{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9943851768669285,
  "eval_steps": 500,
  "global_step": 333,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008983717012914094,
      "grad_norm": 5.839048375887439,
      "learning_rate": 2.3529411764705885e-06,
      "loss": 0.851,
      "step": 1
    },
    {
      "epoch": 0.017967434025828188,
      "grad_norm": 5.796622055110998,
      "learning_rate": 4.705882352941177e-06,
      "loss": 0.8589,
      "step": 2
    },
    {
      "epoch": 0.02695115103874228,
      "grad_norm": 5.449611767460796,
      "learning_rate": 7.058823529411766e-06,
      "loss": 0.851,
      "step": 3
    },
    {
      "epoch": 0.035934868051656375,
      "grad_norm": 3.918769913446889,
      "learning_rate": 9.411764705882354e-06,
      "loss": 0.8013,
      "step": 4
    },
    {
      "epoch": 0.044918585064570464,
      "grad_norm": 2.1146491352495644,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.7675,
      "step": 5
    },
    {
      "epoch": 0.05390230207748456,
      "grad_norm": 5.3866836077244535,
      "learning_rate": 1.4117647058823532e-05,
      "loss": 0.8021,
      "step": 6
    },
    {
      "epoch": 0.06288601909039865,
      "grad_norm": 7.1506220356372205,
      "learning_rate": 1.647058823529412e-05,
      "loss": 0.7729,
      "step": 7
    },
    {
      "epoch": 0.07186973610331275,
      "grad_norm": 7.3684198638012575,
      "learning_rate": 1.8823529411764708e-05,
      "loss": 0.7895,
      "step": 8
    },
    {
      "epoch": 0.08085345311622684,
      "grad_norm": 4.2008371048182624,
      "learning_rate": 2.1176470588235296e-05,
      "loss": 0.7502,
      "step": 9
    },
    {
      "epoch": 0.08983717012914093,
      "grad_norm": 3.018632704832958,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.7085,
      "step": 10
    },
    {
      "epoch": 0.09882088714205503,
      "grad_norm": 2.4103057201747564,
      "learning_rate": 2.5882352941176475e-05,
      "loss": 0.68,
      "step": 11
    },
    {
      "epoch": 0.10780460415496912,
      "grad_norm": 1.6029821095889687,
      "learning_rate": 2.8235294117647063e-05,
      "loss": 0.6579,
      "step": 12
    },
    {
      "epoch": 0.11678832116788321,
      "grad_norm": 1.5359955676401171,
      "learning_rate": 3.0588235294117644e-05,
      "loss": 0.6348,
      "step": 13
    },
    {
      "epoch": 0.1257720381807973,
      "grad_norm": 1.2677879393997815,
      "learning_rate": 3.294117647058824e-05,
      "loss": 0.6259,
      "step": 14
    },
    {
      "epoch": 0.13475575519371139,
      "grad_norm": 1.1307239717924904,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.6217,
      "step": 15
    },
    {
      "epoch": 0.1437394722066255,
      "grad_norm": 1.1919595690235936,
      "learning_rate": 3.7647058823529415e-05,
      "loss": 0.6117,
      "step": 16
    },
    {
      "epoch": 0.1527231892195396,
      "grad_norm": 1.5137586529647031,
      "learning_rate": 4e-05,
      "loss": 0.5951,
      "step": 17
    },
    {
      "epoch": 0.16170690623245368,
      "grad_norm": 1.0839410029716163,
      "learning_rate": 4.235294117647059e-05,
      "loss": 0.5873,
      "step": 18
    },
    {
      "epoch": 0.17069062324536777,
      "grad_norm": 1.411013539783664,
      "learning_rate": 4.470588235294118e-05,
      "loss": 0.5855,
      "step": 19
    },
    {
      "epoch": 0.17967434025828186,
      "grad_norm": 0.9604494585660002,
      "learning_rate": 4.705882352941177e-05,
      "loss": 0.5855,
      "step": 20
    },
    {
      "epoch": 0.18865805727119594,
      "grad_norm": 1.7966858442462152,
      "learning_rate": 4.941176470588236e-05,
      "loss": 0.5834,
      "step": 21
    },
    {
      "epoch": 0.19764177428411006,
      "grad_norm": 1.2639231326008076,
      "learning_rate": 5.176470588235295e-05,
      "loss": 0.573,
      "step": 22
    },
    {
      "epoch": 0.20662549129702415,
      "grad_norm": 1.6189242064013842,
      "learning_rate": 5.411764705882354e-05,
      "loss": 0.5687,
      "step": 23
    },
    {
      "epoch": 0.21560920830993824,
      "grad_norm": 1.2918939880837212,
      "learning_rate": 5.6470588235294126e-05,
      "loss": 0.5629,
      "step": 24
    },
    {
      "epoch": 0.22459292532285233,
      "grad_norm": 1.8153484123304104,
      "learning_rate": 5.8823529411764714e-05,
      "loss": 0.5644,
      "step": 25
    },
    {
      "epoch": 0.23357664233576642,
      "grad_norm": 1.383080233694585,
      "learning_rate": 6.117647058823529e-05,
      "loss": 0.561,
      "step": 26
    },
    {
      "epoch": 0.2425603593486805,
      "grad_norm": 1.908843640323105,
      "learning_rate": 6.352941176470589e-05,
      "loss": 0.5581,
      "step": 27
    },
    {
      "epoch": 0.2515440763615946,
      "grad_norm": 1.5638595834141664,
      "learning_rate": 6.588235294117648e-05,
      "loss": 0.5497,
      "step": 28
    },
    {
      "epoch": 0.2605277933745087,
      "grad_norm": 1.0799402527915973,
      "learning_rate": 6.823529411764707e-05,
      "loss": 0.5405,
      "step": 29
    },
    {
      "epoch": 0.26951151038742277,
      "grad_norm": 1.2842268786606617,
      "learning_rate": 7.058823529411765e-05,
      "loss": 0.5489,
      "step": 30
    },
    {
      "epoch": 0.2784952274003369,
      "grad_norm": 2.1296312845067953,
      "learning_rate": 7.294117647058824e-05,
      "loss": 0.5429,
      "step": 31
    },
    {
      "epoch": 0.287478944413251,
      "grad_norm": 1.2081689918064082,
      "learning_rate": 7.529411764705883e-05,
      "loss": 0.5439,
      "step": 32
    },
    {
      "epoch": 0.29646266142616506,
      "grad_norm": 2.681914408101531,
      "learning_rate": 7.764705882352942e-05,
      "loss": 0.5409,
      "step": 33
    },
    {
      "epoch": 0.3054463784390792,
      "grad_norm": 2.088525203831682,
      "learning_rate": 8e-05,
      "loss": 0.5448,
      "step": 34
    },
    {
      "epoch": 0.31443009545199324,
      "grad_norm": 2.429087878227431,
      "learning_rate": 7.999779207981935e-05,
      "loss": 0.5376,
      "step": 35
    },
    {
      "epoch": 0.32341381246490736,
      "grad_norm": 2.2118789436187467,
      "learning_rate": 7.999116856302298e-05,
      "loss": 0.5429,
      "step": 36
    },
    {
      "epoch": 0.3323975294778215,
      "grad_norm": 1.998359697047038,
      "learning_rate": 7.998013018082072e-05,
      "loss": 0.5372,
      "step": 37
    },
    {
      "epoch": 0.34138124649073553,
      "grad_norm": 1.4739747352645118,
      "learning_rate": 7.996467815180588e-05,
      "loss": 0.5306,
      "step": 38
    },
    {
      "epoch": 0.35036496350364965,
      "grad_norm": 1.8627456810217522,
      "learning_rate": 7.994481418182082e-05,
      "loss": 0.5309,
      "step": 39
    },
    {
      "epoch": 0.3593486805165637,
      "grad_norm": 1.2832017910437872,
      "learning_rate": 7.992054046376854e-05,
      "loss": 0.5301,
      "step": 40
    },
    {
      "epoch": 0.36833239752947783,
      "grad_norm": 1.9368339421908154,
      "learning_rate": 7.989185967737066e-05,
      "loss": 0.534,
      "step": 41
    },
    {
      "epoch": 0.3773161145423919,
      "grad_norm": 1.4631687551659387,
      "learning_rate": 7.985877498887149e-05,
      "loss": 0.526,
      "step": 42
    },
    {
      "epoch": 0.386299831555306,
      "grad_norm": 1.4992485001174218,
      "learning_rate": 7.982129005068865e-05,
      "loss": 0.5169,
      "step": 43
    },
    {
      "epoch": 0.3952835485682201,
      "grad_norm": 1.2556436640866477,
      "learning_rate": 7.977940900100967e-05,
      "loss": 0.5086,
      "step": 44
    },
    {
      "epoch": 0.4042672655811342,
      "grad_norm": 1.4032135906881233,
      "learning_rate": 7.973313646333532e-05,
      "loss": 0.5214,
      "step": 45
    },
    {
      "epoch": 0.4132509825940483,
      "grad_norm": 1.3909591395416454,
      "learning_rate": 7.968247754596908e-05,
      "loss": 0.5143,
      "step": 46
    },
    {
      "epoch": 0.42223469960696236,
      "grad_norm": 1.9540476320319087,
      "learning_rate": 7.962743784145323e-05,
      "loss": 0.5207,
      "step": 47
    },
    {
      "epoch": 0.4312184166198765,
      "grad_norm": 1.0309999384978459,
      "learning_rate": 7.956802342595152e-05,
      "loss": 0.5069,
      "step": 48
    },
    {
      "epoch": 0.4402021336327906,
      "grad_norm": 1.3811159742393722,
      "learning_rate": 7.950424085857827e-05,
      "loss": 0.5145,
      "step": 49
    },
    {
      "epoch": 0.44918585064570465,
      "grad_norm": 1.4750899151595693,
      "learning_rate": 7.943609718067437e-05,
      "loss": 0.5034,
      "step": 50
    },
    {
      "epoch": 0.45816956765861877,
      "grad_norm": 1.586345005999154,
      "learning_rate": 7.936359991502993e-05,
      "loss": 0.5139,
      "step": 51
    },
    {
      "epoch": 0.46715328467153283,
      "grad_norm": 0.8294429144133623,
      "learning_rate": 7.92867570650537e-05,
      "loss": 0.5024,
      "step": 52
    },
    {
      "epoch": 0.47613700168444695,
      "grad_norm": 1.4768401149936898,
      "learning_rate": 7.920557711388967e-05,
      "loss": 0.5024,
      "step": 53
    },
    {
      "epoch": 0.485120718697361,
      "grad_norm": 1.4206688134227474,
      "learning_rate": 7.912006902348045e-05,
      "loss": 0.5087,
      "step": 54
    },
    {
      "epoch": 0.4941044357102751,
      "grad_norm": 0.9288407004391388,
      "learning_rate": 7.903024223357797e-05,
      "loss": 0.5059,
      "step": 55
    },
    {
      "epoch": 0.5030881527231892,
      "grad_norm": 1.0360922211438188,
      "learning_rate": 7.893610666070134e-05,
      "loss": 0.5025,
      "step": 56
    },
    {
      "epoch": 0.5120718697361033,
      "grad_norm": 1.5478822264915786,
      "learning_rate": 7.883767269704209e-05,
      "loss": 0.5166,
      "step": 57
    },
    {
      "epoch": 0.5210555867490174,
      "grad_norm": 1.2814412272627784,
      "learning_rate": 7.873495120931697e-05,
      "loss": 0.5053,
      "step": 58
    },
    {
      "epoch": 0.5300393037619315,
      "grad_norm": 1.2394259373221026,
      "learning_rate": 7.86279535375683e-05,
      "loss": 0.506,
      "step": 59
    },
    {
      "epoch": 0.5390230207748455,
      "grad_norm": 0.7402164979257034,
      "learning_rate": 7.851669149391198e-05,
      "loss": 0.4972,
      "step": 60
    },
    {
      "epoch": 0.5480067377877597,
      "grad_norm": 1.0415746091150728,
      "learning_rate": 7.84011773612336e-05,
      "loss": 0.503,
      "step": 61
    },
    {
      "epoch": 0.5569904548006738,
      "grad_norm": 1.287772964853304,
      "learning_rate": 7.828142389183239e-05,
      "loss": 0.5002,
      "step": 62
    },
    {
      "epoch": 0.5659741718135879,
      "grad_norm": 1.5812766703675953,
      "learning_rate": 7.815744430601344e-05,
      "loss": 0.5056,
      "step": 63
    },
    {
      "epoch": 0.574957888826502,
      "grad_norm": 0.8792970340323266,
      "learning_rate": 7.802925229062823e-05,
      "loss": 0.4918,
      "step": 64
    },
    {
      "epoch": 0.583941605839416,
      "grad_norm": 1.1398022409490918,
      "learning_rate": 7.789686199756365e-05,
      "loss": 0.4969,
      "step": 65
    },
    {
      "epoch": 0.5929253228523301,
      "grad_norm": 1.7704394806497858,
      "learning_rate": 7.776028804217968e-05,
      "loss": 0.5026,
      "step": 66
    },
    {
      "epoch": 0.6019090398652442,
      "grad_norm": 0.9434307642741575,
      "learning_rate": 7.761954550169593e-05,
      "loss": 0.4932,
      "step": 67
    },
    {
      "epoch": 0.6108927568781584,
      "grad_norm": 2.082446516632287,
      "learning_rate": 7.74746499135272e-05,
      "loss": 0.5172,
      "step": 68
    },
    {
      "epoch": 0.6198764738910725,
      "grad_norm": 1.8262658550358597,
      "learning_rate": 7.732561727356811e-05,
      "loss": 0.5039,
      "step": 69
    },
    {
      "epoch": 0.6288601909039865,
      "grad_norm": 1.2365083047972374,
      "learning_rate": 7.717246403442735e-05,
      "loss": 0.4973,
      "step": 70
    },
    {
      "epoch": 0.6378439079169006,
      "grad_norm": 1.4539572435843178,
      "learning_rate": 7.701520710361129e-05,
      "loss": 0.4988,
      "step": 71
    },
    {
      "epoch": 0.6468276249298147,
      "grad_norm": 1.0089172545802498,
      "learning_rate": 7.685386384165748e-05,
      "loss": 0.4874,
      "step": 72
    },
    {
      "epoch": 0.6558113419427288,
      "grad_norm": 1.3604668572334533,
      "learning_rate": 7.668845206021812e-05,
      "loss": 0.5035,
      "step": 73
    },
    {
      "epoch": 0.664795058955643,
      "grad_norm": 1.0804166408590006,
      "learning_rate": 7.651899002009375e-05,
      "loss": 0.4926,
      "step": 74
    },
    {
      "epoch": 0.673778775968557,
      "grad_norm": 1.4514655022285154,
      "learning_rate": 7.634549642921725e-05,
      "loss": 0.4939,
      "step": 75
    },
    {
      "epoch": 0.6827624929814711,
      "grad_norm": 0.991441799340063,
      "learning_rate": 7.616799044058867e-05,
      "loss": 0.5067,
      "step": 76
    },
    {
      "epoch": 0.6917462099943852,
      "grad_norm": 1.117806439503576,
      "learning_rate": 7.598649165016073e-05,
      "loss": 0.4915,
      "step": 77
    },
    {
      "epoch": 0.7007299270072993,
      "grad_norm": 0.8700173580337157,
      "learning_rate": 7.58010200946755e-05,
      "loss": 0.4934,
      "step": 78
    },
    {
      "epoch": 0.7097136440202133,
      "grad_norm": 0.9327016278300901,
      "learning_rate": 7.561159624945257e-05,
      "loss": 0.4823,
      "step": 79
    },
    {
      "epoch": 0.7186973610331274,
      "grad_norm": 1.357805985616199,
      "learning_rate": 7.541824102612839e-05,
      "loss": 0.5005,
      "step": 80
    },
    {
      "epoch": 0.7276810780460415,
      "grad_norm": 0.8155442555029125,
      "learning_rate": 7.5220975770348e-05,
      "loss": 0.4846,
      "step": 81
    },
    {
      "epoch": 0.7366647950589557,
      "grad_norm": 0.8879306236067503,
      "learning_rate": 7.501982225940833e-05,
      "loss": 0.4716,
      "step": 82
    },
    {
      "epoch": 0.7456485120718698,
      "grad_norm": 1.1674726644441007,
      "learning_rate": 7.48148026998542e-05,
      "loss": 0.4912,
      "step": 83
    },
    {
      "epoch": 0.7546322290847838,
      "grad_norm": 0.9150265900061878,
      "learning_rate": 7.460593972502674e-05,
      "loss": 0.4857,
      "step": 84
    },
    {
      "epoch": 0.7636159460976979,
      "grad_norm": 1.2818859952439805,
      "learning_rate": 7.439325639256483e-05,
      "loss": 0.4815,
      "step": 85
    },
    {
      "epoch": 0.772599663110612,
      "grad_norm": 0.7413198899597869,
      "learning_rate": 7.417677618185955e-05,
      "loss": 0.4837,
      "step": 86
    },
    {
      "epoch": 0.7815833801235261,
      "grad_norm": 0.9478143258160291,
      "learning_rate": 7.39565229914622e-05,
      "loss": 0.4765,
      "step": 87
    },
    {
      "epoch": 0.7905670971364402,
      "grad_norm": 0.963846436149164,
      "learning_rate": 7.373252113644596e-05,
      "loss": 0.4879,
      "step": 88
    },
    {
      "epoch": 0.7995508141493542,
      "grad_norm": 0.9081468443162725,
      "learning_rate": 7.350479534572166e-05,
      "loss": 0.4705,
      "step": 89
    },
    {
      "epoch": 0.8085345311622684,
      "grad_norm": 0.6474133013047098,
      "learning_rate": 7.327337075930775e-05,
      "loss": 0.4806,
      "step": 90
    },
    {
      "epoch": 0.8175182481751825,
      "grad_norm": 0.6809036448679276,
      "learning_rate": 7.303827292555495e-05,
      "loss": 0.481,
      "step": 91
    },
    {
      "epoch": 0.8265019651880966,
      "grad_norm": 1.0720011034315253,
      "learning_rate": 7.279952779832584e-05,
      "loss": 0.4759,
      "step": 92
    },
    {
      "epoch": 0.8354856822010107,
      "grad_norm": 1.0029995670688614,
      "learning_rate": 7.255716173412966e-05,
      "loss": 0.4736,
      "step": 93
    },
    {
      "epoch": 0.8444693992139247,
      "grad_norm": 0.8353160428003878,
      "learning_rate": 7.23112014892126e-05,
      "loss": 0.4754,
      "step": 94
    },
    {
      "epoch": 0.8534531162268388,
      "grad_norm": 0.9066101057502527,
      "learning_rate": 7.20616742166041e-05,
      "loss": 0.4801,
      "step": 95
    },
    {
      "epoch": 0.862436833239753,
      "grad_norm": 1.3700422899595712,
      "learning_rate": 7.180860746311917e-05,
      "loss": 0.4904,
      "step": 96
    },
    {
      "epoch": 0.8714205502526671,
      "grad_norm": 0.8940940289737819,
      "learning_rate": 7.155202916631743e-05,
      "loss": 0.4824,
      "step": 97
    },
    {
      "epoch": 0.8804042672655812,
      "grad_norm": 1.3226382972793307,
      "learning_rate": 7.129196765141886e-05,
      "loss": 0.4679,
      "step": 98
    },
    {
      "epoch": 0.8893879842784952,
      "grad_norm": 0.8696732708344425,
      "learning_rate": 7.10284516281768e-05,
      "loss": 0.4786,
      "step": 99
    },
    {
      "epoch": 0.8983717012914093,
      "grad_norm": 1.0583061403429452,
      "learning_rate": 7.076151018770854e-05,
      "loss": 0.4669,
      "step": 100
    },
    {
      "epoch": 0.9073554183043234,
      "grad_norm": 1.0722660670077375,
      "learning_rate": 7.049117279928374e-05,
      "loss": 0.4801,
      "step": 101
    },
    {
      "epoch": 0.9163391353172375,
      "grad_norm": 1.2855639166116766,
      "learning_rate": 7.021746930707117e-05,
      "loss": 0.4674,
      "step": 102
    },
    {
      "epoch": 0.9253228523301515,
      "grad_norm": 0.6812836053141724,
      "learning_rate": 6.994042992684406e-05,
      "loss": 0.4788,
      "step": 103
    },
    {
      "epoch": 0.9343065693430657,
      "grad_norm": 0.6913667167229325,
      "learning_rate": 6.966008524264429e-05,
      "loss": 0.4679,
      "step": 104
    },
    {
      "epoch": 0.9432902863559798,
      "grad_norm": 0.9914419206668401,
      "learning_rate": 6.937646620340618e-05,
      "loss": 0.4792,
      "step": 105
    },
    {
      "epoch": 0.9522740033688939,
      "grad_norm": 1.3797231509055095,
      "learning_rate": 6.908960411953973e-05,
      "loss": 0.4683,
      "step": 106
    },
    {
      "epoch": 0.961257720381808,
      "grad_norm": 0.5888339311760277,
      "learning_rate": 6.879953065947416e-05,
      "loss": 0.4678,
      "step": 107
    },
    {
      "epoch": 0.970241437394722,
      "grad_norm": 1.038793566408693,
      "learning_rate": 6.850627784616178e-05,
      "loss": 0.4601,
      "step": 108
    },
    {
      "epoch": 0.9792251544076361,
      "grad_norm": 1.0842893632221857,
      "learning_rate": 6.82098780535428e-05,
      "loss": 0.4681,
      "step": 109
    },
    {
      "epoch": 0.9882088714205502,
      "grad_norm": 0.8626180542597122,
      "learning_rate": 6.791036400297142e-05,
      "loss": 0.4761,
      "step": 110
    },
    {
      "epoch": 0.9971925884334644,
      "grad_norm": 0.7965083948145554,
      "learning_rate": 6.760776875960347e-05,
      "loss": 0.4689,
      "step": 111
    },
    {
      "epoch": 1.0075800112296462,
      "grad_norm": 1.786141860362385,
      "learning_rate": 6.730212572874618e-05,
      "loss": 0.8334,
      "step": 112
    },
    {
      "epoch": 1.0165637282425604,
      "grad_norm": 0.8882542778931486,
      "learning_rate": 6.699346865217031e-05,
      "loss": 0.4299,
      "step": 113
    },
    {
      "epoch": 1.0255474452554745,
      "grad_norm": 0.6042242319747863,
      "learning_rate": 6.668183160438531e-05,
      "loss": 0.4431,
      "step": 114
    },
    {
      "epoch": 1.0345311622683886,
      "grad_norm": 0.5382725444188637,
      "learning_rate": 6.636724898887751e-05,
      "loss": 0.4464,
      "step": 115
    },
    {
      "epoch": 1.0435148792813027,
      "grad_norm": 0.6092419399947249,
      "learning_rate": 6.604975553431219e-05,
      "loss": 0.45,
      "step": 116
    },
    {
      "epoch": 1.0524985962942168,
      "grad_norm": 0.952732534414775,
      "learning_rate": 6.572938629069959e-05,
      "loss": 0.4416,
      "step": 117
    },
    {
      "epoch": 1.0614823133071307,
      "grad_norm": 1.2586357901517062,
      "learning_rate": 6.540617662552565e-05,
      "loss": 0.4478,
      "step": 118
    },
    {
      "epoch": 1.0704660303200448,
      "grad_norm": 0.5882439478167614,
      "learning_rate": 6.508016221984747e-05,
      "loss": 0.4346,
      "step": 119
    },
    {
      "epoch": 1.079449747332959,
      "grad_norm": 0.8126561068118403,
      "learning_rate": 6.475137906435435e-05,
      "loss": 0.4373,
      "step": 120
    },
    {
      "epoch": 1.088433464345873,
      "grad_norm": 1.5656070861661977,
      "learning_rate": 6.441986345539446e-05,
      "loss": 0.4461,
      "step": 121
    },
    {
      "epoch": 1.0974171813587872,
      "grad_norm": 0.5066258143638921,
      "learning_rate": 6.408565199096798e-05,
      "loss": 0.4327,
      "step": 122
    },
    {
      "epoch": 1.1064008983717013,
      "grad_norm": 1.6682432153253044,
      "learning_rate": 6.374878156668676e-05,
      "loss": 0.4568,
      "step": 123
    },
    {
      "epoch": 1.1153846153846154,
      "grad_norm": 0.6299972385025496,
      "learning_rate": 6.340928937170118e-05,
      "loss": 0.4285,
      "step": 124
    },
    {
      "epoch": 1.1243683323975295,
      "grad_norm": 1.9783962130226307,
      "learning_rate": 6.30672128845947e-05,
      "loss": 0.4643,
      "step": 125
    },
    {
      "epoch": 1.1333520494104437,
      "grad_norm": 1.2699186089072336,
      "learning_rate": 6.272258986924624e-05,
      "loss": 0.4362,
      "step": 126
    },
    {
      "epoch": 1.1423357664233578,
      "grad_norm": 1.617840953040159,
      "learning_rate": 6.237545837066133e-05,
      "loss": 0.4476,
      "step": 127
    },
    {
      "epoch": 1.1513194834362717,
      "grad_norm": 1.5459662606407818,
      "learning_rate": 6.202585671077204e-05,
      "loss": 0.4452,
      "step": 128
    },
    {
      "epoch": 1.1603032004491858,
      "grad_norm": 0.8520351440042236,
      "learning_rate": 6.167382348420637e-05,
      "loss": 0.4485,
      "step": 129
    },
    {
      "epoch": 1.1692869174621,
      "grad_norm": 1.2109454920551714,
      "learning_rate": 6.131939755402755e-05,
      "loss": 0.4436,
      "step": 130
    },
    {
      "epoch": 1.178270634475014,
      "grad_norm": 0.7621427547756707,
      "learning_rate": 6.09626180474438e-05,
      "loss": 0.4436,
      "step": 131
    },
    {
      "epoch": 1.1872543514879281,
      "grad_norm": 0.9377335412029888,
      "learning_rate": 6.060352435148874e-05,
      "loss": 0.4404,
      "step": 132
    },
    {
      "epoch": 1.1962380685008422,
      "grad_norm": 0.742883271426497,
      "learning_rate": 6.024215610867327e-05,
      "loss": 0.4431,
      "step": 133
    },
    {
      "epoch": 1.2052217855137564,
      "grad_norm": 0.623519085897906,
      "learning_rate": 5.9878553212609184e-05,
      "loss": 0.4349,
      "step": 134
    },
    {
      "epoch": 1.2142055025266705,
      "grad_norm": 0.5738395916616672,
      "learning_rate": 5.95127558036051e-05,
      "loss": 0.4383,
      "step": 135
    },
    {
      "epoch": 1.2231892195395846,
      "grad_norm": 0.5745090863281419,
      "learning_rate": 5.9144804264235066e-05,
      "loss": 0.4353,
      "step": 136
    },
    {
      "epoch": 1.2321729365524985,
      "grad_norm": 0.6173172910448796,
      "learning_rate": 5.8774739214880554e-05,
      "loss": 0.4397,
      "step": 137
    },
    {
      "epoch": 1.2411566535654126,
      "grad_norm": 0.44131140572281446,
      "learning_rate": 5.840260150924609e-05,
      "loss": 0.4358,
      "step": 138
    },
    {
      "epoch": 1.2501403705783267,
      "grad_norm": 0.48929805430187906,
      "learning_rate": 5.802843222984919e-05,
      "loss": 0.4348,
      "step": 139
    },
    {
      "epoch": 1.2591240875912408,
      "grad_norm": 0.46891906471051287,
      "learning_rate": 5.765227268348501e-05,
      "loss": 0.433,
      "step": 140
    },
    {
      "epoch": 1.268107804604155,
      "grad_norm": 0.4197153657865162,
      "learning_rate": 5.727416439666622e-05,
      "loss": 0.4373,
      "step": 141
    },
    {
      "epoch": 1.277091521617069,
      "grad_norm": 0.49836695670940545,
      "learning_rate": 5.689414911103867e-05,
      "loss": 0.4339,
      "step": 142
    },
    {
      "epoch": 1.2860752386299832,
      "grad_norm": 0.4282436546394893,
      "learning_rate": 5.651226877877326e-05,
      "loss": 0.4371,
      "step": 143
    },
    {
      "epoch": 1.2950589556428973,
      "grad_norm": 0.39801460777108744,
      "learning_rate": 5.612856555793459e-05,
      "loss": 0.4422,
      "step": 144
    },
    {
      "epoch": 1.3040426726558114,
      "grad_norm": 0.384138892088657,
      "learning_rate": 5.574308180782693e-05,
      "loss": 0.4235,
      "step": 145
    },
    {
      "epoch": 1.3130263896687255,
      "grad_norm": 0.4344092290487113,
      "learning_rate": 5.5355860084317787e-05,
      "loss": 0.4356,
      "step": 146
    },
    {
      "epoch": 1.3220101066816397,
      "grad_norm": 0.37928071414714964,
      "learning_rate": 5.496694313514009e-05,
      "loss": 0.4362,
      "step": 147
    },
    {
      "epoch": 1.3309938236945535,
      "grad_norm": 0.34476337733896867,
      "learning_rate": 5.457637389517285e-05,
      "loss": 0.4398,
      "step": 148
    },
    {
      "epoch": 1.3399775407074677,
      "grad_norm": 0.4506193928055396,
      "learning_rate": 5.4184195481701425e-05,
      "loss": 0.428,
      "step": 149
    },
    {
      "epoch": 1.3489612577203818,
      "grad_norm": 0.32358078080164415,
      "learning_rate": 5.3790451189657486e-05,
      "loss": 0.4354,
      "step": 150
    },
    {
      "epoch": 1.357944974733296,
      "grad_norm": 0.3714258728709878,
      "learning_rate": 5.339518448683945e-05,
      "loss": 0.4336,
      "step": 151
    },
    {
      "epoch": 1.36692869174621,
      "grad_norm": 0.3698334511165891,
      "learning_rate": 5.2998439009113814e-05,
      "loss": 0.421,
      "step": 152
    },
    {
      "epoch": 1.3759124087591241,
      "grad_norm": 0.3039344886888513,
      "learning_rate": 5.260025855559792e-05,
      "loss": 0.4265,
      "step": 153
    },
    {
      "epoch": 1.3848961257720382,
      "grad_norm": 0.38707783731390033,
      "learning_rate": 5.2200687083824706e-05,
      "loss": 0.4285,
      "step": 154
    },
    {
      "epoch": 1.3938798427849521,
      "grad_norm": 0.3708434779281009,
      "learning_rate": 5.179976870488999e-05,
      "loss": 0.4242,
      "step": 155
    },
    {
      "epoch": 1.4028635597978663,
      "grad_norm": 0.2827560899827463,
      "learning_rate": 5.1397547678582745e-05,
      "loss": 0.4215,
      "step": 156
    },
    {
      "epoch": 1.4118472768107804,
      "grad_norm": 0.36357591101838377,
      "learning_rate": 5.099406840849902e-05,
      "loss": 0.4344,
      "step": 157
    },
    {
      "epoch": 1.4208309938236945,
      "grad_norm": 0.2815376075683387,
      "learning_rate": 5.058937543713999e-05,
      "loss": 0.4234,
      "step": 158
    },
    {
      "epoch": 1.4298147108366086,
      "grad_norm": 0.34011781572335237,
      "learning_rate": 5.018351344099453e-05,
      "loss": 0.438,
      "step": 159
    },
    {
      "epoch": 1.4387984278495227,
      "grad_norm": 0.32400444292287883,
      "learning_rate": 4.9776527225607274e-05,
      "loss": 0.4328,
      "step": 160
    },
    {
      "epoch": 1.4477821448624368,
      "grad_norm": 0.2529768229883554,
      "learning_rate": 4.93684617206321e-05,
      "loss": 0.4339,
      "step": 161
    },
    {
      "epoch": 1.456765861875351,
      "grad_norm": 0.2629261907748739,
      "learning_rate": 4.89593619748722e-05,
      "loss": 0.4394,
      "step": 162
    },
    {
      "epoch": 1.465749578888265,
      "grad_norm": 0.32833332654685193,
      "learning_rate": 4.8549273151306795e-05,
      "loss": 0.4331,
      "step": 163
    },
    {
      "epoch": 1.4747332959011792,
      "grad_norm": 0.37621666962913386,
      "learning_rate": 4.8138240522105365e-05,
      "loss": 0.4294,
      "step": 164
    },
    {
      "epoch": 1.4837170129140933,
      "grad_norm": 0.33680087947524656,
      "learning_rate": 4.7726309463629733e-05,
      "loss": 0.4317,
      "step": 165
    },
    {
      "epoch": 1.4927007299270074,
      "grad_norm": 0.22831905889281706,
      "learning_rate": 4.731352545142478e-05,
      "loss": 0.4265,
      "step": 166
    },
    {
      "epoch": 1.5016844469399215,
      "grad_norm": 0.23855204275905859,
      "learning_rate": 4.689993405519802e-05,
      "loss": 0.4343,
      "step": 167
    },
    {
      "epoch": 1.5106681639528357,
      "grad_norm": 0.3165956064567229,
      "learning_rate": 4.648558093378899e-05,
      "loss": 0.4331,
      "step": 168
    },
    {
      "epoch": 1.5196518809657495,
      "grad_norm": 0.3792988696086109,
      "learning_rate": 4.607051183012862e-05,
      "loss": 0.4214,
      "step": 169
    },
    {
      "epoch": 1.5286355979786637,
      "grad_norm": 0.29743099805783,
      "learning_rate": 4.5654772566189415e-05,
      "loss": 0.4322,
      "step": 170
    },
    {
      "epoch": 1.5376193149915778,
      "grad_norm": 0.22315650518088825,
      "learning_rate": 4.5238409037926905e-05,
      "loss": 0.4286,
      "step": 171
    },
    {
      "epoch": 1.546603032004492,
      "grad_norm": 0.3745132482990998,
      "learning_rate": 4.4821467210212924e-05,
      "loss": 0.4325,
      "step": 172
    },
    {
      "epoch": 1.5555867490174058,
      "grad_norm": 0.4349435850403825,
      "learning_rate": 4.4403993111761265e-05,
      "loss": 0.4278,
      "step": 173
    },
    {
      "epoch": 1.56457046603032,
      "grad_norm": 0.34417305619739286,
      "learning_rate": 4.398603283004626e-05,
      "loss": 0.421,
      "step": 174
    },
    {
      "epoch": 1.573554183043234,
      "grad_norm": 0.2697125284825644,
      "learning_rate": 4.356763250621496e-05,
      "loss": 0.4355,
      "step": 175
    },
    {
      "epoch": 1.5825379000561481,
      "grad_norm": 0.3759486979388303,
      "learning_rate": 4.314883832999326e-05,
      "loss": 0.4336,
      "step": 176
    },
    {
      "epoch": 1.5915216170690623,
      "grad_norm": 0.42378540392492775,
      "learning_rate": 4.272969653458685e-05,
      "loss": 0.4258,
      "step": 177
    },
    {
      "epoch": 1.6005053340819764,
      "grad_norm": 0.36295049407047797,
      "learning_rate": 4.231025339157714e-05,
      "loss": 0.4214,
      "step": 178
    },
    {
      "epoch": 1.6094890510948905,
      "grad_norm": 0.22472885668491016,
      "learning_rate": 4.189055520581315e-05,
      "loss": 0.4302,
      "step": 179
    },
    {
      "epoch": 1.6184727681078046,
      "grad_norm": 0.24842552091044895,
      "learning_rate": 4.147064831029959e-05,
      "loss": 0.4267,
      "step": 180
    },
    {
      "epoch": 1.6274564851207187,
      "grad_norm": 0.27444720145312784,
      "learning_rate": 4.105057906108189e-05,
      "loss": 0.4306,
      "step": 181
    },
    {
      "epoch": 1.6364402021336328,
      "grad_norm": 0.24170155717400021,
      "learning_rate": 4.063039383212866e-05,
      "loss": 0.4226,
      "step": 182
    },
    {
      "epoch": 1.645423919146547,
      "grad_norm": 0.2158635971279587,
      "learning_rate": 4.021013901021225e-05,
      "loss": 0.4182,
      "step": 183
    },
    {
      "epoch": 1.654407636159461,
      "grad_norm": 0.2392521978049934,
      "learning_rate": 3.978986098978777e-05,
      "loss": 0.4184,
      "step": 184
    },
    {
      "epoch": 1.6633913531723752,
      "grad_norm": 0.2333669024637109,
      "learning_rate": 3.936960616787135e-05,
      "loss": 0.4272,
      "step": 185
    },
    {
      "epoch": 1.6723750701852893,
      "grad_norm": 0.303971238157218,
      "learning_rate": 3.8949420938918124e-05,
      "loss": 0.4278,
      "step": 186
    },
    {
      "epoch": 1.6813587871982034,
      "grad_norm": 0.20260216895624977,
      "learning_rate": 3.852935168970042e-05,
      "loss": 0.4274,
      "step": 187
    },
    {
      "epoch": 1.6903425042111173,
      "grad_norm": 0.27940230833327445,
      "learning_rate": 3.810944479418686e-05,
      "loss": 0.4235,
      "step": 188
    },
    {
      "epoch": 1.6993262212240314,
      "grad_norm": 0.24151727930977668,
      "learning_rate": 3.768974660842287e-05,
      "loss": 0.43,
      "step": 189
    },
    {
      "epoch": 1.7083099382369455,
      "grad_norm": 0.28305027971796287,
      "learning_rate": 3.727030346541317e-05,
      "loss": 0.4204,
      "step": 190
    },
    {
      "epoch": 1.7172936552498597,
      "grad_norm": 0.23944450177743382,
      "learning_rate": 3.685116167000675e-05,
      "loss": 0.4303,
      "step": 191
    },
    {
      "epoch": 1.7262773722627736,
      "grad_norm": 0.2649036336219584,
      "learning_rate": 3.6432367493785056e-05,
      "loss": 0.4306,
      "step": 192
    },
    {
      "epoch": 1.7352610892756877,
      "grad_norm": 0.20536449694593853,
      "learning_rate": 3.601396716995375e-05,
      "loss": 0.4298,
      "step": 193
    },
    {
      "epoch": 1.7442448062886018,
      "grad_norm": 0.2752898250202777,
      "learning_rate": 3.559600688823875e-05,
      "loss": 0.4213,
      "step": 194
    },
    {
      "epoch": 1.753228523301516,
      "grad_norm": 0.2441828080744484,
      "learning_rate": 3.517853278978708e-05,
      "loss": 0.4317,
      "step": 195
    },
    {
      "epoch": 1.76221224031443,
      "grad_norm": 0.18921602583278035,
      "learning_rate": 3.4761590962073115e-05,
      "loss": 0.4352,
      "step": 196
    },
    {
      "epoch": 1.7711959573273441,
      "grad_norm": 0.25381845245915774,
      "learning_rate": 3.434522743381061e-05,
      "loss": 0.4227,
      "step": 197
    },
    {
      "epoch": 1.7801796743402583,
      "grad_norm": 0.22098025447635186,
      "learning_rate": 3.39294881698714e-05,
      "loss": 0.4145,
      "step": 198
    },
    {
      "epoch": 1.7891633913531724,
      "grad_norm": 0.21253762725360797,
      "learning_rate": 3.3514419066211025e-05,
      "loss": 0.4204,
      "step": 199
    },
    {
      "epoch": 1.7981471083660865,
      "grad_norm": 0.1993914725116837,
      "learning_rate": 3.310006594480199e-05,
      "loss": 0.4212,
      "step": 200
    },
    {
      "epoch": 1.8071308253790006,
      "grad_norm": 0.21949029430642605,
      "learning_rate": 3.268647454857524e-05,
      "loss": 0.4246,
      "step": 201
    },
    {
      "epoch": 1.8161145423919147,
      "grad_norm": 0.18093321704969498,
      "learning_rate": 3.227369053637028e-05,
      "loss": 0.4379,
      "step": 202
    },
    {
      "epoch": 1.8250982594048288,
      "grad_norm": 0.21469795266380035,
      "learning_rate": 3.1861759477894656e-05,
      "loss": 0.4227,
      "step": 203
    },
    {
      "epoch": 1.834081976417743,
      "grad_norm": 0.18433389487278787,
      "learning_rate": 3.145072684869322e-05,
      "loss": 0.4166,
      "step": 204
    },
    {
      "epoch": 1.843065693430657,
      "grad_norm": 0.19979893625833967,
      "learning_rate": 3.104063802512782e-05,
      "loss": 0.4281,
      "step": 205
    },
    {
      "epoch": 1.8520494104435712,
      "grad_norm": 0.19103290364400521,
      "learning_rate": 3.063153827936792e-05,
      "loss": 0.4161,
      "step": 206
    },
    {
      "epoch": 1.861033127456485,
      "grad_norm": 0.21106537662314218,
      "learning_rate": 3.0223472774392753e-05,
      "loss": 0.4242,
      "step": 207
    },
    {
      "epoch": 1.8700168444693992,
      "grad_norm": 0.20844328823177255,
      "learning_rate": 2.9816486559005482e-05,
      "loss": 0.4333,
      "step": 208
    },
    {
      "epoch": 1.8790005614823133,
      "grad_norm": 0.19732371134712767,
      "learning_rate": 2.9410624562860026e-05,
      "loss": 0.4211,
      "step": 209
    },
    {
      "epoch": 1.8879842784952274,
      "grad_norm": 0.23608540869194397,
      "learning_rate": 2.9005931591500974e-05,
      "loss": 0.422,
      "step": 210
    },
    {
      "epoch": 1.8969679955081415,
      "grad_norm": 0.18142970460558788,
      "learning_rate": 2.860245232141726e-05,
      "loss": 0.4215,
      "step": 211
    },
    {
      "epoch": 1.9059517125210554,
      "grad_norm": 0.21920791583501226,
      "learning_rate": 2.8200231295110012e-05,
      "loss": 0.4304,
      "step": 212
    },
    {
      "epoch": 1.9149354295339696,
      "grad_norm": 0.1575572213846906,
      "learning_rate": 2.7799312916175294e-05,
      "loss": 0.4115,
      "step": 213
    },
    {
      "epoch": 1.9239191465468837,
      "grad_norm": 0.21181976577415998,
      "learning_rate": 2.7399741444402087e-05,
      "loss": 0.4301,
      "step": 214
    },
    {
      "epoch": 1.9329028635597978,
      "grad_norm": 0.1817247821525619,
      "learning_rate": 2.7001560990886196e-05,
      "loss": 0.4202,
      "step": 215
    },
    {
      "epoch": 1.941886580572712,
      "grad_norm": 0.17883228911738394,
      "learning_rate": 2.6604815513160556e-05,
      "loss": 0.4241,
      "step": 216
    },
    {
      "epoch": 1.950870297585626,
      "grad_norm": 0.19947193861762386,
      "learning_rate": 2.6209548810342517e-05,
      "loss": 0.4137,
      "step": 217
    },
    {
      "epoch": 1.9598540145985401,
      "grad_norm": 0.19071364537376845,
      "learning_rate": 2.5815804518298575e-05,
      "loss": 0.4327,
      "step": 218
    },
    {
      "epoch": 1.9688377316114543,
      "grad_norm": 0.1712467330551581,
      "learning_rate": 2.542362610482715e-05,
      "loss": 0.4239,
      "step": 219
    },
    {
      "epoch": 1.9778214486243684,
      "grad_norm": 0.16971518326737378,
      "learning_rate": 2.503305686485991e-05,
      "loss": 0.4197,
      "step": 220
    },
    {
      "epoch": 1.9868051656372825,
      "grad_norm": 0.15655415639126638,
      "learning_rate": 2.464413991568222e-05,
      "loss": 0.4085,
      "step": 221
    },
    {
      "epoch": 1.9957888826501966,
      "grad_norm": 0.19809705687476845,
      "learning_rate": 2.4256918192173088e-05,
      "loss": 0.4325,
      "step": 222
    },
    {
      "epoch": 2.0061763054463784,
      "grad_norm": 0.32148951189801367,
      "learning_rate": 2.3871434442065414e-05,
      "loss": 0.7473,
      "step": 223
    },
    {
      "epoch": 2.0151600224592925,
      "grad_norm": 0.20734628721061774,
      "learning_rate": 2.3487731221226754e-05,
      "loss": 0.3817,
      "step": 224
    },
    {
      "epoch": 2.0241437394722066,
      "grad_norm": 0.2283335322281334,
      "learning_rate": 2.3105850888961348e-05,
      "loss": 0.4043,
      "step": 225
    },
    {
      "epoch": 2.0331274564851207,
      "grad_norm": 0.20409956046524397,
      "learning_rate": 2.272583560333379e-05,
      "loss": 0.3908,
      "step": 226
    },
    {
      "epoch": 2.042111173498035,
      "grad_norm": 0.21233165191280134,
      "learning_rate": 2.2347727316515e-05,
      "loss": 0.3963,
      "step": 227
    },
    {
      "epoch": 2.051094890510949,
      "grad_norm": 0.2058022752445305,
      "learning_rate": 2.1971567770150814e-05,
      "loss": 0.3875,
      "step": 228
    },
    {
      "epoch": 2.060078607523863,
      "grad_norm": 0.2202126073153398,
      "learning_rate": 2.1597398490753917e-05,
      "loss": 0.4031,
      "step": 229
    },
    {
      "epoch": 2.069062324536777,
      "grad_norm": 0.20478388239536333,
      "learning_rate": 2.1225260785119456e-05,
      "loss": 0.3891,
      "step": 230
    },
    {
      "epoch": 2.0780460415496913,
      "grad_norm": 0.1937488744583445,
      "learning_rate": 2.0855195735764947e-05,
      "loss": 0.3906,
      "step": 231
    },
    {
      "epoch": 2.0870297585626054,
      "grad_norm": 0.2163161528806818,
      "learning_rate": 2.0487244196394912e-05,
      "loss": 0.3941,
      "step": 232
    },
    {
      "epoch": 2.0960134755755195,
      "grad_norm": 0.18524457129950797,
      "learning_rate": 2.0121446787390822e-05,
      "loss": 0.3799,
      "step": 233
    },
    {
      "epoch": 2.1049971925884337,
      "grad_norm": 0.217114553144753,
      "learning_rate": 1.9757843891326736e-05,
      "loss": 0.3951,
      "step": 234
    },
    {
      "epoch": 2.1139809096013478,
      "grad_norm": 0.2017737198987517,
      "learning_rate": 1.939647564851127e-05,
      "loss": 0.3928,
      "step": 235
    },
    {
      "epoch": 2.1229646266142614,
      "grad_norm": 0.20060221109136825,
      "learning_rate": 1.9037381952556217e-05,
      "loss": 0.3891,
      "step": 236
    },
    {
      "epoch": 2.1319483436271756,
      "grad_norm": 0.19362260661168698,
      "learning_rate": 1.8680602445972463e-05,
      "loss": 0.3967,
      "step": 237
    },
    {
      "epoch": 2.1409320606400897,
      "grad_norm": 0.1898518071335132,
      "learning_rate": 1.832617651579365e-05,
      "loss": 0.3859,
      "step": 238
    },
    {
      "epoch": 2.149915777653004,
      "grad_norm": 0.1817097392847635,
      "learning_rate": 1.797414328922797e-05,
      "loss": 0.3878,
      "step": 239
    },
    {
      "epoch": 2.158899494665918,
      "grad_norm": 0.1438401587120875,
      "learning_rate": 1.7624541629338676e-05,
      "loss": 0.3902,
      "step": 240
    },
    {
      "epoch": 2.167883211678832,
      "grad_norm": 0.1785624789625333,
      "learning_rate": 1.7277410130753775e-05,
      "loss": 0.3803,
      "step": 241
    },
    {
      "epoch": 2.176866928691746,
      "grad_norm": 0.16410314706005405,
      "learning_rate": 1.6932787115405318e-05,
      "loss": 0.4038,
      "step": 242
    },
    {
      "epoch": 2.1858506457046603,
      "grad_norm": 0.17495931093245287,
      "learning_rate": 1.6590710628298826e-05,
      "loss": 0.4097,
      "step": 243
    },
    {
      "epoch": 2.1948343627175744,
      "grad_norm": 0.15732504490550855,
      "learning_rate": 1.6251218433313254e-05,
      "loss": 0.3841,
      "step": 244
    },
    {
      "epoch": 2.2038180797304885,
      "grad_norm": 0.15370170593613078,
      "learning_rate": 1.591434800903203e-05,
      "loss": 0.3879,
      "step": 245
    },
    {
      "epoch": 2.2128017967434026,
      "grad_norm": 0.14225993974716095,
      "learning_rate": 1.558013654460555e-05,
      "loss": 0.3831,
      "step": 246
    },
    {
      "epoch": 2.2217855137563167,
      "grad_norm": 0.16769227472528905,
      "learning_rate": 1.5248620935645666e-05,
      "loss": 0.3932,
      "step": 247
    },
    {
      "epoch": 2.230769230769231,
      "grad_norm": 0.15697609805604978,
      "learning_rate": 1.4919837780152544e-05,
      "loss": 0.3793,
      "step": 248
    },
    {
      "epoch": 2.239752947782145,
      "grad_norm": 0.14834997366132227,
      "learning_rate": 1.4593823374474374e-05,
      "loss": 0.3884,
      "step": 249
    },
    {
      "epoch": 2.248736664795059,
      "grad_norm": 0.1416591291545358,
      "learning_rate": 1.4270613709300429e-05,
      "loss": 0.3883,
      "step": 250
    },
    {
      "epoch": 2.257720381807973,
      "grad_norm": 0.1416137601943575,
      "learning_rate": 1.3950244465687833e-05,
      "loss": 0.3925,
      "step": 251
    },
    {
      "epoch": 2.2667040988208873,
      "grad_norm": 0.14486611430866134,
      "learning_rate": 1.3632751011122497e-05,
      "loss": 0.3988,
      "step": 252
    },
    {
      "epoch": 2.2756878158338014,
      "grad_norm": 0.1438931082489217,
      "learning_rate": 1.3318168395614697e-05,
      "loss": 0.3857,
      "step": 253
    },
    {
      "epoch": 2.2846715328467155,
      "grad_norm": 0.1376572002883416,
      "learning_rate": 1.3006531347829699e-05,
      "loss": 0.4054,
      "step": 254
    },
    {
      "epoch": 2.293655249859629,
      "grad_norm": 0.13374618220430773,
      "learning_rate": 1.2697874271253844e-05,
      "loss": 0.389,
      "step": 255
    },
    {
      "epoch": 2.3026389668725433,
      "grad_norm": 0.1344986927289295,
      "learning_rate": 1.2392231240396542e-05,
      "loss": 0.398,
      "step": 256
    },
    {
      "epoch": 2.3116226838854574,
      "grad_norm": 0.13248310416193929,
      "learning_rate": 1.2089635997028592e-05,
      "loss": 0.3964,
      "step": 257
    },
    {
      "epoch": 2.3206064008983716,
      "grad_norm": 0.1250202978575819,
      "learning_rate": 1.1790121946457212e-05,
      "loss": 0.3827,
      "step": 258
    },
    {
      "epoch": 2.3295901179112857,
      "grad_norm": 0.1267954036736396,
      "learning_rate": 1.1493722153838239e-05,
      "loss": 0.403,
      "step": 259
    },
    {
      "epoch": 2.3385738349242,
      "grad_norm": 0.11963916536373062,
      "learning_rate": 1.120046934052585e-05,
      "loss": 0.3774,
      "step": 260
    },
    {
      "epoch": 2.347557551937114,
      "grad_norm": 0.11989536513409248,
      "learning_rate": 1.0910395880460274e-05,
      "loss": 0.3994,
      "step": 261
    },
    {
      "epoch": 2.356541268950028,
      "grad_norm": 0.13086834428014207,
      "learning_rate": 1.062353379659383e-05,
      "loss": 0.3959,
      "step": 262
    },
    {
      "epoch": 2.365524985962942,
      "grad_norm": 0.12262091225654496,
      "learning_rate": 1.0339914757355718e-05,
      "loss": 0.3909,
      "step": 263
    },
    {
      "epoch": 2.3745087029758563,
      "grad_norm": 0.13332341565208733,
      "learning_rate": 1.0059570073155953e-05,
      "loss": 0.3964,
      "step": 264
    },
    {
      "epoch": 2.3834924199887704,
      "grad_norm": 0.1390720043482773,
      "learning_rate": 9.782530692928832e-06,
      "loss": 0.3962,
      "step": 265
    },
    {
      "epoch": 2.3924761370016845,
      "grad_norm": 0.12593592732440334,
      "learning_rate": 9.508827200716273e-06,
      "loss": 0.3987,
      "step": 266
    },
    {
      "epoch": 2.4014598540145986,
      "grad_norm": 0.14058113905698136,
      "learning_rate": 9.238489812291469e-06,
      "loss": 0.3851,
      "step": 267
    },
    {
      "epoch": 2.4104435710275127,
      "grad_norm": 0.1293391140497839,
      "learning_rate": 8.971548371823205e-06,
      "loss": 0.3972,
      "step": 268
    },
    {
      "epoch": 2.419427288040427,
      "grad_norm": 0.1312457950118106,
      "learning_rate": 8.708032348581144e-06,
      "loss": 0.401,
      "step": 269
    },
    {
      "epoch": 2.428411005053341,
      "grad_norm": 0.13054369787139117,
      "learning_rate": 8.447970833682584e-06,
      "loss": 0.3801,
      "step": 270
    },
    {
      "epoch": 2.437394722066255,
      "grad_norm": 0.12662575874177268,
      "learning_rate": 8.191392536880852e-06,
      "loss": 0.3835,
      "step": 271
    },
    {
      "epoch": 2.446378439079169,
      "grad_norm": 0.14035374781938695,
      "learning_rate": 7.938325783395924e-06,
      "loss": 0.3924,
      "step": 272
    },
    {
      "epoch": 2.4553621560920833,
      "grad_norm": 0.11900240575463085,
      "learning_rate": 7.68879851078741e-06,
      "loss": 0.3839,
      "step": 273
    },
    {
      "epoch": 2.464345873104997,
      "grad_norm": 0.13981444396533035,
      "learning_rate": 7.442838265870347e-06,
      "loss": 0.4029,
      "step": 274
    },
    {
      "epoch": 2.473329590117911,
      "grad_norm": 0.13223802646070204,
      "learning_rate": 7.2004722016741605e-06,
      "loss": 0.3933,
      "step": 275
    },
    {
      "epoch": 2.482313307130825,
      "grad_norm": 0.1225885147923774,
      "learning_rate": 6.961727074445055e-06,
      "loss": 0.3984,
      "step": 276
    },
    {
      "epoch": 2.4912970241437393,
      "grad_norm": 0.11467365860762357,
      "learning_rate": 6.726629240692255e-06,
      "loss": 0.378,
      "step": 277
    },
    {
      "epoch": 2.5002807411566534,
      "grad_norm": 0.11696089178213367,
      "learning_rate": 6.4952046542783395e-06,
      "loss": 0.3794,
      "step": 278
    },
    {
      "epoch": 2.5092644581695676,
      "grad_norm": 0.12498140576448633,
      "learning_rate": 6.2674788635540415e-06,
      "loss": 0.3905,
      "step": 279
    },
    {
      "epoch": 2.5182481751824817,
      "grad_norm": 0.11214127166395674,
      "learning_rate": 6.04347700853781e-06,
      "loss": 0.4005,
      "step": 280
    },
    {
      "epoch": 2.527231892195396,
      "grad_norm": 0.10515201118000413,
      "learning_rate": 5.823223818140458e-06,
      "loss": 0.3826,
      "step": 281
    },
    {
      "epoch": 2.53621560920831,
      "grad_norm": 0.10659878155108474,
      "learning_rate": 5.606743607435183e-06,
      "loss": 0.3862,
      "step": 282
    },
    {
      "epoch": 2.545199326221224,
      "grad_norm": 0.12414903554961848,
      "learning_rate": 5.394060274973267e-06,
      "loss": 0.4022,
      "step": 283
    },
    {
      "epoch": 2.554183043234138,
      "grad_norm": 0.1086137186577079,
      "learning_rate": 5.185197300145817e-06,
      "loss": 0.3772,
      "step": 284
    },
    {
      "epoch": 2.5631667602470523,
      "grad_norm": 0.11392438691496676,
      "learning_rate": 4.980177740591678e-06,
      "loss": 0.4074,
      "step": 285
    },
    {
      "epoch": 2.5721504772599664,
      "grad_norm": 0.10614011676956907,
      "learning_rate": 4.779024229652005e-06,
      "loss": 0.3946,
      "step": 286
    },
    {
      "epoch": 2.5811341942728805,
      "grad_norm": 0.10589648874830734,
      "learning_rate": 4.581758973871609e-06,
      "loss": 0.3948,
      "step": 287
    },
    {
      "epoch": 2.5901179112857946,
      "grad_norm": 0.10117190574073207,
      "learning_rate": 4.3884037505474455e-06,
      "loss": 0.3834,
      "step": 288
    },
    {
      "epoch": 2.5991016282987087,
      "grad_norm": 0.10088190358164044,
      "learning_rate": 4.198979905324496e-06,
      "loss": 0.3839,
      "step": 289
    },
    {
      "epoch": 2.608085345311623,
      "grad_norm": 0.10998415397637723,
      "learning_rate": 4.0135083498392905e-06,
      "loss": 0.4005,
      "step": 290
    },
    {
      "epoch": 2.6170690623245365,
      "grad_norm": 0.10291443940485356,
      "learning_rate": 3.832009559411338e-06,
      "loss": 0.3946,
      "step": 291
    },
    {
      "epoch": 2.626052779337451,
      "grad_norm": 0.0987497217572695,
      "learning_rate": 3.654503570782755e-06,
      "loss": 0.3797,
      "step": 292
    },
    {
      "epoch": 2.6350364963503647,
      "grad_norm": 0.10073485737297722,
      "learning_rate": 3.481009979906258e-06,
      "loss": 0.3977,
      "step": 293
    },
    {
      "epoch": 2.6440202133632793,
      "grad_norm": 0.10086816317959164,
      "learning_rate": 3.311547939781887e-06,
      "loss": 0.3944,
      "step": 294
    },
    {
      "epoch": 2.653003930376193,
      "grad_norm": 0.09803317553764752,
      "learning_rate": 3.14613615834253e-06,
      "loss": 0.3851,
      "step": 295
    },
    {
      "epoch": 2.661987647389107,
      "grad_norm": 0.10547156290892046,
      "learning_rate": 2.9847928963887198e-06,
      "loss": 0.3955,
      "step": 296
    },
    {
      "epoch": 2.670971364402021,
      "grad_norm": 0.09842701976978478,
      "learning_rate": 2.8275359655726586e-06,
      "loss": 0.3918,
      "step": 297
    },
    {
      "epoch": 2.6799550814149353,
      "grad_norm": 0.10279243698030634,
      "learning_rate": 2.6743827264319012e-06,
      "loss": 0.3916,
      "step": 298
    },
    {
      "epoch": 2.6889387984278494,
      "grad_norm": 0.09391104128159572,
      "learning_rate": 2.5253500864728155e-06,
      "loss": 0.3846,
      "step": 299
    },
    {
      "epoch": 2.6979225154407636,
      "grad_norm": 0.0914654057659311,
      "learning_rate": 2.3804544983040724e-06,
      "loss": 0.3824,
      "step": 300
    },
    {
      "epoch": 2.7069062324536777,
      "grad_norm": 0.09870686168033262,
      "learning_rate": 2.23971195782033e-06,
      "loss": 0.3927,
      "step": 301
    },
    {
      "epoch": 2.715889949466592,
      "grad_norm": 0.1061384997031586,
      "learning_rate": 2.1031380024363645e-06,
      "loss": 0.3993,
      "step": 302
    },
    {
      "epoch": 2.724873666479506,
      "grad_norm": 0.10190369293256683,
      "learning_rate": 1.9707477093717786e-06,
      "loss": 0.386,
      "step": 303
    },
    {
      "epoch": 2.73385738349242,
      "grad_norm": 0.09359682387036863,
      "learning_rate": 1.8425556939865696e-06,
      "loss": 0.3816,
      "step": 304
    },
    {
      "epoch": 2.742841100505334,
      "grad_norm": 0.0951294949828999,
      "learning_rate": 1.7185761081676222e-06,
      "loss": 0.3914,
      "step": 305
    },
    {
      "epoch": 2.7518248175182483,
      "grad_norm": 0.09488726435556502,
      "learning_rate": 1.5988226387664151e-06,
      "loss": 0.3826,
      "step": 306
    },
    {
      "epoch": 2.7608085345311624,
      "grad_norm": 0.08946031352013306,
      "learning_rate": 1.4833085060880349e-06,
      "loss": 0.3933,
      "step": 307
    },
    {
      "epoch": 2.7697922515440765,
      "grad_norm": 0.09936760457984616,
      "learning_rate": 1.3720464624317108e-06,
      "loss": 0.3754,
      "step": 308
    },
    {
      "epoch": 2.7787759685569906,
      "grad_norm": 0.09605033066883667,
      "learning_rate": 1.2650487906830234e-06,
      "loss": 0.4084,
      "step": 309
    },
    {
      "epoch": 2.7877596855699043,
      "grad_norm": 0.099913289552659,
      "learning_rate": 1.1623273029579195e-06,
      "loss": 0.3757,
      "step": 310
    },
    {
      "epoch": 2.796743402582819,
      "grad_norm": 0.09596634894133811,
      "learning_rate": 1.063893339298674e-06,
      "loss": 0.3957,
      "step": 311
    },
    {
      "epoch": 2.8057271195957325,
      "grad_norm": 0.09539713481428638,
      "learning_rate": 9.697577664220303e-07,
      "loss": 0.3943,
      "step": 312
    },
    {
      "epoch": 2.814710836608647,
      "grad_norm": 0.09238341661295082,
      "learning_rate": 8.799309765195452e-07,
      "loss": 0.3865,
      "step": 313
    },
    {
      "epoch": 2.8236945536215607,
      "grad_norm": 0.09215370347773162,
      "learning_rate": 7.944228861103264e-07,
      "loss": 0.393,
      "step": 314
    },
    {
      "epoch": 2.832678270634475,
      "grad_norm": 0.09454052595038595,
      "learning_rate": 7.132429349463011e-07,
      "loss": 0.387,
      "step": 315
    },
    {
      "epoch": 2.841661987647389,
      "grad_norm": 0.09250173345082788,
      "learning_rate": 6.364000849700791e-07,
      "loss": 0.391,
      "step": 316
    },
    {
      "epoch": 2.850645704660303,
      "grad_norm": 0.08839044291030243,
      "learning_rate": 5.639028193256257e-07,
      "loss": 0.3974,
      "step": 317
    },
    {
      "epoch": 2.859629421673217,
      "grad_norm": 0.08678885222396349,
      "learning_rate": 4.957591414217344e-07,
      "loss": 0.3912,
      "step": 318
    },
    {
      "epoch": 2.8686131386861313,
      "grad_norm": 0.09688413018501994,
      "learning_rate": 4.3197657404848935e-07,
      "loss": 0.3871,
      "step": 319
    },
    {
      "epoch": 2.8775968556990454,
      "grad_norm": 0.09442855901495942,
      "learning_rate": 3.725621585467698e-07,
      "loss": 0.3955,
      "step": 320
    },
    {
      "epoch": 2.8865805727119596,
      "grad_norm": 0.08776447719254343,
      "learning_rate": 3.1752245403092963e-07,
      "loss": 0.3889,
      "step": 321
    },
    {
      "epoch": 2.8955642897248737,
      "grad_norm": 0.08591645042050655,
      "learning_rate": 2.6686353666468323e-07,
      "loss": 0.3876,
      "step": 322
    },
    {
      "epoch": 2.904548006737788,
      "grad_norm": 0.08956097003735121,
      "learning_rate": 2.2059099899033098e-07,
      "loss": 0.3928,
      "step": 323
    },
    {
      "epoch": 2.913531723750702,
      "grad_norm": 0.08874803315016298,
      "learning_rate": 1.7870994931135977e-07,
      "loss": 0.3856,
      "step": 324
    },
    {
      "epoch": 2.922515440763616,
      "grad_norm": 0.09325131199811001,
      "learning_rate": 1.412250111285074e-07,
      "loss": 0.3976,
      "step": 325
    },
    {
      "epoch": 2.93149915777653,
      "grad_norm": 0.08572072856798726,
      "learning_rate": 1.0814032262935315e-07,
      "loss": 0.3809,
      "step": 326
    },
    {
      "epoch": 2.9404828747894443,
      "grad_norm": 0.08771028382043292,
      "learning_rate": 7.945953623146096e-08,
      "loss": 0.3791,
      "step": 327
    },
    {
      "epoch": 2.9494665918023584,
      "grad_norm": 0.08835820422720063,
      "learning_rate": 5.518581817918645e-08,
      "loss": 0.3928,
      "step": 328
    },
    {
      "epoch": 2.958450308815272,
      "grad_norm": 0.08971780481475433,
      "learning_rate": 3.532184819412532e-08,
      "loss": 0.3875,
      "step": 329
    },
    {
      "epoch": 2.9674340258281866,
      "grad_norm": 0.09086168349256409,
      "learning_rate": 1.9869819179292315e-08,
      "loss": 0.3902,
      "step": 330
    },
    {
      "epoch": 2.9764177428411003,
      "grad_norm": 0.09110615713231646,
      "learning_rate": 8.83143697702149e-09,
      "loss": 0.3928,
      "step": 331
    },
    {
      "epoch": 2.985401459854015,
      "grad_norm": 0.08717148170316144,
      "learning_rate": 2.2079201806501916e-09,
      "loss": 0.3867,
      "step": 332
    },
    {
      "epoch": 2.9943851768669285,
      "grad_norm": 0.09194318654950019,
      "learning_rate": 0.0,
      "loss": 0.3902,
      "step": 333
    },
    {
      "epoch": 2.9943851768669285,
      "step": 333,
      "total_flos": 4607052149424128.0,
      "train_loss": 0.4566754535869793,
      "train_runtime": 115300.1106,
      "train_samples_per_second": 2.965,
      "train_steps_per_second": 0.003
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 333,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4607052149424128.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}