
      "step": 1644
    },
    {
      "epoch": 1.2340585146286571,
      "grad_norm": 0.12451205402612686,
      "learning_rate": 3.841847896473001e-05,
      "loss": 0.0157,
      "step": 1645
    },
    {
      "epoch": 1.2348087021755438,
      "grad_norm": 0.09027544409036636,
      "learning_rate": 3.83547925096636e-05,
      "loss": 0.0143,
      "step": 1646
    },
    {
      "epoch": 1.2355588897224306,
      "grad_norm": 0.15218143165111542,
      "learning_rate": 3.829112602498132e-05,
      "loss": 0.0265,
      "step": 1647
    },
    {
      "epoch": 1.2363090772693173,
      "grad_norm": 0.13812431693077087,
      "learning_rate": 3.822747961986493e-05,
      "loss": 0.0179,
      "step": 1648
    },
    {
      "epoch": 1.237059264816204,
      "grad_norm": 0.07685697078704834,
      "learning_rate": 3.816385340346171e-05,
      "loss": 0.0111,
      "step": 1649
    },
    {
      "epoch": 1.2378094523630907,
      "grad_norm": 0.12320845574140549,
      "learning_rate": 3.81002474848844e-05,
      "loss": 0.0163,
      "step": 1650
    },
    {
      "epoch": 1.2385596399099774,
      "grad_norm": 0.09731096774339676,
      "learning_rate": 3.803666197321084e-05,
      "loss": 0.0166,
      "step": 1651
    },
    {
      "epoch": 1.2393098274568641,
      "grad_norm": 0.1925540417432785,
      "learning_rate": 3.797309697748396e-05,
      "loss": 0.0228,
      "step": 1652
    },
    {
      "epoch": 1.2400600150037508,
      "grad_norm": 0.08675350993871689,
      "learning_rate": 3.7909552606711454e-05,
      "loss": 0.012,
      "step": 1653
    },
    {
      "epoch": 1.2408102025506376,
      "grad_norm": 0.13772478699684143,
      "learning_rate": 3.784602896986566e-05,
      "loss": 0.0158,
      "step": 1654
    },
    {
      "epoch": 1.2415603900975243,
      "grad_norm": 0.12192375957965851,
      "learning_rate": 3.778252617588334e-05,
      "loss": 0.0109,
      "step": 1655
    },
    {
      "epoch": 1.2423105776444112,
      "grad_norm": 0.1430065929889679,
      "learning_rate": 3.771904433366557e-05,
      "loss": 0.0238,
      "step": 1656
    },
    {
      "epoch": 1.2430607651912977,
      "grad_norm": 0.09612610936164856,
      "learning_rate": 3.7655583552077446e-05,
      "loss": 0.0168,
      "step": 1657
    },
    {
      "epoch": 1.2438109527381846,
      "grad_norm": 0.10175983607769012,
      "learning_rate": 3.7592143939947955e-05,
      "loss": 0.0121,
      "step": 1658
    },
    {
      "epoch": 1.2445611402850711,
      "grad_norm": 0.09988542646169662,
      "learning_rate": 3.7528725606069774e-05,
      "loss": 0.0156,
      "step": 1659
    },
    {
      "epoch": 1.245311327831958,
      "grad_norm": 0.3816523849964142,
      "learning_rate": 3.746532865919913e-05,
      "loss": 0.038,
      "step": 1660
    },
    {
      "epoch": 1.2460615153788448,
      "grad_norm": 0.1465483009815216,
      "learning_rate": 3.740195320805551e-05,
      "loss": 0.0234,
      "step": 1661
    },
    {
      "epoch": 1.2468117029257315,
      "grad_norm": 0.1917344033718109,
      "learning_rate": 3.733859936132158e-05,
      "loss": 0.0182,
      "step": 1662
    },
    {
      "epoch": 1.2475618904726182,
      "grad_norm": 0.1488763391971588,
      "learning_rate": 3.727526722764297e-05,
      "loss": 0.0208,
      "step": 1663
    },
    {
      "epoch": 1.248312078019505,
      "grad_norm": 0.11044822633266449,
      "learning_rate": 3.7211956915628035e-05,
      "loss": 0.0175,
      "step": 1664
    },
    {
      "epoch": 1.2490622655663917,
      "grad_norm": 0.10706579685211182,
      "learning_rate": 3.7148668533847744e-05,
      "loss": 0.0158,
      "step": 1665
    },
    {
      "epoch": 1.2498124531132784,
      "grad_norm": 0.10008851438760757,
      "learning_rate": 3.7085402190835406e-05,
      "loss": 0.0148,
      "step": 1666
    },
    {
      "epoch": 1.250562640660165,
      "grad_norm": 0.10831057280302048,
      "learning_rate": 3.702215799508659e-05,
      "loss": 0.0175,
      "step": 1667
    },
    {
      "epoch": 1.2513128282070518,
      "grad_norm": 0.10080386698246002,
      "learning_rate": 3.695893605505887e-05,
      "loss": 0.0142,
      "step": 1668
    },
    {
      "epoch": 1.2520630157539385,
      "grad_norm": 0.10447536408901215,
      "learning_rate": 3.689573647917162e-05,
      "loss": 0.0181,
      "step": 1669
    },
    {
      "epoch": 1.2528132033008252,
      "grad_norm": 0.1283860206604004,
      "learning_rate": 3.683255937580592e-05,
      "loss": 0.0165,
      "step": 1670
    },
    {
      "epoch": 1.253563390847712,
      "grad_norm": 0.14649006724357605,
      "learning_rate": 3.6769404853304276e-05,
      "loss": 0.0125,
      "step": 1671
    },
    {
      "epoch": 1.2543135783945987,
      "grad_norm": 0.09096325188875198,
      "learning_rate": 3.670627301997047e-05,
      "loss": 0.0175,
      "step": 1672
    },
    {
      "epoch": 1.2550637659414854,
      "grad_norm": 0.11983542889356613,
      "learning_rate": 3.664316398406939e-05,
      "loss": 0.0154,
      "step": 1673
    },
    {
      "epoch": 1.255813953488372,
      "grad_norm": 0.077834352850914,
      "learning_rate": 3.658007785382679e-05,
      "loss": 0.0104,
      "step": 1674
    },
    {
      "epoch": 1.2565641410352588,
      "grad_norm": 0.12809742987155914,
      "learning_rate": 3.65170147374292e-05,
      "loss": 0.0214,
      "step": 1675
    },
    {
      "epoch": 1.2573143285821455,
      "grad_norm": 0.11058270931243896,
      "learning_rate": 3.645397474302363e-05,
      "loss": 0.015,
      "step": 1676
    },
    {
      "epoch": 1.2580645161290323,
      "grad_norm": 0.1564393937587738,
      "learning_rate": 3.639095797871748e-05,
      "loss": 0.0179,
      "step": 1677
    },
    {
      "epoch": 1.258814703675919,
      "grad_norm": 0.11511152237653732,
      "learning_rate": 3.63279645525783e-05,
      "loss": 0.0194,
      "step": 1678
    },
    {
      "epoch": 1.2595648912228057,
      "grad_norm": 0.05084642395377159,
      "learning_rate": 3.626499457263359e-05,
      "loss": 0.0075,
      "step": 1679
    },
    {
      "epoch": 1.2603150787696924,
      "grad_norm": 0.16043655574321747,
      "learning_rate": 3.620204814687069e-05,
      "loss": 0.0181,
      "step": 1680
    },
    {
      "epoch": 1.2610652663165791,
      "grad_norm": 0.09519901871681213,
      "learning_rate": 3.61391253832365e-05,
      "loss": 0.0131,
      "step": 1681
    },
    {
      "epoch": 1.2618154538634658,
      "grad_norm": 0.0991373062133789,
      "learning_rate": 3.607622638963739e-05,
      "loss": 0.0153,
      "step": 1682
    },
    {
      "epoch": 1.2625656414103525,
      "grad_norm": 0.15621377527713776,
      "learning_rate": 3.601335127393889e-05,
      "loss": 0.0159,
      "step": 1683
    },
    {
      "epoch": 1.2633158289572393,
      "grad_norm": 0.16243353486061096,
      "learning_rate": 3.59505001439657e-05,
      "loss": 0.0229,
      "step": 1684
    },
    {
      "epoch": 1.264066016504126,
      "grad_norm": 0.169635608792305,
      "learning_rate": 3.588767310750127e-05,
      "loss": 0.0233,
      "step": 1685
    },
    {
      "epoch": 1.2648162040510127,
      "grad_norm": 0.13255098462104797,
      "learning_rate": 3.5824870272287815e-05,
      "loss": 0.0227,
      "step": 1686
    },
    {
      "epoch": 1.2655663915978994,
      "grad_norm": 0.15259072184562683,
      "learning_rate": 3.576209174602597e-05,
      "loss": 0.0198,
      "step": 1687
    },
    {
      "epoch": 1.2663165791447861,
      "grad_norm": 0.15271423757076263,
      "learning_rate": 3.569933763637477e-05,
      "loss": 0.0186,
      "step": 1688
    },
    {
      "epoch": 1.2670667666916728,
      "grad_norm": 0.0755571573972702,
      "learning_rate": 3.56366080509513e-05,
      "loss": 0.0114,
      "step": 1689
    },
    {
      "epoch": 1.2678169542385596,
      "grad_norm": 0.14046405255794525,
      "learning_rate": 3.557390309733065e-05,
      "loss": 0.0191,
      "step": 1690
    },
    {
      "epoch": 1.2685671417854465,
      "grad_norm": 0.10311423987150192,
      "learning_rate": 3.551122288304561e-05,
      "loss": 0.0175,
      "step": 1691
    },
    {
      "epoch": 1.269317329332333,
      "grad_norm": 0.12403605133295059,
      "learning_rate": 3.544856751558659e-05,
      "loss": 0.0264,
      "step": 1692
    },
    {
      "epoch": 1.27006751687922,
      "grad_norm": 0.088347427546978,
      "learning_rate": 3.538593710240139e-05,
      "loss": 0.0105,
      "step": 1693
    },
    {
      "epoch": 1.2708177044261064,
      "grad_norm": 0.1014111265540123,
      "learning_rate": 3.532333175089498e-05,
      "loss": 0.0131,
      "step": 1694
    },
    {
      "epoch": 1.2715678919729934,
      "grad_norm": 0.1430293172597885,
      "learning_rate": 3.526075156842938e-05,
      "loss": 0.0187,
      "step": 1695
    },
    {
      "epoch": 1.2723180795198799,
      "grad_norm": 0.13322706520557404,
      "learning_rate": 3.519819666232345e-05,
      "loss": 0.0152,
      "step": 1696
    },
    {
      "epoch": 1.2730682670667668,
      "grad_norm": 0.09196791052818298,
      "learning_rate": 3.5135667139852654e-05,
      "loss": 0.0143,
      "step": 1697
    },
    {
      "epoch": 1.2738184546136533,
      "grad_norm": 0.11764800548553467,
      "learning_rate": 3.507316310824902e-05,
      "loss": 0.0174,
      "step": 1698
    },
    {
      "epoch": 1.2745686421605402,
      "grad_norm": 0.0816640555858612,
      "learning_rate": 3.50106846747008e-05,
      "loss": 0.0118,
      "step": 1699
    },
    {
      "epoch": 1.275318829707427,
      "grad_norm": 0.1807868927717209,
      "learning_rate": 3.4948231946352314e-05,
      "loss": 0.0217,
      "step": 1700
    },
    {
      "epoch": 1.2760690172543137,
      "grad_norm": 0.08453043550252914,
      "learning_rate": 3.488580503030389e-05,
      "loss": 0.0161,
      "step": 1701
    },
    {
      "epoch": 1.2768192048012004,
      "grad_norm": 0.11556815356016159,
      "learning_rate": 3.482340403361151e-05,
      "loss": 0.0186,
      "step": 1702
    },
    {
      "epoch": 1.277569392348087,
      "grad_norm": 0.0988595187664032,
      "learning_rate": 3.4761029063286745e-05,
      "loss": 0.015,
      "step": 1703
    },
    {
      "epoch": 1.2783195798949738,
      "grad_norm": 0.11000659316778183,
      "learning_rate": 3.4698680226296526e-05,
      "loss": 0.0134,
      "step": 1704
    },
    {
      "epoch": 1.2790697674418605,
      "grad_norm": 0.16800323128700256,
      "learning_rate": 3.4636357629562986e-05,
      "loss": 0.0298,
      "step": 1705
    },
    {
      "epoch": 1.2798199549887472,
      "grad_norm": 0.1235019639134407,
      "learning_rate": 3.457406137996321e-05,
      "loss": 0.0222,
      "step": 1706
    },
    {
      "epoch": 1.280570142535634,
      "grad_norm": 0.09599635750055313,
      "learning_rate": 3.4511791584329154e-05,
      "loss": 0.0165,
      "step": 1707
    },
    {
      "epoch": 1.2813203300825207,
      "grad_norm": 0.10348288714885712,
      "learning_rate": 3.4449548349447394e-05,
      "loss": 0.0161,
      "step": 1708
    },
    {
      "epoch": 1.2820705176294074,
      "grad_norm": 0.11185697466135025,
      "learning_rate": 3.438733178205892e-05,
      "loss": 0.0107,
      "step": 1709
    },
    {
      "epoch": 1.282820705176294,
      "grad_norm": 0.13753622770309448,
      "learning_rate": 3.4325141988859046e-05,
      "loss": 0.0243,
      "step": 1710
    },
    {
      "epoch": 1.2835708927231808,
      "grad_norm": 0.07685783505439758,
      "learning_rate": 3.426297907649711e-05,
      "loss": 0.0114,
      "step": 1711
    },
    {
      "epoch": 1.2843210802700675,
      "grad_norm": 0.142874076962471,
      "learning_rate": 3.4200843151576414e-05,
      "loss": 0.0253,
      "step": 1712
    },
    {
      "epoch": 1.2850712678169542,
      "grad_norm": 0.12762536108493805,
      "learning_rate": 3.413873432065394e-05,
      "loss": 0.0266,
      "step": 1713
    },
    {
      "epoch": 1.285821455363841,
      "grad_norm": 0.07396502792835236,
      "learning_rate": 3.407665269024024e-05,
      "loss": 0.011,
      "step": 1714
    },
    {
      "epoch": 1.2865716429107277,
      "grad_norm": 0.33753538131713867,
      "learning_rate": 3.401459836679917e-05,
      "loss": 0.0264,
      "step": 1715
    },
    {
      "epoch": 1.2873218304576144,
      "grad_norm": 0.10064946115016937,
      "learning_rate": 3.39525714567478e-05,
      "loss": 0.0182,
      "step": 1716
    },
    {
      "epoch": 1.288072018004501,
      "grad_norm": 0.16794368624687195,
      "learning_rate": 3.389057206645614e-05,
      "loss": 0.0284,
      "step": 1717
    },
    {
      "epoch": 1.2888222055513878,
      "grad_norm": 0.15257765352725983,
      "learning_rate": 3.382860030224708e-05,
      "loss": 0.023,
      "step": 1718
    },
    {
      "epoch": 1.2895723930982745,
      "grad_norm": 0.12148189544677734,
      "learning_rate": 3.3766656270396074e-05,
      "loss": 0.0232,
      "step": 1719
    },
    {
      "epoch": 1.2903225806451613,
      "grad_norm": 0.09580284357070923,
      "learning_rate": 3.3704740077131036e-05,
      "loss": 0.0144,
      "step": 1720
    },
    {
      "epoch": 1.291072768192048,
      "grad_norm": 0.08760684728622437,
      "learning_rate": 3.3642851828632155e-05,
      "loss": 0.0126,
      "step": 1721
    },
    {
      "epoch": 1.2918229557389347,
      "grad_norm": 0.11275685578584671,
      "learning_rate": 3.3580991631031656e-05,
      "loss": 0.0157,
      "step": 1722
    },
    {
      "epoch": 1.2925731432858214,
      "grad_norm": 0.09197841584682465,
      "learning_rate": 3.3519159590413715e-05,
      "loss": 0.0127,
      "step": 1723
    },
    {
      "epoch": 1.2933233308327081,
      "grad_norm": 0.11137830466032028,
      "learning_rate": 3.345735581281417e-05,
      "loss": 0.019,
      "step": 1724
    },
    {
      "epoch": 1.2940735183795948,
      "grad_norm": 0.09153483808040619,
      "learning_rate": 3.339558040422042e-05,
      "loss": 0.0128,
      "step": 1725
    },
    {
      "epoch": 1.2948237059264815,
      "grad_norm": 0.10245629400014877,
      "learning_rate": 3.333383347057123e-05,
      "loss": 0.0195,
      "step": 1726
    },
    {
      "epoch": 1.2955738934733683,
      "grad_norm": 0.18266040086746216,
      "learning_rate": 3.3272115117756476e-05,
      "loss": 0.0235,
      "step": 1727
    },
    {
      "epoch": 1.296324081020255,
      "grad_norm": 0.15308219194412231,
      "learning_rate": 3.3210425451617074e-05,
      "loss": 0.0152,
      "step": 1728
    },
    {
      "epoch": 1.2970742685671417,
      "grad_norm": 0.0983470156788826,
      "learning_rate": 3.314876457794474e-05,
      "loss": 0.0138,
      "step": 1729
    },
    {
      "epoch": 1.2978244561140286,
      "grad_norm": 0.1688724309206009,
      "learning_rate": 3.3087132602481774e-05,
      "loss": 0.0165,
      "step": 1730
    },
    {
      "epoch": 1.2985746436609151,
      "grad_norm": 0.08947484195232391,
      "learning_rate": 3.302552963092096e-05,
      "loss": 0.0138,
      "step": 1731
    },
    {
      "epoch": 1.299324831207802,
      "grad_norm": 0.14829127490520477,
      "learning_rate": 3.296395576890532e-05,
      "loss": 0.0171,
      "step": 1732
    },
    {
      "epoch": 1.3000750187546886,
      "grad_norm": 0.12239532917737961,
      "learning_rate": 3.290241112202797e-05,
      "loss": 0.0201,
      "step": 1733
    },
    {
      "epoch": 1.3008252063015755,
      "grad_norm": 0.16367493569850922,
      "learning_rate": 3.284089579583192e-05,
      "loss": 0.0195,
      "step": 1734
    },
    {
      "epoch": 1.301575393848462,
      "grad_norm": 0.11469226330518723,
      "learning_rate": 3.2779409895809886e-05,
      "loss": 0.0161,
      "step": 1735
    },
    {
      "epoch": 1.302325581395349,
      "grad_norm": 0.10691020637750626,
      "learning_rate": 3.2717953527404155e-05,
      "loss": 0.0128,
      "step": 1736
    },
    {
      "epoch": 1.3030757689422354,
      "grad_norm": 0.1380341500043869,
      "learning_rate": 3.265652679600631e-05,
      "loss": 0.022,
      "step": 1737
    },
    {
      "epoch": 1.3038259564891224,
      "grad_norm": 0.10156739503145218,
      "learning_rate": 3.25951298069572e-05,
      "loss": 0.0173,
      "step": 1738
    },
    {
      "epoch": 1.304576144036009,
      "grad_norm": 0.0986672043800354,
      "learning_rate": 3.253376266554655e-05,
      "loss": 0.016,
      "step": 1739
    },
    {
      "epoch": 1.3053263315828958,
      "grad_norm": 0.14217688143253326,
      "learning_rate": 3.247242547701301e-05,
      "loss": 0.02,
      "step": 1740
    },
    {
      "epoch": 1.3060765191297825,
      "grad_norm": 0.15207204222679138,
      "learning_rate": 3.241111834654382e-05,
      "loss": 0.0253,
      "step": 1741
    },
    {
      "epoch": 1.3068267066766692,
      "grad_norm": 0.13661247491836548,
      "learning_rate": 3.234984137927464e-05,
      "loss": 0.0193,
      "step": 1742
    },
    {
      "epoch": 1.307576894223556,
      "grad_norm": 0.10947258770465851,
      "learning_rate": 3.228859468028946e-05,
      "loss": 0.0199,
      "step": 1743
    },
    {
      "epoch": 1.3083270817704427,
      "grad_norm": 0.2023872286081314,
      "learning_rate": 3.222737835462034e-05,
      "loss": 0.027,
      "step": 1744
    },
    {
      "epoch": 1.3090772693173294,
      "grad_norm": 0.0802428126335144,
      "learning_rate": 3.216619250724724e-05,
      "loss": 0.012,
      "step": 1745
    },
    {
      "epoch": 1.309827456864216,
      "grad_norm": 0.11808689683675766,
      "learning_rate": 3.2105037243097866e-05,
      "loss": 0.0168,
      "step": 1746
    },
    {
      "epoch": 1.3105776444111028,
      "grad_norm": 0.18175119161605835,
      "learning_rate": 3.2043912667047465e-05,
      "loss": 0.0261,
      "step": 1747
    },
    {
      "epoch": 1.3113278319579895,
      "grad_norm": 0.08211642503738403,
      "learning_rate": 3.198281888391869e-05,
      "loss": 0.0104,
      "step": 1748
    },
    {
      "epoch": 1.3120780195048762,
      "grad_norm": 0.14124037325382233,
      "learning_rate": 3.192175599848133e-05,
      "loss": 0.0222,
      "step": 1749
    },
    {
      "epoch": 1.312828207051763,
      "grad_norm": 0.11715236306190491,
      "learning_rate": 3.1860724115452234e-05,
      "loss": 0.014,
      "step": 1750
    },
    {
      "epoch": 1.3135783945986497,
      "grad_norm": 0.07071872800588608,
      "learning_rate": 3.179972333949509e-05,
      "loss": 0.012,
      "step": 1751
    },
    {
      "epoch": 1.3143285821455364,
      "grad_norm": 0.0900421068072319,
      "learning_rate": 3.173875377522019e-05,
      "loss": 0.0142,
      "step": 1752
    },
    {
      "epoch": 1.315078769692423,
      "grad_norm": 0.18912558257579803,
      "learning_rate": 3.167781552718435e-05,
      "loss": 0.0248,
      "step": 1753
    },
    {
      "epoch": 1.3158289572393098,
      "grad_norm": 0.09656878560781479,
      "learning_rate": 3.161690869989068e-05,
      "loss": 0.0149,
      "step": 1754
    },
    {
      "epoch": 1.3165791447861965,
      "grad_norm": 0.10428894311189651,
      "learning_rate": 3.155603339778837e-05,
      "loss": 0.0132,
      "step": 1755
    },
    {
      "epoch": 1.3173293323330832,
      "grad_norm": 0.15280795097351074,
      "learning_rate": 3.149518972527257e-05,
      "loss": 0.0189,
      "step": 1756
    },
    {
      "epoch": 1.31807951987997,
      "grad_norm": 0.1433146446943283,
      "learning_rate": 3.1434377786684197e-05,
      "loss": 0.0233,
      "step": 1757
    },
    {
      "epoch": 1.3188297074268567,
      "grad_norm": 0.10560217499732971,
      "learning_rate": 3.137359768630972e-05,
      "loss": 0.0184,
      "step": 1758
    },
    {
      "epoch": 1.3195798949737434,
      "grad_norm": 0.14866486191749573,
      "learning_rate": 3.131284952838106e-05,
      "loss": 0.0131,
      "step": 1759
    },
    {
      "epoch": 1.32033008252063,
      "grad_norm": 0.11577482521533966,
      "learning_rate": 3.125213341707528e-05,
      "loss": 0.0203,
      "step": 1760
    },
    {
      "epoch": 1.3210802700675168,
      "grad_norm": 0.13666844367980957,
      "learning_rate": 3.1191449456514575e-05,
      "loss": 0.0202,
      "step": 1761
    },
    {
      "epoch": 1.3218304576144035,
      "grad_norm": 0.09610434621572495,
      "learning_rate": 3.113079775076593e-05,
      "loss": 0.0126,
      "step": 1762
    },
    {
      "epoch": 1.3225806451612903,
      "grad_norm": 0.13390542566776276,
      "learning_rate": 3.107017840384107e-05,
      "loss": 0.0155,
      "step": 1763
    },
    {
      "epoch": 1.323330832708177,
      "grad_norm": 0.09720296412706375,
      "learning_rate": 3.100959151969619e-05,
      "loss": 0.0157,
      "step": 1764
    },
    {
      "epoch": 1.3240810202550637,
      "grad_norm": 0.16918721795082092,
      "learning_rate": 3.0949037202231826e-05,
      "loss": 0.024,
      "step": 1765
    },
    {
      "epoch": 1.3248312078019504,
      "grad_norm": 0.09024802595376968,
      "learning_rate": 3.08885155552927e-05,
      "loss": 0.0121,
      "step": 1766
    },
    {
      "epoch": 1.3255813953488373,
      "grad_norm": 0.0722404420375824,
      "learning_rate": 3.082802668266743e-05,
      "loss": 0.01,
      "step": 1767
    },
    {
      "epoch": 1.3263315828957238,
      "grad_norm": 0.11001674830913544,
      "learning_rate": 3.076757068808852e-05,
      "loss": 0.0154,
      "step": 1768
    },
    {
      "epoch": 1.3270817704426108,
      "grad_norm": 0.2680615782737732,
      "learning_rate": 3.070714767523203e-05,
      "loss": 0.0486,
      "step": 1769
    },
    {
      "epoch": 1.3278319579894973,
      "grad_norm": 0.11515659838914871,
      "learning_rate": 3.0646757747717475e-05,
      "loss": 0.0131,
      "step": 1770
    },
    {
      "epoch": 1.3285821455363842,
      "grad_norm": 0.11640623956918716,
      "learning_rate": 3.0586401009107636e-05,
      "loss": 0.0153,
      "step": 1771
    },
    {
      "epoch": 1.3293323330832707,
      "grad_norm": 0.22977732121944427,
      "learning_rate": 3.0526077562908386e-05,
      "loss": 0.0315,
      "step": 1772
    },
    {
      "epoch": 1.3300825206301576,
      "grad_norm": 0.09609303623437881,
      "learning_rate": 3.0465787512568466e-05,
      "loss": 0.0137,
      "step": 1773
    },
    {
      "epoch": 1.3308327081770441,
      "grad_norm": 0.10839638113975525,
      "learning_rate": 3.040553096147942e-05,
      "loss": 0.0203,
      "step": 1774
    },
    {
      "epoch": 1.331582895723931,
      "grad_norm": 0.12790155410766602,
      "learning_rate": 3.0345308012975255e-05,
      "loss": 0.0195,
      "step": 1775
    },
    {
      "epoch": 1.3323330832708178,
      "grad_norm": 0.08059082180261612,
      "learning_rate": 3.0285118770332428e-05,
      "loss": 0.0101,
      "step": 1776
    },
    {
      "epoch": 1.3330832708177045,
      "grad_norm": 0.1413569152355194,
      "learning_rate": 3.022496333676954e-05,
      "loss": 0.0194,
      "step": 1777
    },
    {
      "epoch": 1.3338334583645912,
      "grad_norm": 0.10592398792505264,
      "learning_rate": 3.0164841815447263e-05,
      "loss": 0.0156,
      "step": 1778
    },
    {
      "epoch": 1.334583645911478,
      "grad_norm": 0.15778115391731262,
      "learning_rate": 3.0104754309468066e-05,
      "loss": 0.0194,
      "step": 1779
    },
    {
      "epoch": 1.3353338334583646,
      "grad_norm": 0.2557612359523773,
      "learning_rate": 3.00447009218761e-05,
      "loss": 0.03,
      "step": 1780
    },
    {
      "epoch": 1.3360840210052514,
      "grad_norm": 0.1358753889799118,
      "learning_rate": 2.9984681755657017e-05,
      "loss": 0.0197,
      "step": 1781
    },
    {
      "epoch": 1.336834208552138,
      "grad_norm": 0.11550354212522507,
      "learning_rate": 2.9924696913737792e-05,
      "loss": 0.0217,
      "step": 1782
    },
    {
      "epoch": 1.3375843960990248,
      "grad_norm": 0.15435385704040527,
      "learning_rate": 2.986474649898651e-05,
      "loss": 0.021,
      "step": 1783
    },
    {
      "epoch": 1.3383345836459115,
      "grad_norm": 0.08404883742332458,
      "learning_rate": 2.9804830614212242e-05,
      "loss": 0.0146,
      "step": 1784
    },
    {
      "epoch": 1.3390847711927982,
      "grad_norm": 0.122297503054142,
      "learning_rate": 2.9744949362164798e-05,
      "loss": 0.0157,
      "step": 1785
    },
    {
      "epoch": 1.339834958739685,
      "grad_norm": 0.08815350383520126,
      "learning_rate": 2.9685102845534658e-05,
      "loss": 0.0126,
      "step": 1786
    },
    {
      "epoch": 1.3405851462865717,
      "grad_norm": 0.1538134217262268,
      "learning_rate": 2.9625291166952702e-05,
      "loss": 0.0319,
      "step": 1787
    },
    {
      "epoch": 1.3413353338334584,
      "grad_norm": 0.12486133724451065,
      "learning_rate": 2.956551442899005e-05,
      "loss": 0.0216,
      "step": 1788
    },
    {
      "epoch": 1.342085521380345,
      "grad_norm": 0.12948335707187653,
      "learning_rate": 2.9505772734157948e-05,
      "loss": 0.0187,
      "step": 1789
    },
    {
      "epoch": 1.3428357089272318,
      "grad_norm": 0.09788129478693008,
      "learning_rate": 2.9446066184907495e-05,
      "loss": 0.0125,
      "step": 1790
    },
    {
      "epoch": 1.3435858964741185,
      "grad_norm": 0.1548995077610016,
      "learning_rate": 2.9386394883629565e-05,
      "loss": 0.0227,
      "step": 1791
    },
    {
      "epoch": 1.3443360840210052,
      "grad_norm": 0.11524860560894012,
      "learning_rate": 2.932675893265454e-05,
      "loss": 0.0174,
      "step": 1792
    },
    {
      "epoch": 1.345086271567892,
      "grad_norm": 0.1371203064918518,
      "learning_rate": 2.926715843425223e-05,
      "loss": 0.0247,
      "step": 1793
    },
    {
      "epoch": 1.3458364591147787,
      "grad_norm": 0.10300806164741516,
      "learning_rate": 2.9207593490631592e-05,
      "loss": 0.0173,
      "step": 1794
    },
    {
      "epoch": 1.3465866466616654,
      "grad_norm": 0.17150884866714478,
      "learning_rate": 2.914806420394064e-05,
      "loss": 0.0284,
      "step": 1795
    },
    {
      "epoch": 1.347336834208552,
      "grad_norm": 0.13951390981674194,
      "learning_rate": 2.908857067626629e-05,
      "loss": 0.0158,
      "step": 1796
    },
    {
      "epoch": 1.3480870217554388,
      "grad_norm": 0.12770432233810425,
      "learning_rate": 2.902911300963403e-05,
      "loss": 0.0231,
      "step": 1797
    },
    {
      "epoch": 1.3488372093023255,
      "grad_norm": 0.09566380083560944,
      "learning_rate": 2.8969691306007918e-05,
      "loss": 0.0135,
      "step": 1798
    },
    {
      "epoch": 1.3495873968492123,
      "grad_norm": 0.09345205873250961,
      "learning_rate": 2.891030566729032e-05,
      "loss": 0.018,
      "step": 1799
    },
    {
      "epoch": 1.350337584396099,
      "grad_norm": 0.11738814413547516,
      "learning_rate": 2.8850956195321795e-05,
      "loss": 0.0178,
      "step": 1800
    },
    {
      "epoch": 1.350337584396099,
      "eval_loss": 0.0302781630307436,
      "eval_runtime": 5.1365,
      "eval_samples_per_second": 10.513,
      "eval_steps_per_second": 2.726,
      "step": 1800
    },
    {
      "epoch": 1.3510877719429857,
      "grad_norm": 0.16350629925727844,
      "learning_rate": 2.8791642991880784e-05,
      "loss": 0.0222,
      "step": 1801
    },
    {
      "epoch": 1.3518379594898724,
      "grad_norm": 0.09744306653738022,
      "learning_rate": 2.873236615868362e-05,
      "loss": 0.0184,
      "step": 1802
    },
    {
      "epoch": 1.3525881470367591,
      "grad_norm": 0.14381620287895203,
      "learning_rate": 2.8673125797384243e-05,
      "loss": 0.0166,
      "step": 1803
    },
    {
      "epoch": 1.3533383345836458,
      "grad_norm": 0.09115853160619736,
      "learning_rate": 2.8613922009574024e-05,
      "loss": 0.0122,
      "step": 1804
    },
    {
      "epoch": 1.3540885221305325,
      "grad_norm": 0.1326061189174652,
      "learning_rate": 2.8554754896781656e-05,
      "loss": 0.0187,
      "step": 1805
    },
    {
      "epoch": 1.3548387096774195,
      "grad_norm": 0.0877179428935051,
      "learning_rate": 2.8495624560472866e-05,
      "loss": 0.0111,
      "step": 1806
    },
    {
      "epoch": 1.355588897224306,
      "grad_norm": 0.13962623476982117,
      "learning_rate": 2.843653110205039e-05,
      "loss": 0.0191,
      "step": 1807
    },
    {
      "epoch": 1.356339084771193,
      "grad_norm": 0.11053589731454849,
      "learning_rate": 2.8377474622853683e-05,
      "loss": 0.0168,
      "step": 1808
    },
    {
      "epoch": 1.3570892723180794,
      "grad_norm": 0.09899614006280899,
      "learning_rate": 2.8318455224158786e-05,
      "loss": 0.0142,
      "step": 1809
    },
    {
      "epoch": 1.3578394598649663,
      "grad_norm": 0.11597689986228943,
      "learning_rate": 2.8259473007178163e-05,
      "loss": 0.0149,
      "step": 1810
    },
    {
      "epoch": 1.3585896474118528,
      "grad_norm": 0.08869005739688873,
      "learning_rate": 2.8200528073060507e-05,
      "loss": 0.0128,
      "step": 1811
    },
    {
      "epoch": 1.3593398349587398,
      "grad_norm": 0.12468276917934418,
      "learning_rate": 2.814162052289058e-05,
      "loss": 0.0171,
      "step": 1812
    },
    {
      "epoch": 1.3600900225056263,
      "grad_norm": 0.06973183900117874,
      "learning_rate": 2.8082750457689033e-05,
      "loss": 0.0079,
      "step": 1813
    },
    {
      "epoch": 1.3608402100525132,
      "grad_norm": 0.3414054811000824,
      "learning_rate": 2.8023917978412207e-05,
      "loss": 0.0179,
      "step": 1814
    },
    {
      "epoch": 1.3615903975994,
      "grad_norm": 0.10484994947910309,
      "learning_rate": 2.7965123185952023e-05,
      "loss": 0.0163,
      "step": 1815
    },
    {
      "epoch": 1.3623405851462866,
      "grad_norm": 0.1568952351808548,
      "learning_rate": 2.7906366181135775e-05,
      "loss": 0.0205,
      "step": 1816
    },
    {
      "epoch": 1.3630907726931734,
      "grad_norm": 0.14470714330673218,
      "learning_rate": 2.7847647064725924e-05,
      "loss": 0.0161,
      "step": 1817
    },
    {
      "epoch": 1.36384096024006,
      "grad_norm": 0.08876904845237732,
      "learning_rate": 2.778896593741999e-05,
      "loss": 0.0091,
      "step": 1818
    },
    {
      "epoch": 1.3645911477869468,
      "grad_norm": 0.07300567626953125,
      "learning_rate": 2.77303228998503e-05,
      "loss": 0.0091,
      "step": 1819
    },
    {
      "epoch": 1.3653413353338335,
      "grad_norm": 0.20413988828659058,
      "learning_rate": 2.7671718052583908e-05,
      "loss": 0.0339,
      "step": 1820
    },
    {
      "epoch": 1.3660915228807202,
      "grad_norm": 0.1549023687839508,
      "learning_rate": 2.7613151496122347e-05,
      "loss": 0.0264,
      "step": 1821
    },
    {
      "epoch": 1.366841710427607,
      "grad_norm": 0.1032576784491539,
      "learning_rate": 2.7554623330901524e-05,
      "loss": 0.0138,
      "step": 1822
    },
    {
      "epoch": 1.3675918979744937,
      "grad_norm": 0.19302760064601898,
      "learning_rate": 2.749613365729141e-05,
      "loss": 0.0302,
      "step": 1823
    },
    {
      "epoch": 1.3683420855213804,
      "grad_norm": 0.13430748879909515,
      "learning_rate": 2.7437682575596104e-05,
      "loss": 0.0225,
      "step": 1824
    },
    {
      "epoch": 1.369092273068267,
      "grad_norm": 0.11612657457590103,
      "learning_rate": 2.7379270186053428e-05,
      "loss": 0.0167,
      "step": 1825
    },
    {
      "epoch": 1.3698424606151538,
      "grad_norm": 0.09877978265285492,
      "learning_rate": 2.7320896588834903e-05,
      "loss": 0.0144,
      "step": 1826
    },
    {
      "epoch": 1.3705926481620405,
      "grad_norm": 0.12179384380578995,
      "learning_rate": 2.7262561884045457e-05,
      "loss": 0.0138,
      "step": 1827
    },
    {
      "epoch": 1.3713428357089272,
      "grad_norm": 0.08559141308069229,
      "learning_rate": 2.720426617172339e-05,
      "loss": 0.0124,
      "step": 1828
    },
    {
      "epoch": 1.372093023255814,
      "grad_norm": 0.16383309662342072,
      "learning_rate": 2.71460095518401e-05,
      "loss": 0.0227,
      "step": 1829
    },
    {
      "epoch": 1.3728432108027007,
      "grad_norm": 0.14575624465942383,
      "learning_rate": 2.708779212429996e-05,
      "loss": 0.0218,
      "step": 1830
    },
    {
      "epoch": 1.3735933983495874,
      "grad_norm": 0.08648881316184998,
      "learning_rate": 2.702961398894014e-05,
      "loss": 0.013,
      "step": 1831
    },
    {
      "epoch": 1.374343585896474,
      "grad_norm": 0.21686382591724396,
      "learning_rate": 2.6971475245530375e-05,
      "loss": 0.0299,
      "step": 1832
    },
    {
      "epoch": 1.3750937734433608,
      "grad_norm": 0.13174933195114136,
      "learning_rate": 2.6913375993772915e-05,
      "loss": 0.0183,
      "step": 1833
    },
    {
      "epoch": 1.3758439609902475,
      "grad_norm": 0.12637682259082794,
      "learning_rate": 2.6855316333302237e-05,
      "loss": 0.0179,
      "step": 1834
    },
    {
      "epoch": 1.3765941485371342,
      "grad_norm": 0.1379723697900772,
      "learning_rate": 2.6797296363684977e-05,
      "loss": 0.0157,
      "step": 1835
    },
    {
      "epoch": 1.377344336084021,
      "grad_norm": 0.1410985141992569,
      "learning_rate": 2.6739316184419622e-05,
      "loss": 0.0162,
      "step": 1836
    },
    {
      "epoch": 1.3780945236309077,
      "grad_norm": 0.12312553822994232,
      "learning_rate": 2.6681375894936472e-05,
      "loss": 0.016,
      "step": 1837
    },
    {
      "epoch": 1.3788447111777944,
      "grad_norm": 0.1267862617969513,
      "learning_rate": 2.662347559459746e-05,
      "loss": 0.0192,
      "step": 1838
    },
    {
      "epoch": 1.379594898724681,
      "grad_norm": 0.07263024896383286,
      "learning_rate": 2.6565615382695896e-05,
      "loss": 0.0117,
      "step": 1839
    },
    {
      "epoch": 1.3803450862715678,
      "grad_norm": 0.0848407968878746,
      "learning_rate": 2.6507795358456307e-05,
      "loss": 0.0157,
      "step": 1840
    },
    {
      "epoch": 1.3810952738184545,
      "grad_norm": 0.09617772698402405,
      "learning_rate": 2.6450015621034362e-05,
      "loss": 0.0133,
      "step": 1841
    },
    {
      "epoch": 1.3818454613653413,
      "grad_norm": 0.09032897651195526,
      "learning_rate": 2.6392276269516613e-05,
      "loss": 0.0114,
      "step": 1842
    },
    {
      "epoch": 1.3825956489122282,
      "grad_norm": 0.1260712742805481,
      "learning_rate": 2.63345774029204e-05,
      "loss": 0.0219,
      "step": 1843
    },
    {
      "epoch": 1.3833458364591147,
      "grad_norm": 0.1832832247018814,
      "learning_rate": 2.6276919120193543e-05,
      "loss": 0.0275,
      "step": 1844
    },
    {
      "epoch": 1.3840960240060016,
      "grad_norm": 0.1576533019542694,
      "learning_rate": 2.621930152021434e-05,
      "loss": 0.0249,
      "step": 1845
    },
    {
      "epoch": 1.3848462115528881,
      "grad_norm": 0.15020745992660522,
      "learning_rate": 2.6161724701791306e-05,
      "loss": 0.017,
      "step": 1846
    },
    {
      "epoch": 1.385596399099775,
      "grad_norm": 0.11638973653316498,
      "learning_rate": 2.6104188763663018e-05,
      "loss": 0.0142,
      "step": 1847
    },
    {
      "epoch": 1.3863465866466615,
      "grad_norm": 0.12805978953838348,
      "learning_rate": 2.604669380449795e-05,
      "loss": 0.0151,
      "step": 1848
    },
    {
      "epoch": 1.3870967741935485,
      "grad_norm": 0.12945251166820526,
      "learning_rate": 2.598923992289427e-05,
      "loss": 0.0153,
      "step": 1849
    },
    {
      "epoch": 1.387846961740435,
      "grad_norm": 0.07249733805656433,
      "learning_rate": 2.5931827217379746e-05,
      "loss": 0.0098,
      "step": 1850
    },
    {
      "epoch": 1.388597149287322,
      "grad_norm": 0.08274883031845093,
      "learning_rate": 2.5874455786411505e-05,
      "loss": 0.0126,
      "step": 1851
    },
    {
      "epoch": 1.3893473368342086,
      "grad_norm": 0.0918225646018982,
      "learning_rate": 2.5817125728375912e-05,
      "loss": 0.0118,
      "step": 1852
    },
    {
      "epoch": 1.3900975243810954,
      "grad_norm": 0.08083860576152802,
      "learning_rate": 2.5759837141588362e-05,
      "loss": 0.011,
      "step": 1853
    },
    {
      "epoch": 1.390847711927982,
      "grad_norm": 0.2597929537296295,
      "learning_rate": 2.5702590124293147e-05,
      "loss": 0.0213,
      "step": 1854
    },
    {
      "epoch": 1.3915978994748688,
      "grad_norm": 0.11949342489242554,
      "learning_rate": 2.5645384774663262e-05,
      "loss": 0.0195,
      "step": 1855
    },
    {
      "epoch": 1.3923480870217555,
      "grad_norm": 0.054844655096530914,
      "learning_rate": 2.5588221190800264e-05,
      "loss": 0.0106,
      "step": 1856
    },
    {
      "epoch": 1.3930982745686422,
      "grad_norm": 0.2637576460838318,
      "learning_rate": 2.5531099470734038e-05,
      "loss": 0.0293,
      "step": 1857
    },
    {
      "epoch": 1.393848462115529,
      "grad_norm": 0.17199398577213287,
      "learning_rate": 2.5474019712422724e-05,
      "loss": 0.0171,
      "step": 1858
    },
    {
      "epoch": 1.3945986496624156,
      "grad_norm": 0.16428735852241516,
      "learning_rate": 2.541698201375249e-05,
      "loss": 0.0173,
      "step": 1859
    },
    {
      "epoch": 1.3953488372093024,
      "grad_norm": 0.10029322654008865,
      "learning_rate": 2.5359986472537373e-05,
      "loss": 0.0118,
      "step": 1860
    },
    {
      "epoch": 1.396099024756189,
      "grad_norm": 0.10694640129804611,
      "learning_rate": 2.530303318651913e-05,
      "loss": 0.0144,
      "step": 1861
    },
    {
      "epoch": 1.3968492123030758,
      "grad_norm": 0.12824860215187073,
      "learning_rate": 2.5246122253366998e-05,
      "loss": 0.0137,
      "step": 1862
    },
    {
      "epoch": 1.3975993998499625,
      "grad_norm": 0.10325340926647186,
      "learning_rate": 2.5189253770677644e-05,
      "loss": 0.0112,
      "step": 1863
    },
    {
      "epoch": 1.3983495873968492,
      "grad_norm": 0.09274782985448837,
      "learning_rate": 2.5132427835974926e-05,
      "loss": 0.0129,
      "step": 1864
    },
    {
      "epoch": 1.399099774943736,
      "grad_norm": 0.06569540500640869,
      "learning_rate": 2.507564454670971e-05,
      "loss": 0.0068,
      "step": 1865
    },
    {
      "epoch": 1.3998499624906227,
      "grad_norm": 0.07178516685962677,
      "learning_rate": 2.5018904000259757e-05,
      "loss": 0.0126,
      "step": 1866
    },
    {
      "epoch": 1.4006001500375094,
      "grad_norm": 0.09361020475625992,
      "learning_rate": 2.4962206293929512e-05,
      "loss": 0.0111,
      "step": 1867
    },
    {
      "epoch": 1.401350337584396,
      "grad_norm": 0.09147719293832779,
      "learning_rate": 2.490555152494996e-05,
      "loss": 0.0093,
      "step": 1868
    },
    {
      "epoch": 1.4021005251312828,
      "grad_norm": 0.12089736014604568,
      "learning_rate": 2.4848939790478463e-05,
      "loss": 0.012,
      "step": 1869
    },
    {
      "epoch": 1.4028507126781695,
      "grad_norm": 0.07719265669584274,
      "learning_rate": 2.4792371187598544e-05,
      "loss": 0.0102,
      "step": 1870
    },
    {
      "epoch": 1.4036009002250562,
      "grad_norm": 0.07259789109230042,
      "learning_rate": 2.4735845813319804e-05,
      "loss": 0.0105,
      "step": 1871
    },
    {
      "epoch": 1.404351087771943,
      "grad_norm": 0.1539996713399887,
      "learning_rate": 2.4679363764577683e-05,
      "loss": 0.0188,
      "step": 1872
    },
    {
      "epoch": 1.4051012753188297,
      "grad_norm": 0.10606228560209274,
      "learning_rate": 2.462292513823336e-05,
      "loss": 0.0168,
      "step": 1873
    },
    {
      "epoch": 1.4058514628657164,
      "grad_norm": 0.09349866211414337,
      "learning_rate": 2.4566530031073486e-05,
      "loss": 0.013,
      "step": 1874
    },
    {
      "epoch": 1.406601650412603,
      "grad_norm": 0.20843681693077087,
      "learning_rate": 2.451017853981013e-05,
      "loss": 0.0161,
      "step": 1875
    },
    {
      "epoch": 1.4073518379594898,
      "grad_norm": 0.12823422253131866,
      "learning_rate": 2.4453870761080554e-05,
      "loss": 0.0222,
      "step": 1876
    },
    {
      "epoch": 1.4081020255063765,
      "grad_norm": 0.14354470372200012,
      "learning_rate": 2.4397606791447052e-05,
      "loss": 0.0188,
      "step": 1877
    },
    {
      "epoch": 1.4088522130532632,
      "grad_norm": 0.1241222470998764,
      "learning_rate": 2.4341386727396793e-05,
      "loss": 0.0204,
      "step": 1878
    },
    {
      "epoch": 1.40960240060015,
      "grad_norm": 0.11772765964269638,
      "learning_rate": 2.4285210665341646e-05,
      "loss": 0.0146,
      "step": 1879
    },
    {
      "epoch": 1.4103525881470367,
      "grad_norm": 0.14254790544509888,
      "learning_rate": 2.422907870161803e-05,
      "loss": 0.0168,
      "step": 1880
    },
    {
      "epoch": 1.4111027756939234,
      "grad_norm": 0.22304561734199524,
      "learning_rate": 2.4172990932486733e-05,
      "loss": 0.0146,
      "step": 1881
    },
    {
      "epoch": 1.4118529632408103,
      "grad_norm": 0.16511446237564087,
      "learning_rate": 2.4116947454132782e-05,
      "loss": 0.0187,
      "step": 1882
    },
    {
      "epoch": 1.4126031507876968,
      "grad_norm": 0.12960758805274963,
      "learning_rate": 2.4060948362665176e-05,
      "loss": 0.0145,
      "step": 1883
    },
    {
      "epoch": 1.4133533383345838,
      "grad_norm": 0.11755431443452835,
      "learning_rate": 2.4004993754116867e-05,
      "loss": 0.0171,
      "step": 1884
    },
    {
      "epoch": 1.4141035258814703,
      "grad_norm": 0.08797797560691833,
      "learning_rate": 2.39490837244445e-05,
      "loss": 0.011,
      "step": 1885
    },
    {
      "epoch": 1.4148537134283572,
      "grad_norm": 0.08972885459661484,
      "learning_rate": 2.389321836952828e-05,
      "loss": 0.012,
      "step": 1886
    },
    {
      "epoch": 1.4156039009752437,
      "grad_norm": 0.1484125852584839,
      "learning_rate": 2.383739778517176e-05,
      "loss": 0.0164,
      "step": 1887
    },
    {
      "epoch": 1.4163540885221306,
      "grad_norm": 0.20713378489017487,
      "learning_rate": 2.3781622067101767e-05,
      "loss": 0.0283,
      "step": 1888
    },
    {
      "epoch": 1.4171042760690171,
      "grad_norm": 0.13764242827892303,
      "learning_rate": 2.372589131096816e-05,
      "loss": 0.0225,
      "step": 1889
    },
    {
      "epoch": 1.417854463615904,
      "grad_norm": 0.12255063652992249,
      "learning_rate": 2.36702056123437e-05,
      "loss": 0.0153,
      "step": 1890
    },
    {
      "epoch": 1.4186046511627908,
      "grad_norm": 0.09976071864366531,
      "learning_rate": 2.3614565066723892e-05,
      "loss": 0.0147,
      "step": 1891
    },
    {
      "epoch": 1.4193548387096775,
      "grad_norm": 0.14297650754451752,
      "learning_rate": 2.355896976952674e-05,
      "loss": 0.0195,
      "step": 1892
    },
    {
      "epoch": 1.4201050262565642,
      "grad_norm": 0.1632770597934723,
      "learning_rate": 2.350341981609276e-05,
      "loss": 0.0217,
      "step": 1893
    },
    {
      "epoch": 1.420855213803451,
      "grad_norm": 0.13608597218990326,
      "learning_rate": 2.344791530168465e-05,
      "loss": 0.0178,
      "step": 1894
    },
    {
      "epoch": 1.4216054013503376,
      "grad_norm": 0.09987060725688934,
      "learning_rate": 2.339245632148715e-05,
      "loss": 0.0168,
      "step": 1895
    },
    {
      "epoch": 1.4223555888972244,
      "grad_norm": 0.15993699431419373,
      "learning_rate": 2.3337042970606965e-05,
      "loss": 0.0261,
      "step": 1896
    },
    {
      "epoch": 1.423105776444111,
      "grad_norm": 0.11352405697107315,
      "learning_rate": 2.3281675344072545e-05,
      "loss": 0.0166,
      "step": 1897
    },
    {
      "epoch": 1.4238559639909978,
      "grad_norm": 0.09080108255147934,
      "learning_rate": 2.3226353536833907e-05,
      "loss": 0.0145,
      "step": 1898
    },
    {
      "epoch": 1.4246061515378845,
      "grad_norm": 0.11767710000276566,
      "learning_rate": 2.317107764376253e-05,
      "loss": 0.0185,
      "step": 1899
    },
    {
      "epoch": 1.4253563390847712,
      "grad_norm": 0.10477675497531891,
      "learning_rate": 2.3115847759651082e-05,
      "loss": 0.0147,
      "step": 1900
    },
    {
      "epoch": 1.426106526631658,
      "grad_norm": 0.1232522651553154,
      "learning_rate": 2.3060663979213404e-05,
      "loss": 0.0192,
      "step": 1901
    },
    {
      "epoch": 1.4268567141785446,
      "grad_norm": 0.1216619461774826,
      "learning_rate": 2.300552639708423e-05,
      "loss": 0.0199,
      "step": 1902
    },
    {
      "epoch": 1.4276069017254314,
      "grad_norm": 0.0820419117808342,
      "learning_rate": 2.2950435107819124e-05,
      "loss": 0.0103,
      "step": 1903
    },
    {
      "epoch": 1.428357089272318,
      "grad_norm": 0.13588947057724,
      "learning_rate": 2.2895390205894164e-05,
      "loss": 0.0185,
      "step": 1904
    },
    {
      "epoch": 1.4291072768192048,
      "grad_norm": 0.30323174595832825,
      "learning_rate": 2.2840391785705967e-05,
      "loss": 0.0401,
      "step": 1905
    },
    {
      "epoch": 1.4298574643660915,
      "grad_norm": 0.15225709974765778,
      "learning_rate": 2.278543994157139e-05,
      "loss": 0.0165,
      "step": 1906
    },
    {
      "epoch": 1.4306076519129782,
      "grad_norm": 0.10808411240577698,
      "learning_rate": 2.2730534767727483e-05,
      "loss": 0.0143,
      "step": 1907
    },
    {
      "epoch": 1.431357839459865,
      "grad_norm": 0.1031549796462059,
      "learning_rate": 2.267567635833116e-05,
      "loss": 0.0224,
      "step": 1908
    },
    {
      "epoch": 1.4321080270067517,
      "grad_norm": 0.10170621424913406,
      "learning_rate": 2.2620864807459213e-05,
      "loss": 0.014,
      "step": 1909
    },
    {
      "epoch": 1.4328582145536384,
      "grad_norm": 0.10480276495218277,
      "learning_rate": 2.2566100209108048e-05,
      "loss": 0.0147,
      "step": 1910
    },
    {
      "epoch": 1.433608402100525,
      "grad_norm": 0.17218300700187683,
      "learning_rate": 2.2511382657193565e-05,
      "loss": 0.0171,
      "step": 1911
    },
    {
      "epoch": 1.4343585896474118,
      "grad_norm": 0.10617750883102417,
      "learning_rate": 2.2456712245550993e-05,
      "loss": 0.0175,
      "step": 1912
    },
    {
      "epoch": 1.4351087771942985,
      "grad_norm": 0.10631793737411499,
      "learning_rate": 2.2402089067934668e-05,
      "loss": 0.016,
      "step": 1913
    },
    {
      "epoch": 1.4358589647411852,
      "grad_norm": 0.15100601315498352,
      "learning_rate": 2.2347513218017974e-05,
      "loss": 0.0236,
      "step": 1914
    },
    {
      "epoch": 1.436609152288072,
      "grad_norm": 0.15311028063297272,
      "learning_rate": 2.2292984789393122e-05,
      "loss": 0.0237,
      "step": 1915
    },
    {
      "epoch": 1.4373593398349587,
      "grad_norm": 0.0808829814195633,
      "learning_rate": 2.2238503875571028e-05,
      "loss": 0.0093,
      "step": 1916
    },
    {
      "epoch": 1.4381095273818454,
      "grad_norm": 0.18772268295288086,
      "learning_rate": 2.218407056998104e-05,
      "loss": 0.0278,
      "step": 1917
    },
    {
      "epoch": 1.438859714928732,
      "grad_norm": 0.06562066823244095,
      "learning_rate": 2.2129684965970948e-05,
      "loss": 0.0077,
      "step": 1918
    },
    {
      "epoch": 1.439609902475619,
      "grad_norm": 0.1397140920162201,
      "learning_rate": 2.2075347156806697e-05,
      "loss": 0.0193,
      "step": 1919
    },
    {
      "epoch": 1.4403600900225055,
      "grad_norm": 0.1287301629781723,
      "learning_rate": 2.2021057235672288e-05,
      "loss": 0.0231,
      "step": 1920
    },
    {
      "epoch": 1.4411102775693925,
      "grad_norm": 0.10357276350259781,
      "learning_rate": 2.1966815295669585e-05,
      "loss": 0.0176,
      "step": 1921
    },
    {
      "epoch": 1.441860465116279,
      "grad_norm": 0.12166761606931686,
      "learning_rate": 2.1912621429818177e-05,
      "loss": 0.0171,
      "step": 1922
    },
    {
      "epoch": 1.442610652663166,
      "grad_norm": 0.08617536723613739,
      "learning_rate": 2.18584757310552e-05,
      "loss": 0.0111,
      "step": 1923
    },
    {
      "epoch": 1.4433608402100524,
      "grad_norm": 0.14840100705623627,
      "learning_rate": 2.1804378292235224e-05,
      "loss": 0.0181,
      "step": 1924
    },
    {
      "epoch": 1.4441110277569393,
      "grad_norm": 0.1149296835064888,
      "learning_rate": 2.1750329206129988e-05,
      "loss": 0.0198,
      "step": 1925
    },
    {
      "epoch": 1.4448612153038258,
      "grad_norm": 0.13646076619625092,
      "learning_rate": 2.1696328565428364e-05,
      "loss": 0.0169,
      "step": 1926
    },
    {
      "epoch": 1.4456114028507128,
      "grad_norm": 0.09913797676563263,
      "learning_rate": 2.1642376462736148e-05,
      "loss": 0.0109,
      "step": 1927
    },
    {
      "epoch": 1.4463615903975993,
      "grad_norm": 0.22945545613765717,
      "learning_rate": 2.158847299057587e-05,
      "loss": 0.0235,
      "step": 1928
    },
    {
      "epoch": 1.4471117779444862,
      "grad_norm": 0.0962451621890068,
      "learning_rate": 2.1534618241386705e-05,
      "loss": 0.0177,
      "step": 1929
    },
    {
      "epoch": 1.447861965491373,
      "grad_norm": 0.08848632872104645,
      "learning_rate": 2.14808123075242e-05,
      "loss": 0.0122,
      "step": 1930
    },
    {
      "epoch": 1.4486121530382596,
      "grad_norm": 0.13440725207328796,
      "learning_rate": 2.1427055281260255e-05,
      "loss": 0.0169,
      "step": 1931
    },
    {
      "epoch": 1.4493623405851463,
      "grad_norm": 0.08332593739032745,
      "learning_rate": 2.1373347254782882e-05,
      "loss": 0.0145,
      "step": 1932
    },
    {
      "epoch": 1.450112528132033,
      "grad_norm": 0.12370090931653976,
      "learning_rate": 2.1319688320196048e-05,
      "loss": 0.0157,
      "step": 1933
    },
    {
      "epoch": 1.4508627156789198,
      "grad_norm": 0.13575379550457,
      "learning_rate": 2.1266078569519542e-05,
      "loss": 0.0129,
      "step": 1934
    },
    {
      "epoch": 1.4516129032258065,
      "grad_norm": 0.10451852530241013,
      "learning_rate": 2.121251809468882e-05,
      "loss": 0.0181,
      "step": 1935
    },
    {
      "epoch": 1.4523630907726932,
      "grad_norm": 0.10986686497926712,
      "learning_rate": 2.1159006987554807e-05,
      "loss": 0.0178,
      "step": 1936
    },
    {
      "epoch": 1.45311327831958,
      "grad_norm": 0.08338434994220734,
      "learning_rate": 2.1105545339883808e-05,
      "loss": 0.0108,
      "step": 1937
    },
    {
      "epoch": 1.4538634658664666,
      "grad_norm": 0.14483703672885895,
      "learning_rate": 2.1052133243357253e-05,
      "loss": 0.0191,
      "step": 1938
    },
    {
      "epoch": 1.4546136534133534,
      "grad_norm": 0.12387536466121674,
      "learning_rate": 2.0998770789571636e-05,
      "loss": 0.0153,
      "step": 1939
    },
    {
      "epoch": 1.45536384096024,
      "grad_norm": 0.10853541642427444,
      "learning_rate": 2.0945458070038315e-05,
      "loss": 0.0115,
      "step": 1940
    },
    {
      "epoch": 1.4561140285071268,
      "grad_norm": 0.14933103322982788,
      "learning_rate": 2.0892195176183354e-05,
      "loss": 0.0164,
      "step": 1941
    },
    {
      "epoch": 1.4568642160540135,
      "grad_norm": 0.131834477186203,
      "learning_rate": 2.083898219934739e-05,
      "loss": 0.0176,
      "step": 1942
    },
    {
      "epoch": 1.4576144036009002,
      "grad_norm": 0.14950533211231232,
      "learning_rate": 2.0785819230785398e-05,
      "loss": 0.0226,
      "step": 1943
    },
    {
      "epoch": 1.458364591147787,
      "grad_norm": 0.0855354368686676,
      "learning_rate": 2.073270636166666e-05,
      "loss": 0.016,
      "step": 1944
    },
    {
      "epoch": 1.4591147786946737,
      "grad_norm": 0.13274472951889038,
      "learning_rate": 2.0679643683074513e-05,
      "loss": 0.0165,
      "step": 1945
    },
    {
      "epoch": 1.4598649662415604,
      "grad_norm": 0.10879681259393692,
      "learning_rate": 2.0626631286006236e-05,
      "loss": 0.0165,
      "step": 1946
    },
    {
      "epoch": 1.460615153788447,
      "grad_norm": 0.12825417518615723,
      "learning_rate": 2.0573669261372847e-05,
      "loss": 0.0211,
      "step": 1947
    },
    {
      "epoch": 1.4613653413353338,
      "grad_norm": 0.10240969061851501,
      "learning_rate": 2.052075769999899e-05,
      "loss": 0.0141,
      "step": 1948
    },
    {
      "epoch": 1.4621155288822205,
      "grad_norm": 0.08049630373716354,
      "learning_rate": 2.046789669262283e-05,
      "loss": 0.0122,
      "step": 1949
    },
    {
      "epoch": 1.4628657164291072,
      "grad_norm": 0.17087234556674957,
      "learning_rate": 2.0415086329895784e-05,
      "loss": 0.0184,
      "step": 1950
    },
    {
      "epoch": 1.463615903975994,
      "grad_norm": 0.09374770522117615,
      "learning_rate": 2.0362326702382384e-05,
      "loss": 0.0119,
      "step": 1951
    },
    {
      "epoch": 1.4643660915228807,
      "grad_norm": 0.10966373980045319,
      "learning_rate": 2.0309617900560218e-05,
      "loss": 0.0148,
      "step": 1952
    },
    {
      "epoch": 1.4651162790697674,
      "grad_norm": 0.1399017721414566,
      "learning_rate": 2.0256960014819692e-05,
      "loss": 0.02,
      "step": 1953
    },
    {
      "epoch": 1.465866466616654,
      "grad_norm": 0.10857044905424118,
      "learning_rate": 2.020435313546391e-05,
      "loss": 0.0161,
      "step": 1954
    },
    {
      "epoch": 1.4666166541635408,
      "grad_norm": 0.22650955617427826,
      "learning_rate": 2.0151797352708457e-05,
      "loss": 0.0407,
      "step": 1955
    },
    {
      "epoch": 1.4673668417104275,
      "grad_norm": 0.12916965782642365,
      "learning_rate": 2.0099292756681343e-05,
      "loss": 0.0206,
      "step": 1956
    },
    {
      "epoch": 1.4681170292573142,
      "grad_norm": 0.15416133403778076,
      "learning_rate": 2.0046839437422772e-05,
      "loss": 0.021,
      "step": 1957
    },
    {
      "epoch": 1.4688672168042012,
      "grad_norm": 0.08638902008533478,
      "learning_rate": 1.999443748488503e-05,
      "loss": 0.0104,
      "step": 1958
    },
    {
      "epoch": 1.4696174043510877,
      "grad_norm": 0.08310846239328384,
      "learning_rate": 1.9942086988932323e-05,
      "loss": 0.0119,
      "step": 1959
    },
    {
      "epoch": 1.4703675918979746,
      "grad_norm": 0.08909523487091064,
      "learning_rate": 1.9889788039340558e-05,
      "loss": 0.0173,
      "step": 1960
    },
    {
      "epoch": 1.471117779444861,
      "grad_norm": 0.1264401376247406,
      "learning_rate": 1.9837540725797305e-05,
      "loss": 0.0184,
      "step": 1961
    },
    {
      "epoch": 1.471867966991748,
      "grad_norm": 0.1094694435596466,
      "learning_rate": 1.9785345137901533e-05,
      "loss": 0.0153,
      "step": 1962
    },
    {
      "epoch": 1.4726181545386345,
      "grad_norm": 0.08005423098802567,
      "learning_rate": 1.9733201365163607e-05,
      "loss": 0.0126,
      "step": 1963
    },
    {
      "epoch": 1.4733683420855215,
      "grad_norm": 0.1491316258907318,
      "learning_rate": 1.968110949700489e-05,
      "loss": 0.025,
      "step": 1964
    },
    {
      "epoch": 1.474118529632408,
      "grad_norm": 0.10460489988327026,
      "learning_rate": 1.962906962275784e-05,
      "loss": 0.0161,
      "step": 1965
    },
    {
      "epoch": 1.474868717179295,
      "grad_norm": 0.07633949816226959,
      "learning_rate": 1.9577081831665707e-05,
      "loss": 0.0112,
      "step": 1966
    },
    {
      "epoch": 1.4756189047261816,
      "grad_norm": 0.12210875749588013,
      "learning_rate": 1.9525146212882456e-05,
      "loss": 0.0244,
      "step": 1967
    },
    {
      "epoch": 1.4763690922730683,
      "grad_norm": 0.10730652511119843,
      "learning_rate": 1.9473262855472517e-05,
      "loss": 0.0162,
      "step": 1968
    },
    {
      "epoch": 1.477119279819955,
      "grad_norm": 0.1215217113494873,
      "learning_rate": 1.942143184841077e-05,
      "loss": 0.0162,
      "step": 1969
    },
    {
      "epoch": 1.4778694673668418,
      "grad_norm": 0.0921521708369255,
      "learning_rate": 1.9369653280582273e-05,
      "loss": 0.015,
      "step": 1970
    },
    {
      "epoch": 1.4786196549137285,
      "grad_norm": 0.19362576305866241,
      "learning_rate": 1.931792724078218e-05,
      "loss": 0.0275,
      "step": 1971
    },
    {
      "epoch": 1.4793698424606152,
      "grad_norm": 0.12856422364711761,
      "learning_rate": 1.9266253817715575e-05,
      "loss": 0.017,
      "step": 1972
    },
    {
      "epoch": 1.480120030007502,
      "grad_norm": 0.17855559289455414,
      "learning_rate": 1.921463309999724e-05,
      "loss": 0.0222,
      "step": 1973
    },
    {
      "epoch": 1.4808702175543886,
      "grad_norm": 0.10688813775777817,
      "learning_rate": 1.9163065176151662e-05,
      "loss": 0.0149,
      "step": 1974
    },
    {
      "epoch": 1.4816204051012754,
      "grad_norm": 0.10548462718725204,
      "learning_rate": 1.9111550134612738e-05,
      "loss": 0.0143,
      "step": 1975
    },
    {
      "epoch": 1.482370592648162,
      "grad_norm": 0.07540253549814224,
      "learning_rate": 1.9060088063723696e-05,
      "loss": 0.0094,
      "step": 1976
    },
    {
      "epoch": 1.4831207801950488,
      "grad_norm": 0.09331195056438446,
      "learning_rate": 1.900867905173692e-05,
      "loss": 0.015,
      "step": 1977
    },
    {
      "epoch": 1.4838709677419355,
      "grad_norm": 0.09781509637832642,
      "learning_rate": 1.8957323186813803e-05,
      "loss": 0.0107,
      "step": 1978
    },
    {
      "epoch": 1.4846211552888222,
      "grad_norm": 0.12924246490001678,
      "learning_rate": 1.8906020557024597e-05,
      "loss": 0.0132,
      "step": 1979
    },
    {
      "epoch": 1.485371342835709,
      "grad_norm": 0.26540517807006836,
      "learning_rate": 1.885477125034827e-05,
      "loss": 0.0255,
      "step": 1980
    },
    {
      "epoch": 1.4861215303825956,
      "grad_norm": 0.17854297161102295,
      "learning_rate": 1.8803575354672315e-05,
      "loss": 0.0234,
      "step": 1981
    },
    {
      "epoch": 1.4868717179294824,
      "grad_norm": 0.1174192726612091,
      "learning_rate": 1.8752432957792654e-05,
      "loss": 0.0136,
      "step": 1982
    },
    {
      "epoch": 1.487621905476369,
      "grad_norm": 0.14164678752422333,
      "learning_rate": 1.8701344147413474e-05,
      "loss": 0.0221,
      "step": 1983
    },
    {
      "epoch": 1.4883720930232558,
      "grad_norm": 0.08099871128797531,
      "learning_rate": 1.8650309011147053e-05,
      "loss": 0.0129,
      "step": 1984
    },
    {
      "epoch": 1.4891222805701425,
      "grad_norm": 0.09161733090877533,
      "learning_rate": 1.8599327636513636e-05,
      "loss": 0.0118,
      "step": 1985
    },
    {
      "epoch": 1.4898724681170292,
      "grad_norm": 0.1466006189584732,
      "learning_rate": 1.8548400110941228e-05,
      "loss": 0.0109,
      "step": 1986
    },
    {
      "epoch": 1.490622655663916,
      "grad_norm": 0.10579754412174225,
      "learning_rate": 1.8497526521765534e-05,
      "loss": 0.0101,
      "step": 1987
    },
    {
      "epoch": 1.4913728432108027,
      "grad_norm": 0.06394397467374802,
      "learning_rate": 1.844670695622976e-05,
      "loss": 0.0073,
      "step": 1988
    },
    {
      "epoch": 1.4921230307576894,
      "grad_norm": 0.1046217605471611,
      "learning_rate": 1.8395941501484464e-05,
      "loss": 0.0163,
      "step": 1989
    },
    {
      "epoch": 1.492873218304576,
      "grad_norm": 0.09319301694631577,
      "learning_rate": 1.8345230244587354e-05,
      "loss": 0.0138,
      "step": 1990
    },
    {
      "epoch": 1.4936234058514628,
      "grad_norm": 0.10420255362987518,
      "learning_rate": 1.829457327250329e-05,
      "loss": 0.0124,
      "step": 1991
    },
    {
      "epoch": 1.4943735933983495,
      "grad_norm": 0.12257860600948334,
      "learning_rate": 1.8243970672103982e-05,
      "loss": 0.0143,
      "step": 1992
    },
    {
      "epoch": 1.4951237809452362,
      "grad_norm": 0.15633484721183777,
      "learning_rate": 1.8193422530167914e-05,
      "loss": 0.0183,
      "step": 1993
    },
    {
      "epoch": 1.495873968492123,
      "grad_norm": 0.09158031642436981,
      "learning_rate": 1.8142928933380142e-05,
      "loss": 0.0149,
      "step": 1994
    },
    {
      "epoch": 1.49662415603901,
      "grad_norm": 0.10753414034843445,
      "learning_rate": 1.8092489968332233e-05,
      "loss": 0.0138,
      "step": 1995
    },
    {
      "epoch": 1.4973743435858964,
      "grad_norm": 0.09815339744091034,
      "learning_rate": 1.804210572152204e-05,
      "loss": 0.0118,
      "step": 1996
    },
    {
      "epoch": 1.4981245311327833,
      "grad_norm": 0.13312570750713348,
      "learning_rate": 1.7991776279353604e-05,
      "loss": 0.0186,
      "step": 1997
    },
    {
      "epoch": 1.4988747186796698,
      "grad_norm": 0.08981100469827652,
      "learning_rate": 1.794150172813693e-05,
      "loss": 0.0095,
      "step": 1998
    },
    {
      "epoch": 1.4996249062265568,
      "grad_norm": 0.18983112275600433,
      "learning_rate": 1.7891282154087934e-05,
      "loss": 0.0228,
      "step": 1999
    },
    {
      "epoch": 1.5003750937734432,
      "grad_norm": 0.10713919252157211,
      "learning_rate": 1.7841117643328246e-05,
      "loss": 0.0154,
      "step": 2000
    },
    {
      "epoch": 1.5003750937734432,
      "eval_loss": 0.03212602809071541,
      "eval_runtime": 5.1295,
      "eval_samples_per_second": 10.527,
      "eval_steps_per_second": 2.729,
      "step": 2000
    },
    {
      "epoch": 1.5011252813203302,
      "grad_norm": 0.105213463306427,
      "learning_rate": 1.779100828188506e-05,
      "loss": 0.0187,
      "step": 2001
    },
    {
      "epoch": 1.5018754688672167,
      "grad_norm": 0.1180376186966896,
      "learning_rate": 1.774095415569102e-05,
      "loss": 0.0128,
      "step": 2002
    },
    {
      "epoch": 1.5026256564141036,
      "grad_norm": 0.09247060120105743,
      "learning_rate": 1.7690955350583976e-05,
      "loss": 0.0089,
      "step": 2003
    },
    {
      "epoch": 1.50337584396099,
      "grad_norm": 0.22491805255413055,
      "learning_rate": 1.764101195230696e-05,
      "loss": 0.0351,
      "step": 2004
    },
    {
      "epoch": 1.504126031507877,
      "grad_norm": 0.6523817181587219,
      "learning_rate": 1.7591124046508045e-05,
      "loss": 0.0568,
      "step": 2005
    },
    {
      "epoch": 1.5048762190547635,
      "grad_norm": 0.11308412998914719,
      "learning_rate": 1.7541291718740012e-05,
      "loss": 0.0138,
      "step": 2006
    },
    {
      "epoch": 1.5056264066016505,
      "grad_norm": 0.10251988470554352,
      "learning_rate": 1.7491515054460418e-05,
      "loss": 0.0118,
      "step": 2007
    },
    {
      "epoch": 1.506376594148537,
      "grad_norm": 0.09428345412015915,
      "learning_rate": 1.7441794139031337e-05,
      "loss": 0.0191,
      "step": 2008
    },
    {
      "epoch": 1.507126781695424,
      "grad_norm": 0.15063363313674927,
      "learning_rate": 1.7392129057719246e-05,
      "loss": 0.0192,
      "step": 2009
    },
    {
      "epoch": 1.5078769692423106,
      "grad_norm": 0.14728331565856934,
      "learning_rate": 1.7342519895694886e-05,
      "loss": 0.0209,
      "step": 2010
    },
    {
      "epoch": 1.5086271567891973,
      "grad_norm": 0.14803294837474823,
      "learning_rate": 1.7292966738033057e-05,
      "loss": 0.0186,
      "step": 2011
    },
    {
      "epoch": 1.509377344336084,
      "grad_norm": 0.11730030179023743,
      "learning_rate": 1.7243469669712546e-05,
      "loss": 0.0185,
      "step": 2012
    },
    {
      "epoch": 1.5101275318829708,
      "grad_norm": 0.10365916788578033,
      "learning_rate": 1.7194028775615966e-05,
      "loss": 0.0143,
      "step": 2013
    },
    {
      "epoch": 1.5108777194298575,
      "grad_norm": 0.12164322286844254,
      "learning_rate": 1.714464414052958e-05,
      "loss": 0.0144,
      "step": 2014
    },
    {
      "epoch": 1.5116279069767442,
      "grad_norm": 0.1376684308052063,
      "learning_rate": 1.7095315849143184e-05,
      "loss": 0.0236,
      "step": 2015
    },
    {
      "epoch": 1.512378094523631,
      "grad_norm": 0.11822548508644104,
      "learning_rate": 1.704604398604991e-05,
      "loss": 0.016,
      "step": 2016
    },
    {
      "epoch": 1.5131282820705176,
      "grad_norm": 0.11870657652616501,
      "learning_rate": 1.6996828635746165e-05,
      "loss": 0.0202,
      "step": 2017
    },
    {
      "epoch": 1.5138784696174044,
      "grad_norm": 0.098456971347332,
      "learning_rate": 1.6947669882631434e-05,
      "loss": 0.0128,
      "step": 2018
    },
    {
      "epoch": 1.514628657164291,
      "grad_norm": 0.125031977891922,
      "learning_rate": 1.6898567811008135e-05,
      "loss": 0.0134,
      "step": 2019
    },
    {
      "epoch": 1.5153788447111778,
      "grad_norm": 0.08734133839607239,
      "learning_rate": 1.684952250508149e-05,
      "loss": 0.0117,
      "step": 2020
    },
    {
      "epoch": 1.5161290322580645,
      "grad_norm": 0.09543147683143616,
      "learning_rate": 1.6800534048959364e-05,
      "loss": 0.012,
      "step": 2021
    },
    {
      "epoch": 1.5168792198049512,
      "grad_norm": 0.11019328981637955,
      "learning_rate": 1.6751602526652133e-05,
      "loss": 0.0151,
      "step": 2022
    },
    {
      "epoch": 1.517629407351838,
      "grad_norm": 0.16680921614170074,
      "learning_rate": 1.6702728022072562e-05,
      "loss": 0.0181,
      "step": 2023
    },
    {
      "epoch": 1.5183795948987246,
      "grad_norm": 0.09131893515586853,
      "learning_rate": 1.665391061903558e-05,
      "loss": 0.0126,
      "step": 2024
    },
    {
      "epoch": 1.5191297824456114,
      "grad_norm": 0.13133695721626282,
      "learning_rate": 1.660515040125824e-05,
      "loss": 0.0241,
      "step": 2025
    },
    {
      "epoch": 1.519879969992498,
      "grad_norm": 0.09224233776330948,
      "learning_rate": 1.6556447452359512e-05,
      "loss": 0.0148,
      "step": 2026
    },
    {
      "epoch": 1.5206301575393848,
      "grad_norm": 0.1913694441318512,
      "learning_rate": 1.6507801855860177e-05,
      "loss": 0.0211,
      "step": 2027
    },
    {
      "epoch": 1.5213803450862715,
      "grad_norm": 0.09763851761817932,
      "learning_rate": 1.645921369518261e-05,
      "loss": 0.0106,
      "step": 2028
    },
    {
      "epoch": 1.5221305326331582,
      "grad_norm": 0.13741938769817352,
      "learning_rate": 1.6410683053650737e-05,
      "loss": 0.0161,
      "step": 2029
    },
    {
      "epoch": 1.5228807201800452,
      "grad_norm": 0.16641104221343994,
      "learning_rate": 1.636221001448983e-05,
      "loss": 0.0226,
      "step": 2030
    },
    {
      "epoch": 1.5236309077269317,
      "grad_norm": 0.14530512690544128,
      "learning_rate": 1.631379466082638e-05,
      "loss": 0.0234,
      "step": 2031
    },
    {
      "epoch": 1.5243810952738186,
      "grad_norm": 0.1273575872182846,
      "learning_rate": 1.626543707568795e-05,
      "loss": 0.0153,
      "step": 2032
    },
    {
      "epoch": 1.525131282820705,
      "grad_norm": 0.18259398639202118,
      "learning_rate": 1.6217137342003036e-05,
      "loss": 0.0178,
      "step": 2033
    },
    {
      "epoch": 1.525881470367592,
      "grad_norm": 0.1915973424911499,
      "learning_rate": 1.616889554260092e-05,
      "loss": 0.0224,
      "step": 2034
    },
    {
      "epoch": 1.5266316579144785,
      "grad_norm": 0.14290551841259003,
      "learning_rate": 1.6120711760211548e-05,
      "loss": 0.0205,
      "step": 2035
    },
    {
      "epoch": 1.5273818454613655,
      "grad_norm": 0.10475607216358185,
      "learning_rate": 1.607258607746537e-05,
      "loss": 0.0187,
      "step": 2036
    },
    {
      "epoch": 1.528132033008252,
      "grad_norm": 0.1033739373087883,
      "learning_rate": 1.602451857689316e-05,
      "loss": 0.0129,
      "step": 2037
    },
    {
      "epoch": 1.528882220555139,
      "grad_norm": 0.09066800773143768,
      "learning_rate": 1.5976509340925977e-05,
      "loss": 0.0094,
      "step": 2038
    },
    {
      "epoch": 1.5296324081020254,
      "grad_norm": 0.12105967849493027,
      "learning_rate": 1.5928558451894914e-05,
      "loss": 0.0142,
      "step": 2039
    },
    {
      "epoch": 1.5303825956489123,
      "grad_norm": 0.38493460416793823,
      "learning_rate": 1.588066599203106e-05,
      "loss": 0.0288,
      "step": 2040
    },
    {
      "epoch": 1.5311327831957988,
      "grad_norm": 0.08361639082431793,
      "learning_rate": 1.583283204346521e-05,
      "loss": 0.0118,
      "step": 2041
    },
    {
      "epoch": 1.5318829707426858,
      "grad_norm": 0.13456936180591583,
      "learning_rate": 1.5785056688227916e-05,
      "loss": 0.0133,
      "step": 2042
    },
    {
      "epoch": 1.5326331582895723,
      "grad_norm": 0.16199876368045807,
      "learning_rate": 1.5737340008249202e-05,
      "loss": 0.0225,
      "step": 2043
    },
    {
      "epoch": 1.5333833458364592,
      "grad_norm": 0.0884062722325325,
      "learning_rate": 1.5689682085358465e-05,
      "loss": 0.0107,
      "step": 2044
    },
    {
      "epoch": 1.5341335333833457,
      "grad_norm": 0.06545273959636688,
      "learning_rate": 1.564208300128438e-05,
      "loss": 0.0084,
      "step": 2045
    },
    {
      "epoch": 1.5348837209302326,
      "grad_norm": 0.08026087284088135,
      "learning_rate": 1.5594542837654625e-05,
      "loss": 0.0094,
      "step": 2046
    },
    {
      "epoch": 1.5356339084771191,
      "grad_norm": 0.11514996737241745,
      "learning_rate": 1.554706167599596e-05,
      "loss": 0.0161,
      "step": 2047
    },
    {
      "epoch": 1.536384096024006,
      "grad_norm": 0.14320102334022522,
      "learning_rate": 1.5499639597733902e-05,
      "loss": 0.0168,
      "step": 2048
    },
    {
      "epoch": 1.5371342835708928,
      "grad_norm": 0.11476108431816101,
      "learning_rate": 1.54522766841926e-05,
      "loss": 0.0163,
      "step": 2049
    },
    {
      "epoch": 1.5378844711177795,
      "grad_norm": 0.17724426090717316,
      "learning_rate": 1.540497301659482e-05,
      "loss": 0.0182,
      "step": 2050
    },
    {
      "epoch": 1.5386346586646662,
      "grad_norm": 0.15506739914417267,
      "learning_rate": 1.5357728676061685e-05,
      "loss": 0.0185,
      "step": 2051
    },
    {
      "epoch": 1.539384846211553,
      "grad_norm": 0.19287851452827454,
      "learning_rate": 1.5310543743612582e-05,
      "loss": 0.0222,
      "step": 2052
    },
    {
      "epoch": 1.5401350337584396,
      "grad_norm": 0.14303511381149292,
      "learning_rate": 1.526341830016505e-05,
      "loss": 0.0209,
      "step": 2053
    },
    {
      "epoch": 1.5408852213053263,
      "grad_norm": 0.12963083386421204,
      "learning_rate": 1.5216352426534548e-05,
      "loss": 0.0198,
      "step": 2054
    },
    {
      "epoch": 1.541635408852213,
      "grad_norm": 0.1562402844429016,
      "learning_rate": 1.5169346203434425e-05,
      "loss": 0.0189,
      "step": 2055
    },
    {
      "epoch": 1.5423855963990998,
      "grad_norm": 0.15967603027820587,
      "learning_rate": 1.5122399711475732e-05,
      "loss": 0.0253,
      "step": 2056
    },
    {
      "epoch": 1.5431357839459865,
      "grad_norm": 0.16278547048568726,
      "learning_rate": 1.50755130311671e-05,
      "loss": 0.0206,
      "step": 2057
    },
    {
      "epoch": 1.5438859714928732,
      "grad_norm": 0.13421553373336792,
      "learning_rate": 1.502868624291452e-05,
      "loss": 0.019,
      "step": 2058
    },
    {
      "epoch": 1.54463615903976,
      "grad_norm": 0.12211119383573532,
      "learning_rate": 1.4981919427021357e-05,
      "loss": 0.0176,
      "step": 2059
    },
    {
      "epoch": 1.5453863465866466,
      "grad_norm": 0.11205483227968216,
      "learning_rate": 1.493521266368807e-05,
      "loss": 0.0144,
      "step": 2060
    },
    {
      "epoch": 1.5461365341335334,
      "grad_norm": 0.10355428606271744,
      "learning_rate": 1.4888566033012201e-05,
      "loss": 0.0181,
      "step": 2061
    },
    {
      "epoch": 1.54688672168042,
      "grad_norm": 0.07181017100811005,
      "learning_rate": 1.4841979614988094e-05,
      "loss": 0.0092,
      "step": 2062
    },
    {
      "epoch": 1.5476369092273068,
      "grad_norm": 0.09114021062850952,
      "learning_rate": 1.4795453489506878e-05,
      "loss": 0.0159,
      "step": 2063
    },
    {
      "epoch": 1.5483870967741935,
      "grad_norm": 0.10014685988426208,
      "learning_rate": 1.4748987736356273e-05,
      "loss": 0.0149,
      "step": 2064
    },
    {
      "epoch": 1.5491372843210802,
      "grad_norm": 0.10421111434698105,
      "learning_rate": 1.4702582435220475e-05,
      "loss": 0.0174,
      "step": 2065
    },
    {
      "epoch": 1.549887471867967,
      "grad_norm": 0.17178568243980408,
      "learning_rate": 1.4656237665680017e-05,
      "loss": 0.0273,
      "step": 2066
    },
    {
      "epoch": 1.5506376594148539,
      "grad_norm": 0.0894736498594284,
      "learning_rate": 1.4609953507211593e-05,
      "loss": 0.0137,
      "step": 2067
    },
    {
      "epoch": 1.5513878469617404,
      "grad_norm": 0.09273448586463928,
      "learning_rate": 1.4563730039187984e-05,
      "loss": 0.0157,
      "step": 2068
    },
    {
      "epoch": 1.5521380345086273,
      "grad_norm": 0.06843544542789459,
      "learning_rate": 1.4517567340877886e-05,
      "loss": 0.0092,
      "step": 2069
    },
    {
      "epoch": 1.5528882220555138,
      "grad_norm": 0.09910368174314499,
      "learning_rate": 1.4471465491445802e-05,
      "loss": 0.0124,
      "step": 2070
    },
    {
      "epoch": 1.5536384096024007,
      "grad_norm": 0.20299047231674194,
      "learning_rate": 1.4425424569951822e-05,
      "loss": 0.0179,
      "step": 2071
    },
    {
      "epoch": 1.5543885971492872,
      "grad_norm": 0.1219157725572586,
      "learning_rate": 1.4379444655351626e-05,
      "loss": 0.0229,
      "step": 2072
    },
    {
      "epoch": 1.5551387846961742,
      "grad_norm": 0.056685127317905426,
      "learning_rate": 1.4333525826496224e-05,
      "loss": 0.0047,
      "step": 2073
    },
    {
      "epoch": 1.5558889722430607,
      "grad_norm": 0.11681665480136871,
      "learning_rate": 1.4287668162131896e-05,
      "loss": 0.0177,
      "step": 2074
    },
    {
      "epoch": 1.5566391597899476,
      "grad_norm": 0.12039943039417267,
      "learning_rate": 1.4241871740900014e-05,
      "loss": 0.0154,
      "step": 2075
    },
    {
      "epoch": 1.557389347336834,
      "grad_norm": 0.12308166176080704,
      "learning_rate": 1.4196136641336932e-05,
      "loss": 0.0185,
      "step": 2076
    },
    {
      "epoch": 1.558139534883721,
      "grad_norm": 0.12046025693416595,
      "learning_rate": 1.4150462941873843e-05,
      "loss": 0.0149,
      "step": 2077
    },
    {
      "epoch": 1.5588897224306075,
      "grad_norm": 0.09745654463768005,
      "learning_rate": 1.410485072083666e-05,
      "loss": 0.0133,
      "step": 2078
    },
    {
      "epoch": 1.5596399099774945,
      "grad_norm": 0.09617792069911957,
      "learning_rate": 1.4059300056445823e-05,
      "loss": 0.0163,
      "step": 2079
    },
    {
      "epoch": 1.560390097524381,
      "grad_norm": 0.13568072021007538,
      "learning_rate": 1.4013811026816243e-05,
      "loss": 0.0187,
      "step": 2080
    },
    {
      "epoch": 1.561140285071268,
      "grad_norm": 0.15544694662094116,
      "learning_rate": 1.3968383709957133e-05,
      "loss": 0.0136,
      "step": 2081
    },
    {
      "epoch": 1.5618904726181544,
      "grad_norm": 0.0854220986366272,
      "learning_rate": 1.3923018183771868e-05,
      "loss": 0.0126,
      "step": 2082
    },
    {
      "epoch": 1.5626406601650413,
      "grad_norm": 0.16740089654922485,
      "learning_rate": 1.3877714526057872e-05,
      "loss": 0.0207,
      "step": 2083
    },
    {
      "epoch": 1.5633908477119278,
      "grad_norm": 0.20988257229328156,
      "learning_rate": 1.3832472814506425e-05,
      "loss": 0.0261,
      "step": 2084
    },
    {
      "epoch": 1.5641410352588148,
      "grad_norm": 0.13327625393867493,
      "learning_rate": 1.3787293126702622e-05,
      "loss": 0.0171,
      "step": 2085
    },
    {
      "epoch": 1.5648912228057015,
      "grad_norm": 0.22465142607688904,
      "learning_rate": 1.3742175540125179e-05,
      "loss": 0.0305,
      "step": 2086
    },
    {
      "epoch": 1.5656414103525882,
      "grad_norm": 0.11496560275554657,
      "learning_rate": 1.3697120132146318e-05,
      "loss": 0.0179,
      "step": 2087
    },
    {
      "epoch": 1.566391597899475,
      "grad_norm": 0.08796335756778717,
      "learning_rate": 1.3652126980031627e-05,
      "loss": 0.0097,
      "step": 2088
    },
    {
      "epoch": 1.5671417854463616,
      "grad_norm": 0.09953059256076813,
      "learning_rate": 1.3607196160939927e-05,
      "loss": 0.0112,
      "step": 2089
    },
    {
      "epoch": 1.5678919729932483,
      "grad_norm": 0.10293276607990265,
      "learning_rate": 1.3562327751923149e-05,
      "loss": 0.0169,
      "step": 2090
    },
    {
      "epoch": 1.568642160540135,
      "grad_norm": 0.15932396054267883,
      "learning_rate": 1.351752182992621e-05,
      "loss": 0.0227,
      "step": 2091
    },
    {
      "epoch": 1.5693923480870218,
      "grad_norm": 0.06464660167694092,
      "learning_rate": 1.3472778471786829e-05,
      "loss": 0.0069,
      "step": 2092
    },
    {
      "epoch": 1.5701425356339085,
      "grad_norm": 0.11081813275814056,
      "learning_rate": 1.3428097754235475e-05,
      "loss": 0.0131,
      "step": 2093
    },
    {
      "epoch": 1.5708927231807952,
      "grad_norm": 0.1540287286043167,
      "learning_rate": 1.3383479753895174e-05,
      "loss": 0.0267,
      "step": 2094
    },
    {
      "epoch": 1.571642910727682,
      "grad_norm": 0.13682395219802856,
      "learning_rate": 1.33389245472814e-05,
      "loss": 0.0134,
      "step": 2095
    },
    {
      "epoch": 1.5723930982745686,
      "grad_norm": 0.1723647564649582,
      "learning_rate": 1.3294432210801966e-05,
      "loss": 0.0251,
      "step": 2096
    },
    {
      "epoch": 1.5731432858214554,
      "grad_norm": 0.12762096524238586,
      "learning_rate": 1.3250002820756819e-05,
      "loss": 0.0188,
      "step": 2097
    },
    {
      "epoch": 1.573893473368342,
      "grad_norm": 0.14758145809173584,
      "learning_rate": 1.3205636453338e-05,
      "loss": 0.0173,
      "step": 2098
    },
    {
      "epoch": 1.5746436609152288,
      "grad_norm": 0.0950731411576271,
      "learning_rate": 1.316133318462946e-05,
      "loss": 0.0129,
      "step": 2099
    },
    {
      "epoch": 1.5753938484621155,
      "grad_norm": 0.09899888187646866,
      "learning_rate": 1.3117093090606958e-05,
      "loss": 0.0132,
      "step": 2100
    },
    {
      "epoch": 1.5761440360090022,
      "grad_norm": 0.15141884982585907,
      "learning_rate": 1.3072916247137861e-05,
      "loss": 0.0157,
      "step": 2101
    },
    {
      "epoch": 1.576894223555889,
      "grad_norm": 0.09063942730426788,
      "learning_rate": 1.302880272998112e-05,
      "loss": 0.0119,
      "step": 2102
    },
    {
      "epoch": 1.5776444111027756,
      "grad_norm": 0.14490434527397156,
      "learning_rate": 1.29847526147871e-05,
      "loss": 0.0164,
      "step": 2103
    },
    {
      "epoch": 1.5783945986496624,
      "grad_norm": 0.09947270154953003,
      "learning_rate": 1.2940765977097402e-05,
      "loss": 0.0122,
      "step": 2104
    },
    {
      "epoch": 1.579144786196549,
      "grad_norm": 0.24933743476867676,
      "learning_rate": 1.2896842892344751e-05,
      "loss": 0.0188,
      "step": 2105
    },
    {
      "epoch": 1.579894973743436,
      "grad_norm": 0.10064645111560822,
      "learning_rate": 1.2852983435852928e-05,
      "loss": 0.0127,
      "step": 2106
    },
    {
      "epoch": 1.5806451612903225,
      "grad_norm": 0.11761155724525452,
      "learning_rate": 1.2809187682836588e-05,
      "loss": 0.0121,
      "step": 2107
    },
    {
      "epoch": 1.5813953488372094,
      "grad_norm": 0.12103329598903656,
      "learning_rate": 1.2765455708401142e-05,
      "loss": 0.0156,
      "step": 2108
    },
    {
      "epoch": 1.582145536384096,
      "grad_norm": 0.24466180801391602,
      "learning_rate": 1.2721787587542595e-05,
      "loss": 0.0367,
      "step": 2109
    },
    {
      "epoch": 1.5828957239309829,
      "grad_norm": 0.11228577047586441,
      "learning_rate": 1.2678183395147485e-05,
      "loss": 0.0176,
      "step": 2110
    },
    {
      "epoch": 1.5836459114778694,
      "grad_norm": 0.13010624051094055,
      "learning_rate": 1.2634643205992707e-05,
      "loss": 0.0134,
      "step": 2111
    },
    {
      "epoch": 1.5843960990247563,
      "grad_norm": 0.12314150482416153,
      "learning_rate": 1.2591167094745404e-05,
      "loss": 0.0215,
      "step": 2112
    },
    {
      "epoch": 1.5851462865716428,
      "grad_norm": 0.09590944647789001,
      "learning_rate": 1.2547755135962841e-05,
      "loss": 0.0179,
      "step": 2113
    },
    {
      "epoch": 1.5858964741185297,
      "grad_norm": 0.1403607428073883,
      "learning_rate": 1.2504407404092217e-05,
      "loss": 0.0255,
      "step": 2114
    },
    {
      "epoch": 1.5866466616654162,
      "grad_norm": 0.09067095071077347,
      "learning_rate": 1.2461123973470634e-05,
      "loss": 0.0129,
      "step": 2115
    },
    {
      "epoch": 1.5873968492123032,
      "grad_norm": 0.13620354235172272,
      "learning_rate": 1.2417904918324913e-05,
      "loss": 0.018,
      "step": 2116
    },
    {
      "epoch": 1.5881470367591897,
      "grad_norm": 0.0839877724647522,
      "learning_rate": 1.237475031277151e-05,
      "loss": 0.0108,
      "step": 2117
    },
    {
      "epoch": 1.5888972243060766,
      "grad_norm": 0.0976109579205513,
      "learning_rate": 1.2331660230816288e-05,
      "loss": 0.0129,
      "step": 2118
    },
    {
      "epoch": 1.589647411852963,
      "grad_norm": 0.15283679962158203,
      "learning_rate": 1.2288634746354505e-05,
      "loss": 0.0134,
      "step": 2119
    },
    {
      "epoch": 1.59039759939985,
      "grad_norm": 0.09862658381462097,
      "learning_rate": 1.2245673933170626e-05,
      "loss": 0.0117,
      "step": 2120
    },
    {
      "epoch": 1.5911477869467365,
      "grad_norm": 0.09258013963699341,
      "learning_rate": 1.2202777864938236e-05,
      "loss": 0.015,
      "step": 2121
    },
    {
      "epoch": 1.5918979744936235,
      "grad_norm": 0.15144041180610657,
      "learning_rate": 1.2159946615219836e-05,
      "loss": 0.0141,
      "step": 2122
    },
    {
      "epoch": 1.59264816204051,
      "grad_norm": 0.12255115807056427,
      "learning_rate": 1.211718025746682e-05,
      "loss": 0.0095,
      "step": 2123
    },
    {
      "epoch": 1.593398349587397,
      "grad_norm": 0.1431581974029541,
      "learning_rate": 1.2074478865019273e-05,
      "loss": 0.0139,
      "step": 2124
    },
    {
      "epoch": 1.5941485371342836,
      "grad_norm": 0.09224507957696915,
      "learning_rate": 1.2031842511105885e-05,
      "loss": 0.0121,
      "step": 2125
    },
    {
      "epoch": 1.5948987246811703,
      "grad_norm": 0.1134839728474617,
      "learning_rate": 1.1989271268843815e-05,
      "loss": 0.0119,
      "step": 2126
    },
    {
      "epoch": 1.595648912228057,
      "grad_norm": 0.13770152628421783,
      "learning_rate": 1.1946765211238526e-05,
      "loss": 0.0167,
      "step": 2127
    },
    {
      "epoch": 1.5963990997749438,
      "grad_norm": 0.14627668261528015,
      "learning_rate": 1.1904324411183731e-05,
      "loss": 0.0174,
      "step": 2128
    },
    {
      "epoch": 1.5971492873218305,
      "grad_norm": 0.12559792399406433,
      "learning_rate": 1.1861948941461226e-05,
      "loss": 0.0138,
      "step": 2129
    },
    {
      "epoch": 1.5978994748687172,
      "grad_norm": 0.1679619550704956,
      "learning_rate": 1.1819638874740769e-05,
      "loss": 0.0179,
      "step": 2130
    },
    {
      "epoch": 1.598649662415604,
      "grad_norm": 0.11241298913955688,
      "learning_rate": 1.1777394283579956e-05,
      "loss": 0.0115,
      "step": 2131
    },
    {
      "epoch": 1.5993998499624906,
      "grad_norm": 0.09657392650842667,
      "learning_rate": 1.1735215240424102e-05,
      "loss": 0.0107,
      "step": 2132
    },
    {
      "epoch": 1.6001500375093773,
      "grad_norm": 0.12358458340167999,
      "learning_rate": 1.1693101817606117e-05,
      "loss": 0.0204,
      "step": 2133
    },
    {
      "epoch": 1.600900225056264,
      "grad_norm": 0.09444307535886765,
      "learning_rate": 1.165105408734638e-05,
      "loss": 0.0109,
      "step": 2134
    },
    {
      "epoch": 1.6016504126031508,
      "grad_norm": 0.1118428185582161,
      "learning_rate": 1.1609072121752584e-05,
      "loss": 0.0201,
      "step": 2135
    },
    {
      "epoch": 1.6024006001500375,
      "grad_norm": 0.10624516010284424,
      "learning_rate": 1.1567155992819678e-05,
      "loss": 0.016,
      "step": 2136
    },
    {
      "epoch": 1.6031507876969242,
      "grad_norm": 0.20431579649448395,
      "learning_rate": 1.15253057724297e-05,
      "loss": 0.0252,
      "step": 2137
    },
    {
      "epoch": 1.603900975243811,
      "grad_norm": 0.14086978137493134,
      "learning_rate": 1.1483521532351654e-05,
      "loss": 0.0166,
      "step": 2138
    },
    {
      "epoch": 1.6046511627906976,
      "grad_norm": 0.10527478903532028,
      "learning_rate": 1.144180334424141e-05,
      "loss": 0.014,
      "step": 2139
    },
    {
      "epoch": 1.6054013503375844,
      "grad_norm": 0.10598310083150864,
      "learning_rate": 1.1400151279641525e-05,
      "loss": 0.0137,
      "step": 2140
    },
    {
      "epoch": 1.606151537884471,
      "grad_norm": 0.13043269515037537,
      "learning_rate": 1.1358565409981203e-05,
      "loss": 0.014,
      "step": 2141
    },
    {
      "epoch": 1.6069017254313578,
      "grad_norm": 0.1680910736322403,
      "learning_rate": 1.1317045806576121e-05,
      "loss": 0.0215,
      "step": 2142
    },
    {
      "epoch": 1.6076519129782447,
      "grad_norm": 0.16358719766139984,
      "learning_rate": 1.12755925406283e-05,
      "loss": 0.0223,
      "step": 2143
    },
    {
      "epoch": 1.6084021005251312,
      "grad_norm": 0.12614835798740387,
      "learning_rate": 1.1234205683226012e-05,
      "loss": 0.0148,
      "step": 2144
    },
    {
      "epoch": 1.6091522880720182,
      "grad_norm": 0.05313693732023239,
      "learning_rate": 1.1192885305343648e-05,
      "loss": 0.0079,
      "step": 2145
    },
    {
      "epoch": 1.6099024756189046,
      "grad_norm": 0.1304258108139038,
      "learning_rate": 1.1151631477841584e-05,
      "loss": 0.0211,
      "step": 2146
    },
    {
      "epoch": 1.6106526631657916,
      "grad_norm": 0.16122934222221375,
      "learning_rate": 1.1110444271466086e-05,
      "loss": 0.0167,
      "step": 2147
    },
    {
      "epoch": 1.611402850712678,
      "grad_norm": 0.08754117786884308,
      "learning_rate": 1.1069323756849126e-05,
      "loss": 0.0136,
      "step": 2148
    },
    {
      "epoch": 1.612153038259565,
      "grad_norm": 0.1655844748020172,
      "learning_rate": 1.102827000450835e-05,
      "loss": 0.0199,
      "step": 2149
    },
    {
      "epoch": 1.6129032258064515,
      "grad_norm": 0.1335715353488922,
      "learning_rate": 1.0987283084846905e-05,
      "loss": 0.0127,
      "step": 2150
    },
    {
      "epoch": 1.6136534133533385,
      "grad_norm": 0.0860769972205162,
      "learning_rate": 1.0946363068153343e-05,
      "loss": 0.0114,
      "step": 2151
    },
    {
      "epoch": 1.614403600900225,
      "grad_norm": 0.1350153535604477,
      "learning_rate": 1.0905510024601423e-05,
      "loss": 0.0161,
      "step": 2152
    },
    {
      "epoch": 1.6151537884471119,
      "grad_norm": 0.11918716132640839,
      "learning_rate": 1.0864724024250106e-05,
      "loss": 0.0147,
      "step": 2153
    },
    {
      "epoch": 1.6159039759939984,
      "grad_norm": 0.0758005753159523,
      "learning_rate": 1.0824005137043375e-05,
      "loss": 0.0108,
      "step": 2154
    },
    {
      "epoch": 1.6166541635408853,
      "grad_norm": 0.11765742301940918,
      "learning_rate": 1.0783353432810106e-05,
      "loss": 0.0175,
      "step": 2155
    },
    {
      "epoch": 1.6174043510877718,
      "grad_norm": 0.24080726504325867,
      "learning_rate": 1.0742768981263984e-05,
      "loss": 0.0417,
      "step": 2156
    },
    {
      "epoch": 1.6181545386346587,
      "grad_norm": 0.07509870827198029,
      "learning_rate": 1.070225185200331e-05,
      "loss": 0.0099,
      "step": 2157
    },
    {
      "epoch": 1.6189047261815452,
      "grad_norm": 0.12462197244167328,
      "learning_rate": 1.0661802114511005e-05,
      "loss": 0.0222,
      "step": 2158
    },
    {
      "epoch": 1.6196549137284322,
      "grad_norm": 0.1365789920091629,
      "learning_rate": 1.062141983815439e-05,
      "loss": 0.0189,
      "step": 2159
    },
    {
      "epoch": 1.6204051012753187,
      "grad_norm": 0.1191381961107254,
      "learning_rate": 1.0581105092185062e-05,
      "loss": 0.0191,
      "step": 2160
    },
    {
      "epoch": 1.6211552888222056,
      "grad_norm": 0.07630721479654312,
      "learning_rate": 1.0540857945738852e-05,
      "loss": 0.0103,
      "step": 2161
    },
    {
      "epoch": 1.6219054763690923,
      "grad_norm": 0.08390341699123383,
      "learning_rate": 1.0500678467835662e-05,
      "loss": 0.0105,
      "step": 2162
    },
    {
      "epoch": 1.622655663915979,
      "grad_norm": 0.10651591420173645,
      "learning_rate": 1.0460566727379335e-05,
      "loss": 0.0187,
      "step": 2163
    },
    {
      "epoch": 1.6234058514628658,
      "grad_norm": 0.122169628739357,
      "learning_rate": 1.0420522793157567e-05,
      "loss": 0.0166,
      "step": 2164
    },
    {
      "epoch": 1.6241560390097525,
      "grad_norm": 0.1749759167432785,
      "learning_rate": 1.038054673384174e-05,
      "loss": 0.0227,
      "step": 2165
    },
    {
      "epoch": 1.6249062265566392,
      "grad_norm": 0.10475295782089233,
      "learning_rate": 1.0340638617986864e-05,
      "loss": 0.01,
      "step": 2166
    },
    {
      "epoch": 1.625656414103526,
      "grad_norm": 0.09955000877380371,
      "learning_rate": 1.030079851403144e-05,
      "loss": 0.0124,
      "step": 2167
    },
    {
      "epoch": 1.6264066016504126,
      "grad_norm": 0.15151360630989075,
      "learning_rate": 1.0261026490297315e-05,
      "loss": 0.0179,
      "step": 2168
    },
    {
      "epoch": 1.6271567891972993,
      "grad_norm": 0.11382818967103958,
      "learning_rate": 1.022132261498961e-05,
      "loss": 0.0176,
      "step": 2169
    },
    {
      "epoch": 1.627906976744186,
      "grad_norm": 0.15799665451049805,
      "learning_rate": 1.0181686956196529e-05,
      "loss": 0.0214,
      "step": 2170
    },
    {
      "epoch": 1.6286571642910728,
      "grad_norm": 0.11197953671216965,
      "learning_rate": 1.0142119581889332e-05,
      "loss": 0.0144,
      "step": 2171
    },
    {
      "epoch": 1.6294073518379595,
      "grad_norm": 0.19829680025577545,
      "learning_rate": 1.0102620559922204e-05,
      "loss": 0.023,
      "step": 2172
    },
    {
      "epoch": 1.6301575393848462,
      "grad_norm": 0.11381664127111435,
      "learning_rate": 1.0063189958032043e-05,
      "loss": 0.0145,
      "step": 2173
    },
    {
      "epoch": 1.630907726931733,
      "grad_norm": 0.2269834578037262,
      "learning_rate": 1.0023827843838457e-05,
      "loss": 0.0211,
      "step": 2174
    },
    {
      "epoch": 1.6316579144786196,
      "grad_norm": 0.11999978125095367,
      "learning_rate": 9.984534284843594e-06,
      "loss": 0.0166,
      "step": 2175
    },
    {
      "epoch": 1.6324081020255063,
      "grad_norm": 0.12092548608779907,
      "learning_rate": 9.945309348432047e-06,
      "loss": 0.0197,
      "step": 2176
    },
    {
      "epoch": 1.633158289572393,
      "grad_norm": 0.10755470395088196,
      "learning_rate": 9.906153101870725e-06,
      "loss": 0.0127,
      "step": 2177
    },
    {
      "epoch": 1.6339084771192798,
      "grad_norm": 0.17812775075435638,
      "learning_rate": 9.867065612308713e-06,
      "loss": 0.0176,
      "step": 2178
    },
    {
      "epoch": 1.6346586646661665,
      "grad_norm": 0.08771397173404694,
      "learning_rate": 9.82804694677722e-06,
      "loss": 0.0158,
      "step": 2179
    },
    {
      "epoch": 1.6354088522130532,
      "grad_norm": 0.14670372009277344,
      "learning_rate": 9.78909717218941e-06,
      "loss": 0.0181,
      "step": 2180
    },
    {
      "epoch": 1.63615903975994,
      "grad_norm": 0.0950799211859703,
      "learning_rate": 9.75021635534033e-06,
      "loss": 0.0127,
      "step": 2181
    },
    {
      "epoch": 1.6369092273068269,
      "grad_norm": 0.10451541095972061,
      "learning_rate": 9.711404562906717e-06,
      "loss": 0.016,
      "step": 2182
    },
    {
      "epoch": 1.6376594148537134,
      "grad_norm": 0.21459363400936127,
      "learning_rate": 9.672661861447002e-06,
      "loss": 0.029,
      "step": 2183
    },
    {
      "epoch": 1.6384096024006003,
      "grad_norm": 0.161620631814003,
      "learning_rate": 9.633988317401087e-06,
      "loss": 0.019,
      "step": 2184
    },
    {
      "epoch": 1.6391597899474868,
      "grad_norm": 0.17189188301563263,
      "learning_rate": 9.595383997090302e-06,
      "loss": 0.0243,
      "step": 2185
    },
    {
      "epoch": 1.6399099774943737,
      "grad_norm": 0.08551929891109467,
      "learning_rate": 9.556848966717247e-06,
      "loss": 0.0118,
      "step": 2186
    },
    {
      "epoch": 1.6406601650412602,
      "grad_norm": 0.12630008161067963,
      "learning_rate": 9.518383292365713e-06,
      "loss": 0.0155,
      "step": 2187
    },
    {
      "epoch": 1.6414103525881472,
      "grad_norm": 0.08629219233989716,
      "learning_rate": 9.479987040000538e-06,
      "loss": 0.012,
      "step": 2188
    },
    {
      "epoch": 1.6421605401350337,
      "grad_norm": 0.12059628218412399,
      "learning_rate": 9.441660275467512e-06,
      "loss": 0.0151,
      "step": 2189
    },
    {
      "epoch": 1.6429107276819206,
      "grad_norm": 0.13245142996311188,
      "learning_rate": 9.403403064493282e-06,
      "loss": 0.0161,
      "step": 2190
    },
    {
      "epoch": 1.643660915228807,
      "grad_norm": 0.1327851563692093,
      "learning_rate": 9.365215472685163e-06,
      "loss": 0.0209,
      "step": 2191
    },
    {
      "epoch": 1.644411102775694,
      "grad_norm": 0.14042188227176666,
      "learning_rate": 9.32709756553114e-06,
      "loss": 0.0175,
      "step": 2192
    },
    {
      "epoch": 1.6451612903225805,
      "grad_norm": 0.1419353187084198,
      "learning_rate": 9.289049408399659e-06,
      "loss": 0.0141,
      "step": 2193
    },
    {
      "epoch": 1.6459114778694675,
      "grad_norm": 0.10731997340917587,
      "learning_rate": 9.251071066539579e-06,
      "loss": 0.0153,
      "step": 2194
    },
    {
      "epoch": 1.646661665416354,
      "grad_norm": 0.13512557744979858,
      "learning_rate": 9.21316260507999e-06,
      "loss": 0.0164,
      "step": 2195
    },
    {
      "epoch": 1.6474118529632409,
      "grad_norm": 0.14420804381370544,
      "learning_rate": 9.175324089030185e-06,
      "loss": 0.0187,
      "step": 2196
    },
    {
      "epoch": 1.6481620405101274,
      "grad_norm": 0.1138468086719513,
      "learning_rate": 9.137555583279495e-06,
      "loss": 0.0138,
      "step": 2197
    },
    {
      "epoch": 1.6489122280570143,
      "grad_norm": 0.26051953434944153,
      "learning_rate": 9.099857152597185e-06,
      "loss": 0.0229,
      "step": 2198
    },
    {
      "epoch": 1.6496624156039008,
      "grad_norm": 0.13202644884586334,
      "learning_rate": 9.062228861632354e-06,
      "loss": 0.0155,
      "step": 2199
    },
    {
      "epoch": 1.6504126031507877,
      "grad_norm": 0.11023962497711182,
      "learning_rate": 9.024670774913812e-06,
      "loss": 0.0172,
      "step": 2200
    },
    {
      "epoch": 1.6504126031507877,
      "eval_loss": 0.03166023641824722,
      "eval_runtime": 5.1277,
      "eval_samples_per_second": 10.531,
      "eval_steps_per_second": 2.73,
      "step": 2200
    },
    {
      "epoch": 1.6511627906976745,
      "grad_norm": 0.1747378408908844,
      "learning_rate": 8.987182956849983e-06,
      "loss": 0.0208,
      "step": 2201
    },
    {
      "epoch": 1.6519129782445612,
      "grad_norm": 0.104463130235672,
      "learning_rate": 8.949765471728789e-06,
      "loss": 0.0149,
      "step": 2202
    },
    {
      "epoch": 1.652663165791448,
      "grad_norm": 0.25243860483169556,
      "learning_rate": 8.912418383717513e-06,
      "loss": 0.0316,
      "step": 2203
    },
    {
      "epoch": 1.6534133533383346,
      "grad_norm": 0.1255129724740982,
      "learning_rate": 8.875141756862749e-06,
      "loss": 0.0157,
      "step": 2204
    },
    {
      "epoch": 1.6541635408852213,
      "grad_norm": 0.13603879511356354,
      "learning_rate": 8.837935655090241e-06,
      "loss": 0.0159,
      "step": 2205
    },
    {
      "epoch": 1.654913728432108,
      "grad_norm": 0.1283695250749588,
      "learning_rate": 8.800800142204779e-06,
      "loss": 0.0151,
      "step": 2206
    },
    {
      "epoch": 1.6556639159789948,
      "grad_norm": 0.0896884873509407,
      "learning_rate": 8.763735281890133e-06,
      "loss": 0.01,
      "step": 2207
    },
    {
      "epoch": 1.6564141035258815,
      "grad_norm": 0.08420361578464508,
      "learning_rate": 8.726741137708866e-06,
      "loss": 0.0089,
      "step": 2208
    },
    {
      "epoch": 1.6571642910727682,
      "grad_norm": 0.09029056876897812,
      "learning_rate": 8.689817773102293e-06,
      "loss": 0.0095,
      "step": 2209
    },
    {
      "epoch": 1.657914478619655,
      "grad_norm": 0.14621016383171082,
      "learning_rate": 8.65296525139036e-06,
      "loss": 0.0178,
      "step": 2210
    },
    {
      "epoch": 1.6586646661665416,
      "grad_norm": 0.19486641883850098,
      "learning_rate": 8.616183635771525e-06,
      "loss": 0.0121,
      "step": 2211
    },
    {
      "epoch": 1.6594148537134283,
      "grad_norm": 0.12832102179527283,
      "learning_rate": 8.579472989322602e-06,
      "loss": 0.0133,
      "step": 2212
    },
    {
      "epoch": 1.660165041260315,
      "grad_norm": 0.06961192190647125,
      "learning_rate": 8.542833374998744e-06,
      "loss": 0.0086,
      "step": 2213
    },
    {
      "epoch": 1.6609152288072018,
      "grad_norm": 0.23715099692344666,
      "learning_rate": 8.5062648556333e-06,
      "loss": 0.0268,
      "step": 2214
    },
    {
      "epoch": 1.6616654163540885,
      "grad_norm": 0.09155306220054626,
      "learning_rate": 8.469767493937681e-06,
      "loss": 0.0163,
      "step": 2215
    },
    {
      "epoch": 1.6624156039009752,
      "grad_norm": 0.17014844715595245,
      "learning_rate": 8.43334135250125e-06,
      "loss": 0.0207,
      "step": 2216
    },
    {
      "epoch": 1.663165791447862,
      "grad_norm": 0.14714054763317108,
      "learning_rate": 8.39698649379126e-06,
      "loss": 0.0197,
      "step": 2217
    },
    {
      "epoch": 1.6639159789947486,
      "grad_norm": 0.18474572896957397,
      "learning_rate": 8.360702980152713e-06,
      "loss": 0.0242,
      "step": 2218
    },
    {
      "epoch": 1.6646661665416356,
      "grad_norm": 0.20110833644866943,
      "learning_rate": 8.32449087380826e-06,
      "loss": 0.0233,
      "step": 2219
    },
    {
      "epoch": 1.665416354088522,
      "grad_norm": 0.09897205233573914,
      "learning_rate": 8.288350236858117e-06,
      "loss": 0.0157,
      "step": 2220
    },
    {
      "epoch": 1.666166541635409,
      "grad_norm": 0.1320415437221527,
      "learning_rate": 8.252281131279887e-06,
      "loss": 0.0159,
      "step": 2221
    },
    {
      "epoch": 1.6669167291822955,
      "grad_norm": 0.13419120013713837,
      "learning_rate": 8.21628361892855e-06,
      "loss": 0.0202,
      "step": 2222
    },
    {
      "epoch": 1.6676669167291824,
      "grad_norm": 0.11188473552465439,
      "learning_rate": 8.180357761536296e-06,
      "loss": 0.0142,
      "step": 2223
    },
    {
      "epoch": 1.668417104276069,
      "grad_norm": 0.15787629783153534,
      "learning_rate": 8.14450362071244e-06,
      "loss": 0.0256,
      "step": 2224
    },
    {
      "epoch": 1.6691672918229559,
      "grad_norm": 0.14580890536308289,
      "learning_rate": 8.10872125794328e-06,
      "loss": 0.0241,
      "step": 2225
    },
    {
      "epoch": 1.6699174793698424,
      "grad_norm": 0.08850288391113281,
      "learning_rate": 8.073010734592057e-06,
      "loss": 0.0089,
      "step": 2226
    },
    {
      "epoch": 1.6706676669167293,
      "grad_norm": 0.14631128311157227,
      "learning_rate": 8.037372111898789e-06,
      "loss": 0.0124,
      "step": 2227
    },
    {
      "epoch": 1.6714178544636158,
      "grad_norm": 0.14625024795532227,
      "learning_rate": 8.001805450980249e-06,
      "loss": 0.0152,
      "step": 2228
    },
    {
      "epoch": 1.6721680420105027,
      "grad_norm": 0.0920761376619339,
      "learning_rate": 7.966310812829709e-06,
      "loss": 0.0133,
      "step": 2229
    },
    {
      "epoch": 1.6729182295573892,
      "grad_norm": 0.2126186639070511,
      "learning_rate": 7.930888258316998e-06,
      "loss": 0.0243,
      "step": 2230
    },
    {
      "epoch": 1.6736684171042762,
      "grad_norm": 0.16171617805957794,
      "learning_rate": 7.89553784818831e-06,
      "loss": 0.028,
      "step": 2231
    },
    {
      "epoch": 1.6744186046511627,
      "grad_norm": 0.09965240210294724,
      "learning_rate": 7.860259643066126e-06,
      "loss": 0.0111,
      "step": 2232
    },
    {
      "epoch": 1.6751687921980496,
      "grad_norm": 0.13482119143009186,
      "learning_rate": 7.82505370344907e-06,
      "loss": 0.0223,
      "step": 2233
    },
    {
      "epoch": 1.675918979744936,
      "grad_norm": 0.07899262011051178,
      "learning_rate": 7.789920089711871e-06,
      "loss": 0.0096,
      "step": 2234
    },
    {
      "epoch": 1.676669167291823,
      "grad_norm": 0.06677675992250443,
      "learning_rate": 7.754858862105224e-06,
      "loss": 0.0071,
      "step": 2235
    },
    {
      "epoch": 1.6774193548387095,
      "grad_norm": 0.1100536659359932,
      "learning_rate": 7.71987008075568e-06,
      "loss": 0.0104,
      "step": 2236
    },
    {
      "epoch": 1.6781695423855965,
      "grad_norm": 0.12148049473762512,
      "learning_rate": 7.684953805665562e-06,
      "loss": 0.0115,
      "step": 2237
    },
    {
      "epoch": 1.6789197299324832,
      "grad_norm": 0.08337541669607162,
      "learning_rate": 7.65011009671282e-06,
      "loss": 0.0124,
      "step": 2238
    },
    {
      "epoch": 1.67966991747937,
      "grad_norm": 0.1657513678073883,
      "learning_rate": 7.615339013651001e-06,
      "loss": 0.023,
      "step": 2239
    },
    {
      "epoch": 1.6804201050262566,
      "grad_norm": 0.07674112915992737,
      "learning_rate": 7.580640616109081e-06,
      "loss": 0.0087,
      "step": 2240
    },
    {
      "epoch": 1.6811702925731433,
      "grad_norm": 0.17678917944431305,
      "learning_rate": 7.546014963591397e-06,
      "loss": 0.0206,
      "step": 2241
    },
    {
      "epoch": 1.68192048012003,
      "grad_norm": 0.14089573919773102,
      "learning_rate": 7.511462115477536e-06,
      "loss": 0.0194,
      "step": 2242
    },
    {
      "epoch": 1.6826706676669168,
      "grad_norm": 0.17028793692588806,
      "learning_rate": 7.476982131022231e-06,
      "loss": 0.0174,
      "step": 2243
    },
    {
      "epoch": 1.6834208552138035,
      "grad_norm": 0.10286597907543182,
      "learning_rate": 7.442575069355256e-06,
      "loss": 0.0115,
      "step": 2244
    },
    {
      "epoch": 1.6841710427606902,
      "grad_norm": 0.19301265478134155,
      "learning_rate": 7.408240989481347e-06,
      "loss": 0.0204,
      "step": 2245
    },
    {
      "epoch": 1.684921230307577,
      "grad_norm": 0.15151479840278625,
      "learning_rate": 7.373979950280046e-06,
      "loss": 0.0174,
      "step": 2246
    },
    {
      "epoch": 1.6856714178544636,
      "grad_norm": 0.0994754210114479,
      "learning_rate": 7.33979201050568e-06,
      "loss": 0.0105,
      "step": 2247
    },
    {
      "epoch": 1.6864216054013503,
      "grad_norm": 0.19209209084510803,
      "learning_rate": 7.3056772287871886e-06,
      "loss": 0.0195,
      "step": 2248
    },
    {
      "epoch": 1.687171792948237,
      "grad_norm": 0.15901431441307068,
      "learning_rate": 7.2716356636280684e-06,
      "loss": 0.0182,
      "step": 2249
    },
    {
      "epoch": 1.6879219804951238,
      "grad_norm": 0.10334349423646927,
      "learning_rate": 7.237667373406259e-06,
      "loss": 0.0116,
      "step": 2250
    },
    {
      "epoch": 1.6886721680420105,
      "grad_norm": 0.10797673463821411,
      "learning_rate": 7.203772416374016e-06,
      "loss": 0.0178,
      "step": 2251
    },
    {
      "epoch": 1.6894223555888972,
      "grad_norm": 0.13973666727542877,
      "learning_rate": 7.1699508506578636e-06,
      "loss": 0.0171,
      "step": 2252
    },
    {
      "epoch": 1.690172543135784,
      "grad_norm": 0.13739119470119476,
      "learning_rate": 7.136202734258457e-06,
      "loss": 0.018,
      "step": 2253
    },
    {
      "epoch": 1.6909227306826706,
      "grad_norm": 0.19144482910633087,
      "learning_rate": 7.1025281250505006e-06,
      "loss": 0.024,
      "step": 2254
    },
    {
      "epoch": 1.6916729182295573,
      "grad_norm": 0.12330687791109085,
      "learning_rate": 7.0689270807826e-06,
      "loss": 0.0158,
      "step": 2255
    },
    {
      "epoch": 1.692423105776444,
      "grad_norm": 0.10102571547031403,
      "learning_rate": 7.035399659077268e-06,
      "loss": 0.0167,
      "step": 2256
    },
    {
      "epoch": 1.6931732933233308,
      "grad_norm": 0.15109023451805115,
      "learning_rate": 7.00194591743073e-06,
      "loss": 0.0246,
      "step": 2257
    },
    {
      "epoch": 1.6939234808702177,
      "grad_norm": 0.0868748277425766,
      "learning_rate": 6.96856591321286e-06,
      "loss": 0.0093,
      "step": 2258
    },
    {
      "epoch": 1.6946736684171042,
      "grad_norm": 0.11420181393623352,
      "learning_rate": 6.9352597036670575e-06,
      "loss": 0.0186,
      "step": 2259
    },
    {
      "epoch": 1.6954238559639911,
      "grad_norm": 0.10863950848579407,
      "learning_rate": 6.902027345910211e-06,
      "loss": 0.0136,
      "step": 2260
    },
    {
      "epoch": 1.6961740435108776,
      "grad_norm": 0.13581328094005585,
      "learning_rate": 6.868868896932534e-06,
      "loss": 0.0224,
      "step": 2261
    },
    {
      "epoch": 1.6969242310577646,
      "grad_norm": 0.13632053136825562,
      "learning_rate": 6.835784413597512e-06,
      "loss": 0.0204,
      "step": 2262
    },
    {
      "epoch": 1.697674418604651,
      "grad_norm": 0.1366109400987625,
      "learning_rate": 6.802773952641761e-06,
      "loss": 0.0173,
      "step": 2263
    },
    {
      "epoch": 1.698424606151538,
      "grad_norm": 0.08851693570613861,
      "learning_rate": 6.769837570674975e-06,
      "loss": 0.013,
      "step": 2264
    },
    {
      "epoch": 1.6991747936984245,
      "grad_norm": 0.11936666071414948,
      "learning_rate": 6.7369753241798114e-06,
      "loss": 0.0136,
      "step": 2265
    },
    {
      "epoch": 1.6999249812453114,
      "grad_norm": 0.0834725946187973,
      "learning_rate": 6.70418726951178e-06,
      "loss": 0.0108,
      "step": 2266
    },
    {
      "epoch": 1.700675168792198,
      "grad_norm": 0.12028200924396515,
      "learning_rate": 6.671473462899181e-06,
      "loss": 0.014,
      "step": 2267
    },
    {
      "epoch": 1.7014253563390849,
      "grad_norm": 0.20901398360729218,
      "learning_rate": 6.638833960442948e-06,
      "loss": 0.0205,
      "step": 2268
    },
    {
      "epoch": 1.7021755438859714,
      "grad_norm": 0.09815794229507446,
      "learning_rate": 6.606268818116618e-06,
      "loss": 0.0138,
      "step": 2269
    },
    {
      "epoch": 1.7029257314328583,
      "grad_norm": 0.17781807482242584,
      "learning_rate": 6.573778091766219e-06,
      "loss": 0.0199,
      "step": 2270
    },
    {
      "epoch": 1.7036759189797448,
      "grad_norm": 0.10289303958415985,
      "learning_rate": 6.541361837110149e-06,
      "loss": 0.012,
      "step": 2271
    },
    {
      "epoch": 1.7044261065266317,
      "grad_norm": 0.08833058178424835,
      "learning_rate": 6.509020109739078e-06,
      "loss": 0.0134,
      "step": 2272
    },
    {
      "epoch": 1.7051762940735182,
      "grad_norm": 0.12388486415147781,
      "learning_rate": 6.476752965115884e-06,
      "loss": 0.0121,
      "step": 2273
    },
    {
      "epoch": 1.7059264816204052,
      "grad_norm": 0.1161249652504921,
      "learning_rate": 6.444560458575544e-06,
      "loss": 0.0154,
      "step": 2274
    },
    {
      "epoch": 1.7066766691672917,
      "grad_norm": 0.11769194155931473,
      "learning_rate": 6.412442645325057e-06,
      "loss": 0.0133,
      "step": 2275
    },
    {
      "epoch": 1.7074268567141786,
      "grad_norm": 0.08280669152736664,
      "learning_rate": 6.38039958044328e-06,
      "loss": 0.0098,
      "step": 2276
    },
    {
      "epoch": 1.7081770442610653,
      "grad_norm": 0.10545045137405396,
      "learning_rate": 6.3484313188809265e-06,
      "loss": 0.0156,
      "step": 2277
    },
    {
      "epoch": 1.708927231807952,
      "grad_norm": 0.20702876150608063,
      "learning_rate": 6.316537915460418e-06,
      "loss": 0.0305,
      "step": 2278
    },
    {
      "epoch": 1.7096774193548387,
      "grad_norm": 0.09845060855150223,
      "learning_rate": 6.284719424875796e-06,
      "loss": 0.0136,
      "step": 2279
    },
    {
      "epoch": 1.7104276069017255,
      "grad_norm": 0.10509305447340012,
      "learning_rate": 6.252975901692659e-06,
      "loss": 0.0132,
      "step": 2280
    },
    {
      "epoch": 1.7111777944486122,
      "grad_norm": 0.08033392578363419,
      "learning_rate": 6.221307400347992e-06,
      "loss": 0.0082,
      "step": 2281
    },
    {
      "epoch": 1.711927981995499,
      "grad_norm": 0.11224459111690521,
      "learning_rate": 6.1897139751501796e-06,
      "loss": 0.0183,
      "step": 2282
    },
    {
      "epoch": 1.7126781695423856,
      "grad_norm": 0.09149748831987381,
      "learning_rate": 6.158195680278816e-06,
      "loss": 0.0117,
      "step": 2283
    },
    {
      "epoch": 1.7134283570892723,
      "grad_norm": 0.14243397116661072,
      "learning_rate": 6.126752569784694e-06,
      "loss": 0.0186,
      "step": 2284
    },
    {
      "epoch": 1.714178544636159,
      "grad_norm": 0.11700420081615448,
      "learning_rate": 6.095384697589635e-06,
      "loss": 0.0132,
      "step": 2285
    },
    {
      "epoch": 1.7149287321830458,
      "grad_norm": 0.10646030306816101,
      "learning_rate": 6.064092117486464e-06,
      "loss": 0.0152,
      "step": 2286
    },
    {
      "epoch": 1.7156789197299325,
      "grad_norm": 0.14706531167030334,
      "learning_rate": 6.032874883138867e-06,
      "loss": 0.0159,
      "step": 2287
    },
    {
      "epoch": 1.7164291072768192,
      "grad_norm": 0.12105236947536469,
      "learning_rate": 6.001733048081337e-06,
      "loss": 0.0196,
      "step": 2288
    },
    {
      "epoch": 1.717179294823706,
      "grad_norm": 0.17891399562358856,
      "learning_rate": 5.970666665719033e-06,
      "loss": 0.0197,
      "step": 2289
    },
    {
      "epoch": 1.7179294823705926,
      "grad_norm": 0.16054308414459229,
      "learning_rate": 5.939675789327759e-06,
      "loss": 0.0205,
      "step": 2290
    },
    {
      "epoch": 1.7186796699174793,
      "grad_norm": 0.1569271832704544,
      "learning_rate": 5.908760472053809e-06,
      "loss": 0.0238,
      "step": 2291
    },
    {
      "epoch": 1.719429857464366,
      "grad_norm": 0.13982658088207245,
      "learning_rate": 5.877920766913919e-06,
      "loss": 0.0146,
      "step": 2292
    },
    {
      "epoch": 1.7201800450112528,
      "grad_norm": 0.1269511878490448,
      "learning_rate": 5.847156726795133e-06,
      "loss": 0.0174,
      "step": 2293
    },
    {
      "epoch": 1.7209302325581395,
      "grad_norm": 0.12216448783874512,
      "learning_rate": 5.816468404454755e-06,
      "loss": 0.0225,
      "step": 2294
    },
    {
      "epoch": 1.7216804201050264,
      "grad_norm": 0.10995414108037949,
      "learning_rate": 5.7858558525202336e-06,
      "loss": 0.011,
      "step": 2295
    },
    {
      "epoch": 1.722430607651913,
      "grad_norm": 0.10215462744235992,
      "learning_rate": 5.755319123489083e-06,
      "loss": 0.0138,
      "step": 2296
    },
    {
      "epoch": 1.7231807951987999,
      "grad_norm": 0.07460669428110123,
      "learning_rate": 5.724858269728789e-06,
      "loss": 0.0107,
      "step": 2297
    },
    {
      "epoch": 1.7239309827456863,
      "grad_norm": 0.13698826730251312,
      "learning_rate": 5.694473343476714e-06,
      "loss": 0.0137,
      "step": 2298
    },
    {
      "epoch": 1.7246811702925733,
      "grad_norm": 0.11854610592126846,
      "learning_rate": 5.664164396840016e-06,
      "loss": 0.0171,
      "step": 2299
    },
    {
      "epoch": 1.7254313578394598,
      "grad_norm": 0.06490737944841385,
      "learning_rate": 5.633931481795552e-06,
      "loss": 0.0068,
      "step": 2300
    },
    {
      "epoch": 1.7261815453863467,
      "grad_norm": 0.08114010840654373,
      "learning_rate": 5.603774650189808e-06,
      "loss": 0.0108,
      "step": 2301
    },
    {
      "epoch": 1.7269317329332332,
      "grad_norm": 0.13306188583374023,
      "learning_rate": 5.573693953738751e-06,
      "loss": 0.0199,
      "step": 2302
    },
    {
      "epoch": 1.7276819204801201,
      "grad_norm": 0.09569033980369568,
      "learning_rate": 5.543689444027839e-06,
      "loss": 0.0108,
      "step": 2303
    },
    {
      "epoch": 1.7284321080270066,
      "grad_norm": 0.08656707406044006,
      "learning_rate": 5.513761172511833e-06,
      "loss": 0.0111,
      "step": 2304
    },
    {
      "epoch": 1.7291822955738936,
      "grad_norm": 0.14147700369358063,
      "learning_rate": 5.483909190514797e-06,
      "loss": 0.0163,
      "step": 2305
    },
    {
      "epoch": 1.72993248312078,
      "grad_norm": 0.0972064808011055,
      "learning_rate": 5.4541335492299115e-06,
      "loss": 0.0119,
      "step": 2306
    },
    {
      "epoch": 1.730682670667667,
      "grad_norm": 0.14779311418533325,
      "learning_rate": 5.424434299719483e-06,
      "loss": 0.0194,
      "step": 2307
    },
    {
      "epoch": 1.7314328582145535,
      "grad_norm": 0.09764081239700317,
      "learning_rate": 5.394811492914803e-06,
      "loss": 0.0146,
      "step": 2308
    },
    {
      "epoch": 1.7321830457614404,
      "grad_norm": 0.08887841552495956,
      "learning_rate": 5.365265179616063e-06,
      "loss": 0.0102,
      "step": 2309
    },
    {
      "epoch": 1.732933233308327,
      "grad_norm": 0.141043022274971,
      "learning_rate": 5.3357954104922895e-06,
      "loss": 0.0152,
      "step": 2310
    },
    {
      "epoch": 1.7336834208552139,
      "grad_norm": 0.16252657771110535,
      "learning_rate": 5.306402236081209e-06,
      "loss": 0.0349,
      "step": 2311
    },
    {
      "epoch": 1.7344336084021004,
      "grad_norm": 0.09353138506412506,
      "learning_rate": 5.277085706789248e-06,
      "loss": 0.0127,
      "step": 2312
    },
    {
      "epoch": 1.7351837959489873,
      "grad_norm": 0.09399595111608505,
      "learning_rate": 5.247845872891371e-06,
      "loss": 0.0103,
      "step": 2313
    },
    {
      "epoch": 1.735933983495874,
      "grad_norm": 0.10673192143440247,
      "learning_rate": 5.218682784530993e-06,
      "loss": 0.0151,
      "step": 2314
    },
    {
      "epoch": 1.7366841710427607,
      "grad_norm": 0.2325008064508438,
      "learning_rate": 5.1895964917199445e-06,
      "loss": 0.0121,
      "step": 2315
    },
    {
      "epoch": 1.7374343585896475,
      "grad_norm": 0.1543661206960678,
      "learning_rate": 5.160587044338355e-06,
      "loss": 0.0186,
      "step": 2316
    },
    {
      "epoch": 1.7381845461365342,
      "grad_norm": 0.11611554026603699,
      "learning_rate": 5.131654492134574e-06,
      "loss": 0.0134,
      "step": 2317
    },
    {
      "epoch": 1.7389347336834209,
      "grad_norm": 0.09432833641767502,
      "learning_rate": 5.102798884725091e-06,
      "loss": 0.0108,
      "step": 2318
    },
    {
      "epoch": 1.7396849212303076,
      "grad_norm": 0.1359938085079193,
      "learning_rate": 5.074020271594404e-06,
      "loss": 0.0198,
      "step": 2319
    },
    {
      "epoch": 1.7404351087771943,
      "grad_norm": 0.1561252623796463,
      "learning_rate": 5.045318702095014e-06,
      "loss": 0.0232,
      "step": 2320
    },
    {
      "epoch": 1.741185296324081,
      "grad_norm": 0.13938003778457642,
      "learning_rate": 5.016694225447288e-06,
      "loss": 0.0197,
      "step": 2321
    },
    {
      "epoch": 1.7419354838709677,
      "grad_norm": 0.13493551313877106,
      "learning_rate": 4.988146890739381e-06,
      "loss": 0.0214,
      "step": 2322
    },
    {
      "epoch": 1.7426856714178545,
      "grad_norm": 0.09438906610012054,
      "learning_rate": 4.959676746927172e-06,
      "loss": 0.0125,
      "step": 2323
    },
    {
      "epoch": 1.7434358589647412,
      "grad_norm": 0.26109787821769714,
      "learning_rate": 4.931283842834139e-06,
      "loss": 0.0148,
      "step": 2324
    },
    {
      "epoch": 1.744186046511628,
      "grad_norm": 0.12471003085374832,
      "learning_rate": 4.902968227151311e-06,
      "loss": 0.0185,
      "step": 2325
    },
    {
      "epoch": 1.7449362340585146,
      "grad_norm": 0.08934643864631653,
      "learning_rate": 4.874729948437218e-06,
      "loss": 0.0118,
      "step": 2326
    },
    {
      "epoch": 1.7456864216054013,
      "grad_norm": 0.10696186125278473,
      "learning_rate": 4.846569055117684e-06,
      "loss": 0.0135,
      "step": 2327
    },
    {
      "epoch": 1.746436609152288,
      "grad_norm": 0.14127911627292633,
      "learning_rate": 4.818485595485889e-06,
      "loss": 0.0198,
      "step": 2328
    },
    {
      "epoch": 1.7471867966991748,
      "grad_norm": 0.12543734908103943,
      "learning_rate": 4.790479617702198e-06,
      "loss": 0.0146,
      "step": 2329
    },
    {
      "epoch": 1.7479369842460615,
      "grad_norm": 0.1253226101398468,
      "learning_rate": 4.762551169794105e-06,
      "loss": 0.0136,
      "step": 2330
    },
    {
      "epoch": 1.7486871717929482,
      "grad_norm": 0.15639212727546692,
      "learning_rate": 4.734700299656158e-06,
      "loss": 0.0154,
      "step": 2331
    },
    {
      "epoch": 1.749437359339835,
      "grad_norm": 0.12211638689041138,
      "learning_rate": 4.706927055049837e-06,
      "loss": 0.0151,
      "step": 2332
    },
    {
      "epoch": 1.7501875468867216,
      "grad_norm": 0.06649266928434372,
      "learning_rate": 4.6792314836035304e-06,
      "loss": 0.0099,
      "step": 2333
    },
    {
      "epoch": 1.7509377344336086,
      "grad_norm": 0.07582502067089081,
      "learning_rate": 4.651613632812413e-06,
      "loss": 0.0105,
      "step": 2334
    },
    {
      "epoch": 1.751687921980495,
      "grad_norm": 0.1324513703584671,
      "learning_rate": 4.624073550038399e-06,
      "loss": 0.0168,
      "step": 2335
    },
    {
      "epoch": 1.752438109527382,
      "grad_norm": 0.14669281244277954,
      "learning_rate": 4.596611282509989e-06,
      "loss": 0.0198,
      "step": 2336
    },
    {
      "epoch": 1.7531882970742685,
      "grad_norm": 0.0815398246049881,
      "learning_rate": 4.56922687732228e-06,
      "loss": 0.0161,
      "step": 2337
    },
    {
      "epoch": 1.7539384846211554,
      "grad_norm": 0.10307783633470535,
      "learning_rate": 4.5419203814368376e-06,
      "loss": 0.0128,
      "step": 2338
    },
    {
      "epoch": 1.754688672168042,
      "grad_norm": 0.07888856530189514,
      "learning_rate": 4.514691841681601e-06,
      "loss": 0.0143,
      "step": 2339
    },
    {
      "epoch": 1.7554388597149289,
      "grad_norm": 0.14705821871757507,
      "learning_rate": 4.487541304750848e-06,
      "loss": 0.0205,
      "step": 2340
    },
    {
      "epoch": 1.7561890472618154,
      "grad_norm": 0.11970402300357819,
      "learning_rate": 4.4604688172050605e-06,
      "loss": 0.0152,
      "step": 2341
    },
    {
      "epoch": 1.7569392348087023,
      "grad_norm": 0.1779605597257614,
      "learning_rate": 4.433474425470902e-06,
      "loss": 0.0208,
      "step": 2342
    },
    {
      "epoch": 1.7576894223555888,
      "grad_norm": 0.09036190807819366,
      "learning_rate": 4.406558175841097e-06,
      "loss": 0.0115,
      "step": 2343
    },
    {
      "epoch": 1.7584396099024757,
      "grad_norm": 0.10841822624206543,
      "learning_rate": 4.379720114474351e-06,
      "loss": 0.0153,
      "step": 2344
    },
    {
      "epoch": 1.7591897974493622,
      "grad_norm": 0.1275036334991455,
      "learning_rate": 4.352960287395303e-06,
      "loss": 0.0179,
      "step": 2345
    },
    {
      "epoch": 1.7599399849962492,
      "grad_norm": 0.08594145625829697,
      "learning_rate": 4.3262787404944165e-06,
      "loss": 0.011,
      "step": 2346
    },
    {
      "epoch": 1.7606901725431356,
      "grad_norm": 0.05019707977771759,
      "learning_rate": 4.299675519527929e-06,
      "loss": 0.0056,
      "step": 2347
    },
    {
      "epoch": 1.7614403600900226,
      "grad_norm": 0.1445784568786621,
      "learning_rate": 4.273150670117743e-06,
      "loss": 0.0165,
      "step": 2348
    },
    {
      "epoch": 1.762190547636909,
      "grad_norm": 0.15771323442459106,
      "learning_rate": 4.246704237751342e-06,
      "loss": 0.0145,
      "step": 2349
    },
    {
      "epoch": 1.762940735183796,
      "grad_norm": 0.1449221819639206,
      "learning_rate": 4.220336267781777e-06,
      "loss": 0.0235,
      "step": 2350
    },
    {
      "epoch": 1.7636909227306825,
      "grad_norm": 0.09601754695177078,
      "learning_rate": 4.19404680542751e-06,
      "loss": 0.0154,
      "step": 2351
    },
    {
      "epoch": 1.7644411102775694,
      "grad_norm": 0.0986536517739296,
      "learning_rate": 4.167835895772382e-06,
      "loss": 0.0111,
      "step": 2352
    },
    {
      "epoch": 1.7651912978244562,
      "grad_norm": 0.104599729180336,
      "learning_rate": 4.141703583765522e-06,
      "loss": 0.0103,
      "step": 2353
    },
    {
      "epoch": 1.7659414853713429,
      "grad_norm": 0.08715003728866577,
      "learning_rate": 4.11564991422127e-06,
      "loss": 0.0106,
      "step": 2354
    },
    {
      "epoch": 1.7666916729182296,
      "grad_norm": 0.1697830855846405,
      "learning_rate": 4.0896749318191095e-06,
      "loss": 0.0191,
      "step": 2355
    },
    {
      "epoch": 1.7674418604651163,
      "grad_norm": 0.1439453512430191,
      "learning_rate": 4.06377868110358e-06,
      "loss": 0.0157,
      "step": 2356
    },
    {
      "epoch": 1.768192048012003,
      "grad_norm": 0.08522556722164154,
      "learning_rate": 4.037961206484186e-06,
      "loss": 0.0118,
      "step": 2357
    },
    {
      "epoch": 1.7689422355588897,
      "grad_norm": 0.1324508637189865,
      "learning_rate": 4.0122225522353675e-06,
      "loss": 0.0156,
      "step": 2358
    },
    {
      "epoch": 1.7696924231057765,
      "grad_norm": 0.12487998604774475,
      "learning_rate": 3.986562762496376e-06,
      "loss": 0.0177,
      "step": 2359
    },
    {
      "epoch": 1.7704426106526632,
      "grad_norm": 0.22810116410255432,
      "learning_rate": 3.9609818812712255e-06,
      "loss": 0.0299,
      "step": 2360
    },
    {
      "epoch": 1.77119279819955,
      "grad_norm": 0.07804132252931595,
      "learning_rate": 3.935479952428611e-06,
      "loss": 0.0093,
      "step": 2361
    },
    {
      "epoch": 1.7719429857464366,
      "grad_norm": 0.18137666583061218,
      "learning_rate": 3.91005701970183e-06,
      "loss": 0.0205,
      "step": 2362
    },
    {
      "epoch": 1.7726931732933233,
      "grad_norm": 0.14881862699985504,
      "learning_rate": 3.8847131266886935e-06,
      "loss": 0.0154,
      "step": 2363
    },
    {
      "epoch": 1.77344336084021,
      "grad_norm": 0.12017820030450821,
      "learning_rate": 3.859448316851505e-06,
      "loss": 0.0178,
      "step": 2364
    },
    {
      "epoch": 1.7741935483870968,
      "grad_norm": 0.13187788426876068,
      "learning_rate": 3.834262633516916e-06,
      "loss": 0.0136,
      "step": 2365
    },
    {
      "epoch": 1.7749437359339835,
      "grad_norm": 0.10610248148441315,
      "learning_rate": 3.8091561198758897e-06,
      "loss": 0.0148,
      "step": 2366
    },
    {
      "epoch": 1.7756939234808702,
      "grad_norm": 0.11325306445360184,
      "learning_rate": 3.784128818983618e-06,
      "loss": 0.0167,
      "step": 2367
    },
    {
      "epoch": 1.776444111027757,
      "grad_norm": 0.08559075742959976,
      "learning_rate": 3.7591807737594743e-06,
      "loss": 0.0129,
      "step": 2368
    },
    {
      "epoch": 1.7771942985746436,
      "grad_norm": 0.10802838951349258,
      "learning_rate": 3.734312026986897e-06,
      "loss": 0.0129,
      "step": 2369
    },
    {
      "epoch": 1.7779444861215303,
      "grad_norm": 0.13777166604995728,
      "learning_rate": 3.7095226213133272e-06,
      "loss": 0.0232,
      "step": 2370
    },
    {
      "epoch": 1.7786946736684173,
      "grad_norm": 0.09481120854616165,
      "learning_rate": 3.6848125992501592e-06,
      "loss": 0.0155,
      "step": 2371
    },
    {
      "epoch": 1.7794448612153038,
      "grad_norm": 0.13231073319911957,
      "learning_rate": 3.6601820031726517e-06,
      "loss": 0.0143,
      "step": 2372
    },
    {
      "epoch": 1.7801950487621907,
      "grad_norm": 0.11420634388923645,
      "learning_rate": 3.6356308753198454e-06,
      "loss": 0.0179,
      "step": 2373
    },
    {
      "epoch": 1.7809452363090772,
      "grad_norm": 0.07495707273483276,
      "learning_rate": 3.6111592577945217e-06,
      "loss": 0.0083,
      "step": 2374
    },
    {
      "epoch": 1.7816954238559641,
      "grad_norm": 0.09842219948768616,
      "learning_rate": 3.586767192563073e-06,
      "loss": 0.0151,
      "step": 2375
    },
    {
      "epoch": 1.7824456114028506,
      "grad_norm": 0.12309426814317703,
      "learning_rate": 3.562454721455505e-06,
      "loss": 0.0169,
      "step": 2376
    },
    {
      "epoch": 1.7831957989497376,
      "grad_norm": 0.0984698012471199,
      "learning_rate": 3.538221886165299e-06,
      "loss": 0.0131,
      "step": 2377
    },
    {
      "epoch": 1.783945986496624,
      "grad_norm": 0.12303908169269562,
      "learning_rate": 3.514068728249398e-06,
      "loss": 0.013,
      "step": 2378
    },
    {
      "epoch": 1.784696174043511,
      "grad_norm": 0.17969577014446259,
      "learning_rate": 3.489995289128073e-06,
      "loss": 0.0373,
      "step": 2379
    },
    {
      "epoch": 1.7854463615903975,
      "grad_norm": 0.10440315306186676,
      "learning_rate": 3.4660016100849126e-06,
      "loss": 0.015,
      "step": 2380
    },
    {
      "epoch": 1.7861965491372844,
      "grad_norm": 0.09079710394144058,
      "learning_rate": 3.442087732266697e-06,
      "loss": 0.0143,
      "step": 2381
    },
    {
      "epoch": 1.786946736684171,
      "grad_norm": 0.15569381415843964,
      "learning_rate": 3.418253696683399e-06,
      "loss": 0.0183,
      "step": 2382
    },
    {
      "epoch": 1.7876969242310579,
      "grad_norm": 0.11904071271419525,
      "learning_rate": 3.3944995442080185e-06,
      "loss": 0.0165,
      "step": 2383
    },
    {
      "epoch": 1.7884471117779444,
      "grad_norm": 0.14972615242004395,
      "learning_rate": 3.3708253155766033e-06,
      "loss": 0.0232,
      "step": 2384
    },
    {
      "epoch": 1.7891972993248313,
      "grad_norm": 0.0957341119647026,
      "learning_rate": 3.347231051388117e-06,
      "loss": 0.0117,
      "step": 2385
    },
    {
      "epoch": 1.7899474868717178,
      "grad_norm": 0.08932828903198242,
      "learning_rate": 3.323716792104403e-06,
      "loss": 0.0091,
      "step": 2386
    },
    {
      "epoch": 1.7906976744186047,
      "grad_norm": 0.11236236989498138,
      "learning_rate": 3.3002825780500957e-06,
      "loss": 0.018,
      "step": 2387
    },
    {
      "epoch": 1.7914478619654912,
      "grad_norm": 0.10578911006450653,
      "learning_rate": 3.276928449412564e-06,
      "loss": 0.0133,
      "step": 2388
    },
    {
      "epoch": 1.7921980495123782,
      "grad_norm": 0.11684674769639969,
      "learning_rate": 3.253654446241844e-06,
      "loss": 0.0164,
      "step": 2389
    },
    {
      "epoch": 1.7929482370592649,
      "grad_norm": 0.0889001339673996,
      "learning_rate": 3.2304606084505585e-06,
      "loss": 0.013,
      "step": 2390
    },
    {
      "epoch": 1.7936984246061516,
      "grad_norm": 0.1431235671043396,
      "learning_rate": 3.2073469758138577e-06,
      "loss": 0.0248,
      "step": 2391
    },
    {
      "epoch": 1.7944486121530383,
      "grad_norm": 0.10257580131292343,
      "learning_rate": 3.18431358796934e-06,
      "loss": 0.0136,
      "step": 2392
    },
    {
      "epoch": 1.795198799699925,
      "grad_norm": 0.08846593648195267,
      "learning_rate": 3.161360484416992e-06,
      "loss": 0.0119,
      "step": 2393
    },
    {
      "epoch": 1.7959489872468117,
      "grad_norm": 0.07945258170366287,
      "learning_rate": 3.1384877045191384e-06,
      "loss": 0.0102,
      "step": 2394
    },
    {
      "epoch": 1.7966991747936985,
      "grad_norm": 0.1506921648979187,
      "learning_rate": 3.1156952875003365e-06,
      "loss": 0.0197,
      "step": 2395
    },
    {
      "epoch": 1.7974493623405852,
      "grad_norm": 0.0959123820066452,
      "learning_rate": 3.0929832724473416e-06,
      "loss": 0.0111,
      "step": 2396
    },
    {
      "epoch": 1.7981995498874719,
      "grad_norm": 0.12383115291595459,
      "learning_rate": 3.0703516983090207e-06,
      "loss": 0.0123,
      "step": 2397
    },
    {
      "epoch": 1.7989497374343586,
      "grad_norm": 0.20673918724060059,
      "learning_rate": 3.0478006038962947e-06,
      "loss": 0.0221,
      "step": 2398
    },
    {
      "epoch": 1.7996999249812453,
      "grad_norm": 0.07872434705495834,
      "learning_rate": 3.0253300278820783e-06,
      "loss": 0.0088,
      "step": 2399
    },
    {
      "epoch": 1.800450112528132,
      "grad_norm": 0.10210002958774567,
      "learning_rate": 3.002940008801186e-06,
      "loss": 0.0133,
      "step": 2400
    },
    {
      "epoch": 1.800450112528132,
      "eval_loss": 0.031127866357564926,
      "eval_runtime": 5.1038,
      "eval_samples_per_second": 10.58,
      "eval_steps_per_second": 2.743,
      "step": 2400
    },
    {
      "epoch": 1.8012003000750187,
      "grad_norm": 0.13143689930438995,
      "learning_rate": 2.9806305850502923e-06,
      "loss": 0.0172,
      "step": 2401
    },
    {
      "epoch": 1.8019504876219055,
      "grad_norm": 0.1445961445569992,
      "learning_rate": 2.9584017948878717e-06,
      "loss": 0.0238,
      "step": 2402
    },
    {
      "epoch": 1.8027006751687922,
      "grad_norm": 0.15010763704776764,
      "learning_rate": 2.9362536764341085e-06,
      "loss": 0.0183,
      "step": 2403
    },
    {
      "epoch": 1.803450862715679,
      "grad_norm": 0.20283788442611694,
      "learning_rate": 2.9141862676708486e-06,
      "loss": 0.0397,
      "step": 2404
    },
    {
      "epoch": 1.8042010502625656,
      "grad_norm": 0.16290506720542908,
      "learning_rate": 2.8921996064415147e-06,
      "loss": 0.0161,
      "step": 2405
    },
    {
      "epoch": 1.8049512378094523,
      "grad_norm": 0.12406005710363388,
      "learning_rate": 2.870293730451068e-06,
      "loss": 0.0143,
      "step": 2406
    },
    {
      "epoch": 1.805701425356339,
      "grad_norm": 0.10236632823944092,
      "learning_rate": 2.8484686772659308e-06,
      "loss": 0.0186,
      "step": 2407
    },
    {
      "epoch": 1.8064516129032258,
      "grad_norm": 0.1057526171207428,
      "learning_rate": 2.826724484313925e-06,
      "loss": 0.0116,
      "step": 2408
    },
    {
      "epoch": 1.8072018004501125,
      "grad_norm": 0.1698257327079773,
      "learning_rate": 2.8050611888841947e-06,
      "loss": 0.0229,
      "step": 2409
    },
    {
      "epoch": 1.8079519879969994,
      "grad_norm": 0.08635108917951584,
      "learning_rate": 2.7834788281271616e-06,
      "loss": 0.0129,
      "step": 2410
    },
    {
      "epoch": 1.808702175543886,
      "grad_norm": 0.11237742751836777,
      "learning_rate": 2.7619774390544473e-06,
      "loss": 0.0132,
      "step": 2411
    },
    {
      "epoch": 1.8094523630907728,
      "grad_norm": 0.0990079864859581,
      "learning_rate": 2.740557058538823e-06,
      "loss": 0.0139,
      "step": 2412
    },
    {
      "epoch": 1.8102025506376593,
      "grad_norm": 0.11661037802696228,
      "learning_rate": 2.7192177233141215e-06,
      "loss": 0.015,
      "step": 2413
    },
    {
      "epoch": 1.8109527381845463,
      "grad_norm": 0.16443532705307007,
      "learning_rate": 2.697959469975203e-06,
      "loss": 0.0227,
      "step": 2414
    },
    {
      "epoch": 1.8117029257314328,
      "grad_norm": 0.12524110078811646,
      "learning_rate": 2.6767823349778843e-06,
      "loss": 0.0145,
      "step": 2415
    },
    {
      "epoch": 1.8124531132783197,
      "grad_norm": 0.09298525005578995,
      "learning_rate": 2.65568635463887e-06,
      "loss": 0.0163,
      "step": 2416
    },
    {
      "epoch": 1.8132033008252062,
      "grad_norm": 0.10674528032541275,
      "learning_rate": 2.634671565135677e-06,
      "loss": 0.0174,
      "step": 2417
    },
    {
      "epoch": 1.8139534883720931,
      "grad_norm": 0.15459661185741425,
      "learning_rate": 2.613738002506605e-06,
      "loss": 0.0147,
      "step": 2418
    },
    {
      "epoch": 1.8147036759189796,
      "grad_norm": 0.11485212296247482,
      "learning_rate": 2.592885702650655e-06,
      "loss": 0.0175,
      "step": 2419
    },
    {
      "epoch": 1.8154538634658666,
      "grad_norm": 0.13821269571781158,
      "learning_rate": 2.572114701327466e-06,
      "loss": 0.0172,
      "step": 2420
    },
    {
      "epoch": 1.816204051012753,
      "grad_norm": 0.10720129311084747,
      "learning_rate": 2.551425034157262e-06,
      "loss": 0.015,
      "step": 2421
    },
    {
      "epoch": 1.81695423855964,
      "grad_norm": 0.10853177309036255,
      "learning_rate": 2.5308167366207724e-06,
      "loss": 0.0195,
      "step": 2422
    },
    {
      "epoch": 1.8177044261065265,
      "grad_norm": 0.10341428220272064,
      "learning_rate": 2.510289844059216e-06,
      "loss": 0.0164,
      "step": 2423
    },
    {
      "epoch": 1.8184546136534134,
      "grad_norm": 0.07772510498762131,
      "learning_rate": 2.48984439167419e-06,
      "loss": 0.012,
      "step": 2424
    },
    {
      "epoch": 1.8192048012003,
      "grad_norm": 0.12696626782417297,
      "learning_rate": 2.4694804145276305e-06,
      "loss": 0.0159,
      "step": 2425
    },
    {
      "epoch": 1.8199549887471869,
      "grad_norm": 0.16353991627693176,
      "learning_rate": 2.449197947541737e-06,
      "loss": 0.019,
      "step": 2426
    },
    {
      "epoch": 1.8207051762940734,
      "grad_norm": 0.1997886449098587,
      "learning_rate": 2.4289970254989635e-06,
      "loss": 0.02,
      "step": 2427
    },
    {
      "epoch": 1.8214553638409603,
      "grad_norm": 0.42609405517578125,
      "learning_rate": 2.408877683041888e-06,
      "loss": 0.0319,
      "step": 2428
    },
    {
      "epoch": 1.822205551387847,
      "grad_norm": 0.1115378588438034,
      "learning_rate": 2.388839954673222e-06,
      "loss": 0.0118,
      "step": 2429
    },
    {
      "epoch": 1.8229557389347337,
      "grad_norm": 0.1740262359380722,
      "learning_rate": 2.3688838747556674e-06,
      "loss": 0.0192,
      "step": 2430
    },
    {
      "epoch": 1.8237059264816204,
      "grad_norm": 0.08586451411247253,
      "learning_rate": 2.3490094775119597e-06,
      "loss": 0.0098,
      "step": 2431
    },
    {
      "epoch": 1.8244561140285072,
      "grad_norm": 0.10886698216199875,
      "learning_rate": 2.3292167970247193e-06,
      "loss": 0.0137,
      "step": 2432
    },
    {
      "epoch": 1.8252063015753939,
      "grad_norm": 0.10428888350725174,
      "learning_rate": 2.30950586723645e-06,
      "loss": 0.0124,
      "step": 2433
    },
    {
      "epoch": 1.8259564891222806,
      "grad_norm": 0.10169228166341782,
      "learning_rate": 2.2898767219494634e-06,
      "loss": 0.0127,
      "step": 2434
    },
    {
      "epoch": 1.8267066766691673,
      "grad_norm": 0.25430697202682495,
      "learning_rate": 2.270329394825793e-06,
      "loss": 0.0274,
      "step": 2435
    },
    {
      "epoch": 1.827456864216054,
      "grad_norm": 0.110308937728405,
      "learning_rate": 2.2508639193871805e-06,
      "loss": 0.0184,
      "step": 2436
    },
    {
      "epoch": 1.8282070517629407,
      "grad_norm": 0.13315962255001068,
      "learning_rate": 2.2314803290150287e-06,
      "loss": 0.0144,
      "step": 2437
    },
    {
      "epoch": 1.8289572393098275,
      "grad_norm": 0.08486826717853546,
      "learning_rate": 2.2121786569502535e-06,
      "loss": 0.0085,
      "step": 2438
    },
    {
      "epoch": 1.8297074268567142,
      "grad_norm": 0.084154412150383,
      "learning_rate": 2.192958936293338e-06,
      "loss": 0.0139,
      "step": 2439
    },
    {
      "epoch": 1.8304576144036009,
      "grad_norm": 0.12315694987773895,
      "learning_rate": 2.1738212000042e-06,
      "loss": 0.0176,
      "step": 2440
    },
    {
      "epoch": 1.8312078019504876,
      "grad_norm": 0.10296216607093811,
      "learning_rate": 2.1547654809021877e-06,
      "loss": 0.0109,
      "step": 2441
    },
    {
      "epoch": 1.8319579894973743,
      "grad_norm": 0.09593435376882553,
      "learning_rate": 2.135791811665977e-06,
      "loss": 0.0142,
      "step": 2442
    },
    {
      "epoch": 1.832708177044261,
      "grad_norm": 0.08409593999385834,
      "learning_rate": 2.1169002248335346e-06,
      "loss": 0.0101,
      "step": 2443
    },
    {
      "epoch": 1.8334583645911477,
      "grad_norm": 0.11003497987985611,
      "learning_rate": 2.098090752802073e-06,
      "loss": 0.0149,
      "step": 2444
    },
    {
      "epoch": 1.8342085521380345,
      "grad_norm": 0.10805971175432205,
      "learning_rate": 2.0793634278279907e-06,
      "loss": 0.0163,
      "step": 2445
    },
    {
      "epoch": 1.8349587396849212,
      "grad_norm": 0.13255015015602112,
      "learning_rate": 2.0607182820268133e-06,
      "loss": 0.0162,
      "step": 2446
    },
    {
      "epoch": 1.835708927231808,
      "grad_norm": 0.13535521924495697,
      "learning_rate": 2.042155347373109e-06,
      "loss": 0.0206,
      "step": 2447
    },
    {
      "epoch": 1.8364591147786946,
      "grad_norm": 0.13720273971557617,
      "learning_rate": 2.023674655700497e-06,
      "loss": 0.0126,
      "step": 2448
    },
    {
      "epoch": 1.8372093023255816,
      "grad_norm": 0.20435172319412231,
      "learning_rate": 2.0052762387015424e-06,
      "loss": 0.0223,
      "step": 2449
    },
    {
      "epoch": 1.837959489872468,
      "grad_norm": 0.1612643599510193,
      "learning_rate": 1.986960127927717e-06,
      "loss": 0.0147,
      "step": 2450
    },
    {
      "epoch": 1.838709677419355,
      "grad_norm": 0.139284148812294,
      "learning_rate": 1.9687263547893407e-06,
      "loss": 0.0182,
      "step": 2451
    },
    {
      "epoch": 1.8394598649662415,
      "grad_norm": 0.16131514310836792,
      "learning_rate": 1.9505749505555503e-06,
      "loss": 0.0175,
      "step": 2452
    },
    {
      "epoch": 1.8402100525131284,
      "grad_norm": 0.11480918526649475,
      "learning_rate": 1.932505946354213e-06,
      "loss": 0.0136,
      "step": 2453
    },
    {
      "epoch": 1.840960240060015,
      "grad_norm": 0.1593775898218155,
      "learning_rate": 1.9145193731718858e-06,
      "loss": 0.019,
      "step": 2454
    },
    {
      "epoch": 1.8417104276069018,
      "grad_norm": 0.28252995014190674,
      "learning_rate": 1.8966152618537846e-06,
      "loss": 0.0292,
      "step": 2455
    },
    {
      "epoch": 1.8424606151537883,
      "grad_norm": 0.20613250136375427,
      "learning_rate": 1.8787936431036824e-06,
      "loss": 0.0222,
      "step": 2456
    },
    {
      "epoch": 1.8432108027006753,
      "grad_norm": 0.1379401534795761,
      "learning_rate": 1.8610545474839036e-06,
      "loss": 0.0298,
      "step": 2457
    },
    {
      "epoch": 1.8439609902475618,
      "grad_norm": 0.09081874042749405,
      "learning_rate": 1.8433980054152533e-06,
      "loss": 0.0155,
      "step": 2458
    },
    {
      "epoch": 1.8447111777944487,
      "grad_norm": 0.05574246495962143,
      "learning_rate": 1.8258240471769662e-06,
      "loss": 0.0088,
      "step": 2459
    },
    {
      "epoch": 1.8454613653413352,
      "grad_norm": 0.11207207292318344,
      "learning_rate": 1.8083327029066399e-06,
      "loss": 0.0132,
      "step": 2460
    },
    {
      "epoch": 1.8462115528882221,
      "grad_norm": 0.14374390244483948,
      "learning_rate": 1.7909240026002138e-06,
      "loss": 0.0226,
      "step": 2461
    },
    {
      "epoch": 1.8469617404351086,
      "grad_norm": 0.08422519266605377,
      "learning_rate": 1.773597976111896e-06,
      "loss": 0.0087,
      "step": 2462
    },
    {
      "epoch": 1.8477119279819956,
      "grad_norm": 0.1311851292848587,
      "learning_rate": 1.7563546531541132e-06,
      "loss": 0.0167,
      "step": 2463
    },
    {
      "epoch": 1.848462115528882,
      "grad_norm": 0.11781198531389236,
      "learning_rate": 1.7391940632974667e-06,
      "loss": 0.0134,
      "step": 2464
    },
    {
      "epoch": 1.849212303075769,
      "grad_norm": 0.13201270997524261,
      "learning_rate": 1.7221162359706776e-06,
      "loss": 0.0158,
      "step": 2465
    },
    {
      "epoch": 1.8499624906226555,
      "grad_norm": 0.24497579038143158,
      "learning_rate": 1.705121200460541e-06,
      "loss": 0.0396,
      "step": 2466
    },
    {
      "epoch": 1.8507126781695424,
      "grad_norm": 0.16112586855888367,
      "learning_rate": 1.6882089859118766e-06,
      "loss": 0.0258,
      "step": 2467
    },
    {
      "epoch": 1.8514628657164292,
      "grad_norm": 0.12652385234832764,
      "learning_rate": 1.6713796213274457e-06,
      "loss": 0.0127,
      "step": 2468
    },
    {
      "epoch": 1.8522130532633159,
      "grad_norm": 0.12414011359214783,
      "learning_rate": 1.6546331355679623e-06,
      "loss": 0.0148,
      "step": 2469
    },
    {
      "epoch": 1.8529632408102026,
      "grad_norm": 0.1247783824801445,
      "learning_rate": 1.6379695573520093e-06,
      "loss": 0.016,
      "step": 2470
    },
    {
      "epoch": 1.8537134283570893,
      "grad_norm": 0.1910862773656845,
      "learning_rate": 1.621388915255967e-06,
      "loss": 0.0302,
      "step": 2471
    },
    {
      "epoch": 1.854463615903976,
      "grad_norm": 0.09329649060964584,
      "learning_rate": 1.604891237714018e-06,
      "loss": 0.0132,
      "step": 2472
    },
    {
      "epoch": 1.8552138034508627,
      "grad_norm": 0.12266834080219269,
      "learning_rate": 1.5884765530180478e-06,
      "loss": 0.0175,
      "step": 2473
    },
    {
      "epoch": 1.8559639909977494,
      "grad_norm": 0.10107134282588959,
      "learning_rate": 1.5721448893176228e-06,
      "loss": 0.0113,
      "step": 2474
    },
    {
      "epoch": 1.8567141785446362,
      "grad_norm": 0.1326693296432495,
      "learning_rate": 1.5558962746199335e-06,
      "loss": 0.0198,
      "step": 2475
    },
    {
      "epoch": 1.8574643660915229,
      "grad_norm": 0.0919971838593483,
      "learning_rate": 1.5397307367897684e-06,
      "loss": 0.013,
      "step": 2476
    },
    {
      "epoch": 1.8582145536384096,
      "grad_norm": 0.09108588844537735,
      "learning_rate": 1.5236483035494297e-06,
      "loss": 0.014,
      "step": 2477
    },
    {
      "epoch": 1.8589647411852963,
      "grad_norm": 0.1272948980331421,
      "learning_rate": 1.5076490024786893e-06,
      "loss": 0.0175,
      "step": 2478
    },
    {
      "epoch": 1.859714928732183,
      "grad_norm": 0.1336841732263565,
      "learning_rate": 1.4917328610147885e-06,
      "loss": 0.0179,
      "step": 2479
    },
    {
      "epoch": 1.8604651162790697,
      "grad_norm": 0.19899259507656097,
      "learning_rate": 1.4758999064523493e-06,
      "loss": 0.0218,
      "step": 2480
    },
    {
      "epoch": 1.8612153038259565,
      "grad_norm": 0.12127574533224106,
      "learning_rate": 1.4601501659433137e-06,
      "loss": 0.0147,
      "step": 2481
    },
    {
      "epoch": 1.8619654913728432,
      "grad_norm": 0.11681947857141495,
      "learning_rate": 1.444483666496943e-06,
      "loss": 0.0144,
      "step": 2482
    },
    {
      "epoch": 1.86271567891973,
      "grad_norm": 0.0795118510723114,
      "learning_rate": 1.4289004349797409e-06,
      "loss": 0.0128,
      "step": 2483
    },
    {
      "epoch": 1.8634658664666166,
      "grad_norm": 0.11302927881479263,
      "learning_rate": 1.4134004981154137e-06,
      "loss": 0.019,
      "step": 2484
    },
    {
      "epoch": 1.8642160540135033,
      "grad_norm": 0.12013009190559387,
      "learning_rate": 1.3979838824848378e-06,
      "loss": 0.0174,
      "step": 2485
    },
    {
      "epoch": 1.8649662415603903,
      "grad_norm": 0.15739822387695312,
      "learning_rate": 1.382650614525971e-06,
      "loss": 0.0179,
      "step": 2486
    },
    {
      "epoch": 1.8657164291072768,
      "grad_norm": 0.15660080313682556,
      "learning_rate": 1.3674007205338678e-06,
      "loss": 0.0199,
      "step": 2487
    },
    {
      "epoch": 1.8664666166541637,
      "grad_norm": 0.10661745071411133,
      "learning_rate": 1.3522342266605925e-06,
      "loss": 0.0144,
      "step": 2488
    },
    {
      "epoch": 1.8672168042010502,
      "grad_norm": 0.1589604765176773,
      "learning_rate": 1.3371511589152008e-06,
      "loss": 0.0189,
      "step": 2489
    },
    {
      "epoch": 1.8679669917479371,
      "grad_norm": 0.11324826627969742,
      "learning_rate": 1.3221515431636522e-06,
      "loss": 0.0135,
      "step": 2490
    },
    {
      "epoch": 1.8687171792948236,
      "grad_norm": 0.19825062155723572,
      "learning_rate": 1.307235405128815e-06,
      "loss": 0.0279,
      "step": 2491
    },
    {
      "epoch": 1.8694673668417106,
      "grad_norm": 0.18215356767177582,
      "learning_rate": 1.292402770390394e-06,
      "loss": 0.0225,
      "step": 2492
    },
    {
      "epoch": 1.870217554388597,
      "grad_norm": 0.09638606756925583,
      "learning_rate": 1.2776536643849145e-06,
      "loss": 0.0132,
      "step": 2493
    },
    {
      "epoch": 1.870967741935484,
      "grad_norm": 0.0821240171790123,
      "learning_rate": 1.2629881124056274e-06,
      "loss": 0.0109,
      "step": 2494
    },
    {
      "epoch": 1.8717179294823705,
      "grad_norm": 0.14057280123233795,
      "learning_rate": 1.2484061396025038e-06,
      "loss": 0.0144,
      "step": 2495
    },
    {
      "epoch": 1.8724681170292574,
      "grad_norm": 0.11429706960916519,
      "learning_rate": 1.2339077709822067e-06,
      "loss": 0.0161,
      "step": 2496
    },
    {
      "epoch": 1.873218304576144,
      "grad_norm": 0.11428690701723099,
      "learning_rate": 1.2194930314080032e-06,
      "loss": 0.0176,
      "step": 2497
    },
    {
      "epoch": 1.8739684921230308,
      "grad_norm": 0.1507968008518219,
      "learning_rate": 1.2051619455997476e-06,
      "loss": 0.0199,
      "step": 2498
    },
    {
      "epoch": 1.8747186796699173,
      "grad_norm": 0.20986028015613556,
      "learning_rate": 1.1909145381338472e-06,
      "loss": 0.0168,
      "step": 2499
    },
    {
      "epoch": 1.8754688672168043,
      "grad_norm": 0.23694069683551788,
      "learning_rate": 1.1767508334431964e-06,
      "loss": 0.0267,
      "step": 2500
    },
    {
      "epoch": 1.8762190547636908,
      "grad_norm": 0.09894689917564392,
      "learning_rate": 1.1626708558171606e-06,
      "loss": 0.0147,
      "step": 2501
    },
    {
      "epoch": 1.8769692423105777,
      "grad_norm": 0.1138412207365036,
      "learning_rate": 1.1486746294015193e-06,
      "loss": 0.0181,
      "step": 2502
    },
    {
      "epoch": 1.8777194298574642,
      "grad_norm": 0.12932085990905762,
      "learning_rate": 1.134762178198412e-06,
      "loss": 0.0169,
      "step": 2503
    },
    {
      "epoch": 1.8784696174043511,
      "grad_norm": 0.09746240079402924,
      "learning_rate": 1.1209335260663256e-06,
      "loss": 0.0121,
      "step": 2504
    },
    {
      "epoch": 1.8792198049512379,
      "grad_norm": 0.11326896399259567,
      "learning_rate": 1.1071886967200352e-06,
      "loss": 0.0208,
      "step": 2505
    },
    {
      "epoch": 1.8799699924981246,
      "grad_norm": 0.09456074237823486,
      "learning_rate": 1.0935277137305744e-06,
      "loss": 0.0124,
      "step": 2506
    },
    {
      "epoch": 1.8807201800450113,
      "grad_norm": 0.1309920698404312,
      "learning_rate": 1.0799506005251814e-06,
      "loss": 0.0181,
      "step": 2507
    },
    {
      "epoch": 1.881470367591898,
      "grad_norm": 0.12602433562278748,
      "learning_rate": 1.06645738038727e-06,
      "loss": 0.0184,
      "step": 2508
    },
    {
      "epoch": 1.8822205551387847,
      "grad_norm": 0.10263347625732422,
      "learning_rate": 1.053048076456381e-06,
      "loss": 0.0161,
      "step": 2509
    },
    {
      "epoch": 1.8829707426856714,
      "grad_norm": 0.09971325099468231,
      "learning_rate": 1.0397227117281528e-06,
      "loss": 0.0115,
      "step": 2510
    },
    {
      "epoch": 1.8837209302325582,
      "grad_norm": 0.11184331774711609,
      "learning_rate": 1.0264813090542725e-06,
      "loss": 0.0146,
      "step": 2511
    },
    {
      "epoch": 1.8844711177794449,
      "grad_norm": 0.16481465101242065,
      "learning_rate": 1.0133238911424426e-06,
      "loss": 0.0182,
      "step": 2512
    },
    {
      "epoch": 1.8852213053263316,
      "grad_norm": 0.16809651255607605,
      "learning_rate": 1.0002504805563362e-06,
      "loss": 0.0144,
      "step": 2513
    },
    {
      "epoch": 1.8859714928732183,
      "grad_norm": 0.0960497260093689,
      "learning_rate": 9.872610997155695e-07,
      "loss": 0.0135,
      "step": 2514
    },
    {
      "epoch": 1.886721680420105,
      "grad_norm": 0.07542908936738968,
      "learning_rate": 9.743557708956575e-07,
      "loss": 0.0089,
      "step": 2515
    },
    {
      "epoch": 1.8874718679669917,
      "grad_norm": 0.07793007045984268,
      "learning_rate": 9.615345162279521e-07,
      "loss": 0.0078,
      "step": 2516
    },
    {
      "epoch": 1.8882220555138785,
      "grad_norm": 0.21286483108997345,
      "learning_rate": 9.48797357699649e-07,
      "loss": 0.0262,
      "step": 2517
    },
    {
      "epoch": 1.8889722430607652,
      "grad_norm": 0.09491562098264694,
      "learning_rate": 9.361443171537254e-07,
      "loss": 0.0074,
      "step": 2518
    },
    {
      "epoch": 1.8897224306076519,
      "grad_norm": 0.14178186655044556,
      "learning_rate": 9.235754162889021e-07,
      "loss": 0.0205,
      "step": 2519
    },
    {
      "epoch": 1.8904726181545386,
      "grad_norm": 0.16438992321491241,
      "learning_rate": 9.110906766595872e-07,
      "loss": 0.0185,
      "step": 2520
    },
    {
      "epoch": 1.8912228057014253,
      "grad_norm": 0.17157548666000366,
      "learning_rate": 8.986901196759046e-07,
      "loss": 0.0291,
      "step": 2521
    },
    {
      "epoch": 1.891972993248312,
      "grad_norm": 0.06102016195654869,
      "learning_rate": 8.863737666035765e-07,
      "loss": 0.009,
      "step": 2522
    },
    {
      "epoch": 1.8927231807951987,
      "grad_norm": 0.18232974410057068,
      "learning_rate": 8.741416385639412e-07,
      "loss": 0.0306,
      "step": 2523
    },
    {
      "epoch": 1.8934733683420855,
      "grad_norm": 0.14115692675113678,
      "learning_rate": 8.619937565338854e-07,
      "loss": 0.0187,
      "step": 2524
    },
    {
      "epoch": 1.8942235558889724,
      "grad_norm": 0.10181062668561935,
      "learning_rate": 8.499301413458338e-07,
      "loss": 0.0114,
      "step": 2525
    },
    {
      "epoch": 1.894973743435859,
      "grad_norm": 0.12386108934879303,
      "learning_rate": 8.37950813687699e-07,
      "loss": 0.0154,
      "step": 2526
    },
    {
      "epoch": 1.8957239309827458,
      "grad_norm": 0.12360105663537979,
      "learning_rate": 8.26055794102848e-07,
      "loss": 0.0184,
      "step": 2527
    },
    {
      "epoch": 1.8964741185296323,
      "grad_norm": 0.12389027327299118,
      "learning_rate": 8.142451029900744e-07,
      "loss": 0.0143,
      "step": 2528
    },
    {
      "epoch": 1.8972243060765193,
      "grad_norm": 0.12951231002807617,
      "learning_rate": 8.025187606035434e-07,
      "loss": 0.0186,
      "step": 2529
    },
    {
      "epoch": 1.8979744936234058,
      "grad_norm": 0.10430709272623062,
      "learning_rate": 7.908767870527745e-07,
      "loss": 0.0135,
      "step": 2530
    },
    {
      "epoch": 1.8987246811702927,
      "grad_norm": 0.09554420411586761,
      "learning_rate": 7.793192023026142e-07,
      "loss": 0.0139,
      "step": 2531
    },
    {
      "epoch": 1.8994748687171792,
      "grad_norm": 0.13989853858947754,
      "learning_rate": 7.678460261731801e-07,
      "loss": 0.0205,
      "step": 2532
    },
    {
      "epoch": 1.9002250562640661,
      "grad_norm": 0.13728971779346466,
      "learning_rate": 7.564572783398339e-07,
      "loss": 0.0131,
      "step": 2533
    },
    {
      "epoch": 1.9009752438109526,
      "grad_norm": 0.09057745337486267,
      "learning_rate": 7.451529783331523e-07,
      "loss": 0.0093,
      "step": 2534
    },
    {
      "epoch": 1.9017254313578396,
      "grad_norm": 0.1505969762802124,
      "learning_rate": 7.339331455389175e-07,
      "loss": 0.0145,
      "step": 2535
    },
    {
      "epoch": 1.902475618904726,
      "grad_norm": 0.1213538721203804,
      "learning_rate": 7.227977991980217e-07,
      "loss": 0.0179,
      "step": 2536
    },
    {
      "epoch": 1.903225806451613,
      "grad_norm": 0.09671944379806519,
      "learning_rate": 7.117469584064895e-07,
      "loss": 0.0119,
      "step": 2537
    },
    {
      "epoch": 1.9039759939984995,
      "grad_norm": 0.08646194636821747,
      "learning_rate": 7.007806421154284e-07,
      "loss": 0.0074,
      "step": 2538
    },
    {
      "epoch": 1.9047261815453864,
      "grad_norm": 0.11140266805887222,
      "learning_rate": 6.898988691309893e-07,
      "loss": 0.0151,
      "step": 2539
    },
    {
      "epoch": 1.905476369092273,
      "grad_norm": 0.09805849939584732,
      "learning_rate": 6.791016581143395e-07,
      "loss": 0.0136,
      "step": 2540
    },
    {
      "epoch": 1.9062265566391599,
      "grad_norm": 0.1474841833114624,
      "learning_rate": 6.683890275816341e-07,
      "loss": 0.0185,
      "step": 2541
    },
    {
      "epoch": 1.9069767441860463,
      "grad_norm": 0.11673633009195328,
      "learning_rate": 6.577609959039776e-07,
      "loss": 0.0155,
      "step": 2542
    },
    {
      "epoch": 1.9077269317329333,
      "grad_norm": 0.13770844042301178,
      "learning_rate": 6.472175813074022e-07,
      "loss": 0.018,
      "step": 2543
    },
    {
      "epoch": 1.90847711927982,
      "grad_norm": 0.10890696942806244,
      "learning_rate": 6.367588018728166e-07,
      "loss": 0.0166,
      "step": 2544
    },
    {
      "epoch": 1.9092273068267067,
      "grad_norm": 0.08831574767827988,
      "learning_rate": 6.263846755360126e-07,
      "loss": 0.0115,
      "step": 2545
    },
    {
      "epoch": 1.9099774943735934,
      "grad_norm": 0.08889841288328171,
      "learning_rate": 6.16095220087587e-07,
      "loss": 0.011,
      "step": 2546
    },
    {
      "epoch": 1.9107276819204801,
      "grad_norm": 0.11765546351671219,
      "learning_rate": 6.05890453172936e-07,
      "loss": 0.0153,
      "step": 2547
    },
    {
      "epoch": 1.9114778694673669,
      "grad_norm": 0.13256019353866577,
      "learning_rate": 5.957703922922386e-07,
      "loss": 0.0199,
      "step": 2548
    },
    {
      "epoch": 1.9122280570142536,
      "grad_norm": 0.0688183531165123,
      "learning_rate": 5.857350548004015e-07,
      "loss": 0.0084,
      "step": 2549
    },
    {
      "epoch": 1.9129782445611403,
      "grad_norm": 0.09796126186847687,
      "learning_rate": 5.757844579070359e-07,
      "loss": 0.0095,
      "step": 2550
    },
    {
      "epoch": 1.913728432108027,
      "grad_norm": 0.09508487582206726,
      "learning_rate": 5.65918618676442e-07,
      "loss": 0.0117,
      "step": 2551
    },
    {
      "epoch": 1.9144786196549137,
      "grad_norm": 0.14749017357826233,
      "learning_rate": 5.561375540275581e-07,
      "loss": 0.0202,
      "step": 2552
    },
    {
      "epoch": 1.9152288072018004,
      "grad_norm": 0.12853582203388214,
      "learning_rate": 5.464412807339558e-07,
      "loss": 0.0171,
      "step": 2553
    },
    {
      "epoch": 1.9159789947486872,
      "grad_norm": 0.13968953490257263,
      "learning_rate": 5.368298154237727e-07,
      "loss": 0.0178,
      "step": 2554
    },
    {
      "epoch": 1.9167291822955739,
      "grad_norm": 0.14459316432476044,
      "learning_rate": 5.273031745797352e-07,
      "loss": 0.0146,
      "step": 2555
    },
    {
      "epoch": 1.9174793698424606,
      "grad_norm": 0.12243661284446716,
      "learning_rate": 5.17861374539097e-07,
      "loss": 0.0106,
      "step": 2556
    },
    {
      "epoch": 1.9182295573893473,
      "grad_norm": 0.14610101282596588,
      "learning_rate": 5.085044314936116e-07,
      "loss": 0.0136,
      "step": 2557
    },
    {
      "epoch": 1.918979744936234,
      "grad_norm": 0.10917095094919205,
      "learning_rate": 4.992323614895156e-07,
      "loss": 0.0184,
      "step": 2558
    },
    {
      "epoch": 1.9197299324831207,
      "grad_norm": 0.16345548629760742,
      "learning_rate": 4.900451804274898e-07,
      "loss": 0.0193,
      "step": 2559
    },
    {
      "epoch": 1.9204801200300075,
      "grad_norm": 0.16431359946727753,
      "learning_rate": 4.809429040626535e-07,
      "loss": 0.0191,
      "step": 2560
    },
    {
      "epoch": 1.9212303075768942,
      "grad_norm": 0.06987474858760834,
      "learning_rate": 4.719255480045148e-07,
      "loss": 0.0109,
      "step": 2561
    },
    {
      "epoch": 1.921980495123781,
      "grad_norm": 0.07272496074438095,
      "learning_rate": 4.6299312771694304e-07,
      "loss": 0.011,
      "step": 2562
    },
    {
      "epoch": 1.9227306826706676,
      "grad_norm": 0.1672581136226654,
      "learning_rate": 4.5414565851816806e-07,
      "loss": 0.018,
      "step": 2563
    },
    {
      "epoch": 1.9234808702175545,
      "grad_norm": 0.18490219116210938,
      "learning_rate": 4.453831555807253e-07,
      "loss": 0.0189,
      "step": 2564
    },
    {
      "epoch": 1.924231057764441,
      "grad_norm": 0.08568117022514343,
      "learning_rate": 4.36705633931439e-07,
      "loss": 0.0088,
      "step": 2565
    },
    {
      "epoch": 1.924981245311328,
      "grad_norm": 0.12976892292499542,
      "learning_rate": 4.281131084514167e-07,
      "loss": 0.0182,
      "step": 2566
    },
    {
      "epoch": 1.9257314328582145,
      "grad_norm": 0.21210747957229614,
      "learning_rate": 4.196055938759824e-07,
      "loss": 0.0219,
      "step": 2567
    },
    {
      "epoch": 1.9264816204051014,
      "grad_norm": 0.09944725036621094,
      "learning_rate": 4.111831047946879e-07,
      "loss": 0.0101,
      "step": 2568
    },
    {
      "epoch": 1.927231807951988,
      "grad_norm": 0.1581333726644516,
      "learning_rate": 4.0284565565127384e-07,
      "loss": 0.0131,
      "step": 2569
    },
    {
      "epoch": 1.9279819954988748,
      "grad_norm": 0.10216685384511948,
      "learning_rate": 3.9459326074364756e-07,
      "loss": 0.015,
      "step": 2570
    },
    {
      "epoch": 1.9287321830457613,
      "grad_norm": 0.0924372598528862,
      "learning_rate": 3.8642593422384965e-07,
      "loss": 0.012,
      "step": 2571
    },
    {
      "epoch": 1.9294823705926483,
      "grad_norm": 0.10501393675804138,
      "learning_rate": 3.7834369009804303e-07,
      "loss": 0.0155,
      "step": 2572
    },
    {
      "epoch": 1.9302325581395348,
      "grad_norm": 0.1474066525697708,
      "learning_rate": 3.703465422264796e-07,
      "loss": 0.0142,
      "step": 2573
    },
    {
      "epoch": 1.9309827456864217,
      "grad_norm": 0.07856673747301102,
      "learning_rate": 3.624345043234778e-07,
      "loss": 0.0118,
      "step": 2574
    },
    {
      "epoch": 1.9317329332333082,
      "grad_norm": 0.10008254647254944,
      "learning_rate": 3.5460758995741194e-07,
      "loss": 0.0105,
      "step": 2575
    },
    {
      "epoch": 1.9324831207801951,
      "grad_norm": 0.13343030214309692,
      "learning_rate": 3.468658125506563e-07,
      "loss": 0.0181,
      "step": 2576
    },
    {
      "epoch": 1.9332333083270816,
      "grad_norm": 0.11500576883554459,
      "learning_rate": 3.3920918537960754e-07,
      "loss": 0.0144,
      "step": 2577
    },
    {
      "epoch": 1.9339834958739686,
      "grad_norm": 0.14943444728851318,
      "learning_rate": 3.3163772157462357e-07,
      "loss": 0.0242,
      "step": 2578
    },
    {
      "epoch": 1.934733683420855,
      "grad_norm": 0.15446636080741882,
      "learning_rate": 3.241514341200236e-07,
      "loss": 0.021,
      "step": 2579
    },
    {
      "epoch": 1.935483870967742,
      "grad_norm": 0.06783124804496765,
      "learning_rate": 3.1675033585404355e-07,
      "loss": 0.0073,
      "step": 2580
    },
    {
      "epoch": 1.9362340585146287,
      "grad_norm": 0.14099209010601044,
      "learning_rate": 3.0943443946884755e-07,
      "loss": 0.0192,
      "step": 2581
    },
    {
      "epoch": 1.9369842460615154,
      "grad_norm": 0.11827205866575241,
      "learning_rate": 3.0220375751047194e-07,
      "loss": 0.0126,
      "step": 2582
    },
    {
      "epoch": 1.9377344336084021,
      "grad_norm": 0.16399821639060974,
      "learning_rate": 2.950583023788256e-07,
      "loss": 0.0216,
      "step": 2583
    },
    {
      "epoch": 1.9384846211552889,
      "grad_norm": 0.10808265209197998,
      "learning_rate": 2.879980863276621e-07,
      "loss": 0.012,
      "step": 2584
    },
    {
      "epoch": 1.9392348087021756,
      "grad_norm": 0.11654945462942123,
      "learning_rate": 2.8102312146455755e-07,
      "loss": 0.017,
      "step": 2585
    },
    {
      "epoch": 1.9399849962490623,
      "grad_norm": 0.12237963080406189,
      "learning_rate": 2.7413341975088824e-07,
      "loss": 0.0263,
      "step": 2586
    },
    {
      "epoch": 1.940735183795949,
      "grad_norm": 0.11718087643384933,
      "learning_rate": 2.6732899300180857e-07,
      "loss": 0.0124,
      "step": 2587
    },
    {
      "epoch": 1.9414853713428357,
      "grad_norm": 0.08054438978433609,
      "learning_rate": 2.606098528862566e-07,
      "loss": 0.012,
      "step": 2588
    },
    {
      "epoch": 1.9422355588897224,
      "grad_norm": 0.10906367003917694,
      "learning_rate": 2.5397601092687627e-07,
      "loss": 0.0188,
      "step": 2589
    },
    {
      "epoch": 1.9429857464366092,
      "grad_norm": 0.10259516537189484,
      "learning_rate": 2.474274785000619e-07,
      "loss": 0.0145,
      "step": 2590
    },
    {
      "epoch": 1.9437359339834959,
      "grad_norm": 0.11857086420059204,
      "learning_rate": 2.40964266835908e-07,
      "loss": 0.0162,
      "step": 2591
    },
    {
      "epoch": 1.9444861215303826,
      "grad_norm": 0.07964332401752472,
      "learning_rate": 2.3458638701817636e-07,
      "loss": 0.0126,
      "step": 2592
    },
    {
      "epoch": 1.9452363090772693,
      "grad_norm": 0.08068211376667023,
      "learning_rate": 2.2829384998430681e-07,
      "loss": 0.0105,
      "step": 2593
    },
    {
      "epoch": 1.945986496624156,
      "grad_norm": 0.11038036644458771,
      "learning_rate": 2.2208666652537846e-07,
      "loss": 0.0163,
      "step": 2594
    },
    {
      "epoch": 1.9467366841710427,
      "grad_norm": 0.12264223396778107,
      "learning_rate": 2.1596484728610421e-07,
      "loss": 0.0222,
      "step": 2595
    },
    {
      "epoch": 1.9474868717179294,
      "grad_norm": 0.1597457379102707,
      "learning_rate": 2.099284027647974e-07,
      "loss": 0.0232,
      "step": 2596
    },
    {
      "epoch": 1.9482370592648162,
      "grad_norm": 0.13513779640197754,
      "learning_rate": 2.039773433133718e-07,
      "loss": 0.0156,
      "step": 2597
    },
    {
      "epoch": 1.9489872468117029,
      "grad_norm": 0.1858983039855957,
      "learning_rate": 1.9811167913729723e-07,
      "loss": 0.0143,
      "step": 2598
    },
    {
      "epoch": 1.9497374343585896,
      "grad_norm": 0.0995759516954422,
      "learning_rate": 1.923314202956217e-07,
      "loss": 0.0099,
      "step": 2599
    },
    {
      "epoch": 1.9504876219054763,
      "grad_norm": 0.11884553730487823,
      "learning_rate": 1.8663657670091595e-07,
      "loss": 0.019,
      "step": 2600
    },
    {
      "epoch": 1.9504876219054763,
      "eval_loss": 0.03117949888110161,
      "eval_runtime": 5.1258,
      "eval_samples_per_second": 10.535,
      "eval_steps_per_second": 2.731,
      "step": 2600
    },
    {
      "epoch": 1.9512378094523632,
      "grad_norm": 0.09198283404111862,
      "learning_rate": 1.810271581192735e-07,
      "loss": 0.0081,
      "step": 2601
    },
    {
      "epoch": 1.9519879969992497,
      "grad_norm": 0.13007594645023346,
      "learning_rate": 1.755031741702995e-07,
      "loss": 0.0131,
      "step": 2602
    },
    {
      "epoch": 1.9527381845461367,
      "grad_norm": 0.17154718935489655,
      "learning_rate": 1.7006463432707177e-07,
      "loss": 0.0161,
      "step": 2603
    },
    {
      "epoch": 1.9534883720930232,
      "grad_norm": 0.08539299666881561,
      "learning_rate": 1.6471154791616317e-07,
      "loss": 0.013,
      "step": 2604
    },
    {
      "epoch": 1.9542385596399101,
      "grad_norm": 0.11189521849155426,
      "learning_rate": 1.59443924117586e-07,
      "loss": 0.0119,
      "step": 2605
    },
    {
      "epoch": 1.9549887471867966,
      "grad_norm": 0.06556031852960587,
      "learning_rate": 1.5426177196479207e-07,
      "loss": 0.0079,
      "step": 2606
    },
    {
      "epoch": 1.9557389347336835,
      "grad_norm": 0.135658398270607,
      "learning_rate": 1.4916510034466702e-07,
      "loss": 0.0231,
      "step": 2607
    },
    {
      "epoch": 1.95648912228057,
      "grad_norm": 0.07447344064712524,
      "learning_rate": 1.441539179974971e-07,
      "loss": 0.0113,
      "step": 2608
    },
    {
      "epoch": 1.957239309827457,
      "grad_norm": 0.09345728904008865,
      "learning_rate": 1.3922823351697479e-07,
      "loss": 0.0114,
      "step": 2609
    },
    {
      "epoch": 1.9579894973743435,
      "grad_norm": 0.11973581463098526,
      "learning_rate": 1.343880553501542e-07,
      "loss": 0.0176,
      "step": 2610
    },
    {
      "epoch": 1.9587396849212304,
      "grad_norm": 0.1319183111190796,
      "learning_rate": 1.2963339179746238e-07,
      "loss": 0.0125,
      "step": 2611
    },
    {
      "epoch": 1.959489872468117,
      "grad_norm": 0.11517900228500366,
      "learning_rate": 1.2496425101268804e-07,
      "loss": 0.0199,
      "step": 2612
    },
    {
      "epoch": 1.9602400600150038,
      "grad_norm": 0.053299639374017715,
      "learning_rate": 1.2038064100294843e-07,
      "loss": 0.0067,
      "step": 2613
    },
    {
      "epoch": 1.9609902475618903,
      "grad_norm": 0.11605609208345413,
      "learning_rate": 1.158825696286725e-07,
      "loss": 0.0127,
      "step": 2614
    },
    {
      "epoch": 1.9617404351087773,
      "grad_norm": 0.12783505022525787,
      "learning_rate": 1.114700446036232e-07,
      "loss": 0.0189,
      "step": 2615
    },
    {
      "epoch": 1.9624906226556638,
      "grad_norm": 0.0946228876709938,
      "learning_rate": 1.0714307349483089e-07,
      "loss": 0.0114,
      "step": 2616
    },
    {
      "epoch": 1.9632408102025507,
      "grad_norm": 0.13363100588321686,
      "learning_rate": 1.029016637226432e-07,
      "loss": 0.0173,
      "step": 2617
    },
    {
      "epoch": 1.9639909977494372,
      "grad_norm": 0.10163934528827667,
      "learning_rate": 9.874582256064192e-08,
      "loss": 0.0127,
      "step": 2618
    },
    {
      "epoch": 1.9647411852963241,
      "grad_norm": 0.14493410289287567,
      "learning_rate": 9.46755571356983e-08,
      "loss": 0.0171,
      "step": 2619
    },
    {
      "epoch": 1.9654913728432108,
      "grad_norm": 0.12048706412315369,
      "learning_rate": 9.069087442791224e-08,
      "loss": 0.0184,
      "step": 2620
    },
    {
      "epoch": 1.9662415603900976,
      "grad_norm": 0.11740259826183319,
      "learning_rate": 8.679178127062871e-08,
      "loss": 0.018,
      "step": 2621
    },
    {
      "epoch": 1.9669917479369843,
      "grad_norm": 0.1652223914861679,
      "learning_rate": 8.297828435039346e-08,
      "loss": 0.0202,
      "step": 2622
    },
    {
      "epoch": 1.967741935483871,
      "grad_norm": 0.09575902670621872,
      "learning_rate": 7.925039020699187e-08,
      "loss": 0.0145,
      "step": 2623
    },
    {
      "epoch": 1.9684921230307577,
      "grad_norm": 0.2031949758529663,
      "learning_rate": 7.56081052333879e-08,
      "loss": 0.0316,
      "step": 2624
    },
    {
      "epoch": 1.9692423105776444,
      "grad_norm": 0.15664900839328766,
      "learning_rate": 7.205143567574624e-08,
      "loss": 0.0267,
      "step": 2625
    },
    {
      "epoch": 1.9699924981245311,
      "grad_norm": 0.07686346024274826,
      "learning_rate": 6.858038763340458e-08,
      "loss": 0.0098,
      "step": 2626
    },
    {
      "epoch": 1.9707426856714179,
      "grad_norm": 0.155964657664299,
      "learning_rate": 6.519496705886252e-08,
      "loss": 0.0257,
      "step": 2627
    },
    {
      "epoch": 1.9714928732183046,
      "grad_norm": 0.13349410891532898,
      "learning_rate": 6.189517975778713e-08,
      "loss": 0.0169,
      "step": 2628
    },
    {
      "epoch": 1.9722430607651913,
      "grad_norm": 0.11973363161087036,
      "learning_rate": 5.8681031388990724e-08,
      "loss": 0.0109,
      "step": 2629
    },
    {
      "epoch": 1.972993248312078,
      "grad_norm": 0.08597532659769058,
      "learning_rate": 5.555252746441975e-08,
      "loss": 0.0108,
      "step": 2630
    },
    {
      "epoch": 1.9737434358589647,
      "grad_norm": 0.10258978605270386,
      "learning_rate": 5.25096733491548e-08,
      "loss": 0.0121,
      "step": 2631
    },
    {
      "epoch": 1.9744936234058514,
      "grad_norm": 0.11236192286014557,
      "learning_rate": 4.9552474261377326e-08,
      "loss": 0.0157,
      "step": 2632
    },
    {
      "epoch": 1.9752438109527382,
      "grad_norm": 0.08419892191886902,
      "learning_rate": 4.6680935272408465e-08,
      "loss": 0.0108,
      "step": 2633
    },
    {
      "epoch": 1.9759939984996249,
      "grad_norm": 0.13061557710170746,
      "learning_rate": 4.3895061306648e-08,
      "loss": 0.0178,
      "step": 2634
    },
    {
      "epoch": 1.9767441860465116,
      "grad_norm": 0.08569876849651337,
      "learning_rate": 4.119485714159099e-08,
      "loss": 0.0088,
      "step": 2635
    },
    {
      "epoch": 1.9774943735933983,
      "grad_norm": 0.10098525881767273,
      "learning_rate": 3.8580327407827796e-08,
      "loss": 0.0171,
      "step": 2636
    },
    {
      "epoch": 1.978244561140285,
      "grad_norm": 0.16379106044769287,
      "learning_rate": 3.605147658901631e-08,
      "loss": 0.0161,
      "step": 2637
    },
    {
      "epoch": 1.978994748687172,
      "grad_norm": 0.13738934695720673,
      "learning_rate": 3.360830902189305e-08,
      "loss": 0.0217,
      "step": 2638
    },
    {
      "epoch": 1.9797449362340584,
      "grad_norm": 0.12114493548870087,
      "learning_rate": 3.125082889623987e-08,
      "loss": 0.0123,
      "step": 2639
    },
    {
      "epoch": 1.9804951237809454,
      "grad_norm": 0.22203579545021057,
      "learning_rate": 2.8979040254911717e-08,
      "loss": 0.0316,
      "step": 2640
    },
    {
      "epoch": 1.9812453113278319,
      "grad_norm": 0.13473734259605408,
      "learning_rate": 2.67929469937922e-08,
      "loss": 0.0215,
      "step": 2641
    },
    {
      "epoch": 1.9819954988747188,
      "grad_norm": 0.15361478924751282,
      "learning_rate": 2.4692552861826925e-08,
      "loss": 0.0195,
      "step": 2642
    },
    {
      "epoch": 1.9827456864216053,
      "grad_norm": 0.16221070289611816,
      "learning_rate": 2.2677861460984607e-08,
      "loss": 0.0156,
      "step": 2643
    },
    {
      "epoch": 1.9834958739684923,
      "grad_norm": 0.1510624885559082,
      "learning_rate": 2.074887624625155e-08,
      "loss": 0.0226,
      "step": 2644
    },
    {
      "epoch": 1.9842460615153787,
      "grad_norm": 0.12968359887599945,
      "learning_rate": 1.890560052565937e-08,
      "loss": 0.0211,
      "step": 2645
    },
    {
      "epoch": 1.9849962490622657,
      "grad_norm": 0.09835748374462128,
      "learning_rate": 1.7148037460235078e-08,
      "loss": 0.0116,
      "step": 2646
    },
    {
      "epoch": 1.9857464366091522,
      "grad_norm": 0.2587953209877014,
      "learning_rate": 1.5476190064034334e-08,
      "loss": 0.0358,
      "step": 2647
    },
    {
      "epoch": 1.9864966241560391,
      "grad_norm": 0.14192144572734833,
      "learning_rate": 1.3890061204108185e-08,
      "loss": 0.0185,
      "step": 2648
    },
    {
      "epoch": 1.9872468117029256,
      "grad_norm": 0.16238971054553986,
      "learning_rate": 1.2389653600508588e-08,
      "loss": 0.0211,
      "step": 2649
    },
    {
      "epoch": 1.9879969992498125,
      "grad_norm": 0.14360111951828003,
      "learning_rate": 1.0974969826288428e-08,
      "loss": 0.0172,
      "step": 2650
    },
    {
      "epoch": 1.988747186796699,
      "grad_norm": 0.10696499794721603,
      "learning_rate": 9.646012307490405e-09,
      "loss": 0.0112,
      "step": 2651
    },
    {
      "epoch": 1.989497374343586,
      "grad_norm": 0.19679325819015503,
      "learning_rate": 8.402783323147034e-09,
      "loss": 0.0238,
      "step": 2652
    },
    {
      "epoch": 1.9902475618904725,
      "grad_norm": 0.10451437532901764,
      "learning_rate": 7.245285005275104e-09,
      "loss": 0.0134,
      "step": 2653
    },
    {
      "epoch": 1.9909977494373594,
      "grad_norm": 0.06075247749686241,
      "learning_rate": 6.1735193388701155e-09,
      "loss": 0.0077,
      "step": 2654
    },
    {
      "epoch": 1.991747936984246,
      "grad_norm": 0.07771611213684082,
      "learning_rate": 5.187488161895182e-09,
      "loss": 0.0071,
      "step": 2655
    },
    {
      "epoch": 1.9924981245311328,
      "grad_norm": 0.1421760767698288,
      "learning_rate": 4.28719316531434e-09,
      "loss": 0.0187,
      "step": 2656
    },
    {
      "epoch": 1.9932483120780196,
      "grad_norm": 0.11837299168109894,
      "learning_rate": 3.4726358930259328e-09,
      "loss": 0.0166,
      "step": 2657
    },
    {
      "epoch": 1.9939984996249063,
      "grad_norm": 0.1275044083595276,
      "learning_rate": 2.743817741929222e-09,
      "loss": 0.0234,
      "step": 2658
    },
    {
      "epoch": 1.994748687171793,
      "grad_norm": 0.11450985819101334,
      "learning_rate": 2.1007399618688807e-09,
      "loss": 0.0121,
      "step": 2659
    },
    {
      "epoch": 1.9954988747186797,
      "grad_norm": 0.13041210174560547,
      "learning_rate": 1.543403655662745e-09,
      "loss": 0.0182,
      "step": 2660
    },
    {
      "epoch": 1.9962490622655664,
      "grad_norm": 0.1500764936208725,
      "learning_rate": 1.0718097790907156e-09,
      "loss": 0.0188,
      "step": 2661
    },
    {
      "epoch": 1.9969992498124531,
      "grad_norm": 0.07319959998130798,
      "learning_rate": 6.859591408836519e-10,
      "loss": 0.0079,
      "step": 2662
    },
    {
      "epoch": 1.9977494373593399,
      "grad_norm": 0.1447007954120636,
      "learning_rate": 3.8585240273447677e-10,
      "loss": 0.0202,
      "step": 2663
    },
    {
      "epoch": 1.9984996249062266,
      "grad_norm": 0.09215506911277771,
      "learning_rate": 1.7149007930927773e-10,
      "loss": 0.0138,
      "step": 2664
    },
    {
      "epoch": 1.9992498124531133,
      "grad_norm": 0.08019605278968811,
      "learning_rate": 4.2872538208449386e-11,
      "loss": 0.0086,
      "step": 2665
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.1368834525346756,
      "learning_rate": 0.0,
      "loss": 0.02,
      "step": 2666
    },
    {
      "epoch": 2.0,
      "step": 2666,
      "total_flos": 1.0511156991219466e+18,
      "train_loss": 0.02966136350740091,
      "train_runtime": 3474.0812,
      "train_samples_per_second": 3.068,
      "train_steps_per_second": 0.767
    }
  ],
  "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": 1.0511156991219466e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}