{
  "best_global_step": 9000,
  "best_metric": 0.5520676374435425,
  "best_model_checkpoint": "/home/work/.workspace/exaone_RL_test/outputs/checkpoint-9000",
  "epoch": 2.0,
  "eval_steps": 1000,
  "global_step": 9766,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0020479213598197828,
      "grad_norm": 154.0,
      "learning_rate": 3.071672354948806e-07,
      "loss": 5.2426,
      "step": 10
    },
    {
      "epoch": 0.0040958427196395655,
      "grad_norm": 134.0,
      "learning_rate": 6.484641638225256e-07,
      "loss": 5.0177,
      "step": 20
    },
    {
      "epoch": 0.006143764079459349,
      "grad_norm": 101.0,
      "learning_rate": 9.897610921501708e-07,
      "loss": 4.6525,
      "step": 30
    },
    {
      "epoch": 0.008191685439279131,
      "grad_norm": 64.0,
      "learning_rate": 1.331058020477816e-06,
      "loss": 3.8346,
      "step": 40
    },
    {
      "epoch": 0.010239606799098914,
      "grad_norm": 44.0,
      "learning_rate": 1.6723549488054607e-06,
      "loss": 2.898,
      "step": 50
    },
    {
      "epoch": 0.012287528158918698,
      "grad_norm": 35.75,
      "learning_rate": 2.013651877133106e-06,
      "loss": 1.9498,
      "step": 60
    },
    {
      "epoch": 0.014335449518738481,
      "grad_norm": 29.375,
      "learning_rate": 2.354948805460751e-06,
      "loss": 1.3878,
      "step": 70
    },
    {
      "epoch": 0.016383370878558262,
      "grad_norm": 10.6875,
      "learning_rate": 2.696245733788396e-06,
      "loss": 1.1501,
      "step": 80
    },
    {
      "epoch": 0.018431292238378045,
      "grad_norm": 9.1875,
      "learning_rate": 3.0375426621160415e-06,
      "loss": 1.0169,
      "step": 90
    },
    {
      "epoch": 0.020479213598197828,
      "grad_norm": 9.6875,
      "learning_rate": 3.378839590443686e-06,
      "loss": 0.8679,
      "step": 100
    },
    {
      "epoch": 0.02252713495801761,
      "grad_norm": 7.59375,
      "learning_rate": 3.7201365187713314e-06,
      "loss": 0.8122,
      "step": 110
    },
    {
      "epoch": 0.024575056317837397,
      "grad_norm": 8.25,
      "learning_rate": 4.061433447098976e-06,
      "loss": 0.7636,
      "step": 120
    },
    {
      "epoch": 0.02662297767765718,
      "grad_norm": 7.9375,
      "learning_rate": 4.402730375426622e-06,
      "loss": 0.7755,
      "step": 130
    },
    {
      "epoch": 0.028670899037476962,
      "grad_norm": 7.90625,
      "learning_rate": 4.744027303754267e-06,
      "loss": 0.749,
      "step": 140
    },
    {
      "epoch": 0.030718820397296745,
      "grad_norm": 8.9375,
      "learning_rate": 5.0853242320819115e-06,
      "loss": 0.7375,
      "step": 150
    },
    {
      "epoch": 0.032766741757116524,
      "grad_norm": 7.25,
      "learning_rate": 5.426621160409556e-06,
      "loss": 0.7443,
      "step": 160
    },
    {
      "epoch": 0.03481466311693631,
      "grad_norm": 7.53125,
      "learning_rate": 5.767918088737202e-06,
      "loss": 0.7866,
      "step": 170
    },
    {
      "epoch": 0.03686258447675609,
      "grad_norm": 6.5,
      "learning_rate": 6.109215017064847e-06,
      "loss": 0.719,
      "step": 180
    },
    {
      "epoch": 0.038910505836575876,
      "grad_norm": 6.125,
      "learning_rate": 6.450511945392492e-06,
      "loss": 0.6936,
      "step": 190
    },
    {
      "epoch": 0.040958427196395655,
      "grad_norm": 5.75,
      "learning_rate": 6.7918088737201375e-06,
      "loss": 0.7347,
      "step": 200
    },
    {
      "epoch": 0.04300634855621544,
      "grad_norm": 5.96875,
      "learning_rate": 7.133105802047782e-06,
      "loss": 0.6707,
      "step": 210
    },
    {
      "epoch": 0.04505426991603522,
      "grad_norm": 7.9375,
      "learning_rate": 7.474402730375427e-06,
      "loss": 0.6804,
      "step": 220
    },
    {
      "epoch": 0.04710219127585501,
      "grad_norm": 5.71875,
      "learning_rate": 7.815699658703072e-06,
      "loss": 0.6831,
      "step": 230
    },
    {
      "epoch": 0.04915011263567479,
      "grad_norm": 5.40625,
      "learning_rate": 8.156996587030718e-06,
      "loss": 0.634,
      "step": 240
    },
    {
      "epoch": 0.05119803399549457,
      "grad_norm": 5.5625,
      "learning_rate": 8.498293515358363e-06,
      "loss": 0.6739,
      "step": 250
    },
    {
      "epoch": 0.05324595535531436,
      "grad_norm": 4.4375,
      "learning_rate": 8.839590443686009e-06,
      "loss": 0.65,
      "step": 260
    },
    {
      "epoch": 0.05529387671513414,
      "grad_norm": 5.5625,
      "learning_rate": 9.180887372013653e-06,
      "loss": 0.6803,
      "step": 270
    },
    {
      "epoch": 0.057341798074953924,
      "grad_norm": 5.0625,
      "learning_rate": 9.522184300341298e-06,
      "loss": 0.7035,
      "step": 280
    },
    {
      "epoch": 0.059389719434773704,
      "grad_norm": 5.875,
      "learning_rate": 9.863481228668942e-06,
      "loss": 0.7119,
      "step": 290
    },
    {
      "epoch": 0.06143764079459349,
      "grad_norm": 6.4375,
      "learning_rate": 9.999990101553574e-06,
      "loss": 0.7058,
      "step": 300
    },
    {
      "epoch": 0.06348556215441327,
      "grad_norm": 5.34375,
      "learning_rate": 9.999929611189557e-06,
      "loss": 0.6344,
      "step": 310
    },
    {
      "epoch": 0.06553348351423305,
      "grad_norm": 3.6875,
      "learning_rate": 9.999814130262912e-06,
      "loss": 0.6832,
      "step": 320
    },
    {
      "epoch": 0.06758140487405284,
      "grad_norm": 4.375,
      "learning_rate": 9.999643660043727e-06,
      "loss": 0.6186,
      "step": 330
    },
    {
      "epoch": 0.06962932623387262,
      "grad_norm": 4.90625,
      "learning_rate": 9.999418202406882e-06,
      "loss": 0.7257,
      "step": 340
    },
    {
      "epoch": 0.0716772475936924,
      "grad_norm": 5.0,
      "learning_rate": 9.999137759832017e-06,
      "loss": 0.6864,
      "step": 350
    },
    {
      "epoch": 0.07372516895351218,
      "grad_norm": 4.53125,
      "learning_rate": 9.998802335403514e-06,
      "loss": 0.6495,
      "step": 360
    },
    {
      "epoch": 0.07577309031333197,
      "grad_norm": 4.875,
      "learning_rate": 9.998411932810465e-06,
      "loss": 0.6305,
      "step": 370
    },
    {
      "epoch": 0.07782101167315175,
      "grad_norm": 4.25,
      "learning_rate": 9.997966556346617e-06,
      "loss": 0.6956,
      "step": 380
    },
    {
      "epoch": 0.07986893303297153,
      "grad_norm": 4.34375,
      "learning_rate": 9.997466210910344e-06,
      "loss": 0.5943,
      "step": 390
    },
    {
      "epoch": 0.08191685439279131,
      "grad_norm": 6.71875,
      "learning_rate": 9.996910902004576e-06,
      "loss": 0.6226,
      "step": 400
    },
    {
      "epoch": 0.0839647757526111,
      "grad_norm": 4.8125,
      "learning_rate": 9.99630063573675e-06,
      "loss": 0.6206,
      "step": 410
    },
    {
      "epoch": 0.08601269711243088,
      "grad_norm": 4.625,
      "learning_rate": 9.995635418818734e-06,
      "loss": 0.6269,
      "step": 420
    },
    {
      "epoch": 0.08806061847225066,
      "grad_norm": 4.5625,
      "learning_rate": 9.994915258566766e-06,
      "loss": 0.5965,
      "step": 430
    },
    {
      "epoch": 0.09010853983207044,
      "grad_norm": 4.5625,
      "learning_rate": 9.994140162901362e-06,
      "loss": 0.6204,
      "step": 440
    },
    {
      "epoch": 0.09215646119189023,
      "grad_norm": 4.40625,
      "learning_rate": 9.993310140347227e-06,
      "loss": 0.687,
      "step": 450
    },
    {
      "epoch": 0.09420438255171001,
      "grad_norm": 4.625,
      "learning_rate": 9.992425200033173e-06,
      "loss": 0.5611,
      "step": 460
    },
    {
      "epoch": 0.0962523039115298,
      "grad_norm": 4.21875,
      "learning_rate": 9.99148535169201e-06,
      "loss": 0.6094,
      "step": 470
    },
    {
      "epoch": 0.09830022527134959,
      "grad_norm": 4.90625,
      "learning_rate": 9.990490605660441e-06,
      "loss": 0.6283,
      "step": 480
    },
    {
      "epoch": 0.10034814663116937,
      "grad_norm": 4.96875,
      "learning_rate": 9.98944097287895e-06,
      "loss": 0.6276,
      "step": 490
    },
    {
      "epoch": 0.10239606799098915,
      "grad_norm": 4.65625,
      "learning_rate": 9.988336464891677e-06,
      "loss": 0.6535,
      "step": 500
    },
    {
      "epoch": 0.10444398935080892,
      "grad_norm": 5.03125,
      "learning_rate": 9.987177093846299e-06,
      "loss": 0.6377,
      "step": 510
    },
    {
      "epoch": 0.10649191071062872,
      "grad_norm": 3.71875,
      "learning_rate": 9.985962872493885e-06,
      "loss": 0.6542,
      "step": 520
    },
    {
      "epoch": 0.1085398320704485,
      "grad_norm": 4.40625,
      "learning_rate": 9.984693814188769e-06,
      "loss": 0.6088,
      "step": 530
    },
    {
      "epoch": 0.11058775343026828,
      "grad_norm": 4.34375,
      "learning_rate": 9.983369932888393e-06,
      "loss": 0.6549,
      "step": 540
    },
    {
      "epoch": 0.11263567479008806,
      "grad_norm": 3.75,
      "learning_rate": 9.981991243153156e-06,
      "loss": 0.5653,
      "step": 550
    },
    {
      "epoch": 0.11468359614990785,
      "grad_norm": 3.484375,
      "learning_rate": 9.980557760146259e-06,
      "loss": 0.6158,
      "step": 560
    },
    {
      "epoch": 0.11673151750972763,
      "grad_norm": 3.875,
      "learning_rate": 9.979069499633528e-06,
      "loss": 0.6301,
      "step": 570
    },
    {
      "epoch": 0.11877943886954741,
      "grad_norm": 3.375,
      "learning_rate": 9.977526477983252e-06,
      "loss": 0.5643,
      "step": 580
    },
    {
      "epoch": 0.12082736022936719,
      "grad_norm": 4.21875,
      "learning_rate": 9.975928712165997e-06,
      "loss": 0.6924,
      "step": 590
    },
    {
      "epoch": 0.12287528158918698,
      "grad_norm": 4.6875,
      "learning_rate": 9.974276219754416e-06,
      "loss": 0.6238,
      "step": 600
    },
    {
      "epoch": 0.12492320294900676,
      "grad_norm": 4.34375,
      "learning_rate": 9.972569018923064e-06,
      "loss": 0.6142,
      "step": 610
    },
    {
      "epoch": 0.12697112430882654,
      "grad_norm": 4.03125,
      "learning_rate": 9.97080712844819e-06,
      "loss": 0.5602,
      "step": 620
    },
    {
      "epoch": 0.12901904566864633,
      "grad_norm": 4.96875,
      "learning_rate": 9.96899056770754e-06,
      "loss": 0.5933,
      "step": 630
    },
    {
      "epoch": 0.1310669670284661,
      "grad_norm": 4.34375,
      "learning_rate": 9.967119356680131e-06,
      "loss": 0.591,
      "step": 640
    },
    {
      "epoch": 0.1331148883882859,
      "grad_norm": 4.46875,
      "learning_rate": 9.965193515946045e-06,
      "loss": 0.594,
      "step": 650
    },
    {
      "epoch": 0.13516280974810568,
      "grad_norm": 4.40625,
      "learning_rate": 9.963213066686193e-06,
      "loss": 0.6271,
      "step": 660
    },
    {
      "epoch": 0.13721073110792545,
      "grad_norm": 4.4375,
      "learning_rate": 9.961178030682083e-06,
      "loss": 0.6486,
      "step": 670
    },
    {
      "epoch": 0.13925865246774524,
      "grad_norm": 4.25,
      "learning_rate": 9.959088430315587e-06,
      "loss": 0.641,
      "step": 680
    },
    {
      "epoch": 0.14130657382756504,
      "grad_norm": 4.375,
      "learning_rate": 9.956944288568689e-06,
      "loss": 0.6058,
      "step": 690
    },
    {
      "epoch": 0.1433544951873848,
      "grad_norm": 4.3125,
      "learning_rate": 9.95474562902323e-06,
      "loss": 0.6139,
      "step": 700
    },
    {
      "epoch": 0.1454024165472046,
      "grad_norm": 4.53125,
      "learning_rate": 9.952492475860662e-06,
      "loss": 0.5889,
      "step": 710
    },
    {
      "epoch": 0.14745033790702436,
      "grad_norm": 3.796875,
      "learning_rate": 9.95018485386176e-06,
      "loss": 0.5904,
      "step": 720
    },
    {
      "epoch": 0.14949825926684415,
      "grad_norm": 4.21875,
      "learning_rate": 9.947822788406367e-06,
      "loss": 0.6628,
      "step": 730
    },
    {
      "epoch": 0.15154618062666395,
      "grad_norm": 3.59375,
      "learning_rate": 9.945406305473116e-06,
      "loss": 0.555,
      "step": 740
    },
    {
      "epoch": 0.1535941019864837,
      "grad_norm": 4.3125,
      "learning_rate": 9.942935431639128e-06,
      "loss": 0.6128,
      "step": 750
    },
    {
      "epoch": 0.1556420233463035,
      "grad_norm": 4.1875,
      "learning_rate": 9.940410194079736e-06,
      "loss": 0.5984,
      "step": 760
    },
    {
      "epoch": 0.1576899447061233,
      "grad_norm": 3.53125,
      "learning_rate": 9.937830620568176e-06,
      "loss": 0.6214,
      "step": 770
    },
    {
      "epoch": 0.15973786606594306,
      "grad_norm": 3.671875,
      "learning_rate": 9.935196739475287e-06,
      "loss": 0.5355,
      "step": 780
    },
    {
      "epoch": 0.16178578742576286,
      "grad_norm": 4.5625,
      "learning_rate": 9.932508579769203e-06,
      "loss": 0.6303,
      "step": 790
    },
    {
      "epoch": 0.16383370878558262,
      "grad_norm": 3.9375,
      "learning_rate": 9.929766171015015e-06,
      "loss": 0.5743,
      "step": 800
    },
    {
      "epoch": 0.16588163014540241,
      "grad_norm": 3.625,
      "learning_rate": 9.926969543374473e-06,
      "loss": 0.5829,
      "step": 810
    },
    {
      "epoch": 0.1679295515052222,
      "grad_norm": 4.75,
      "learning_rate": 9.924118727605633e-06,
      "loss": 0.6498,
      "step": 820
    },
    {
      "epoch": 0.16997747286504197,
      "grad_norm": 4.21875,
      "learning_rate": 9.921213755062531e-06,
      "loss": 0.6007,
      "step": 830
    },
    {
      "epoch": 0.17202539422486177,
      "grad_norm": 4.46875,
      "learning_rate": 9.918254657694831e-06,
      "loss": 0.639,
      "step": 840
    },
    {
      "epoch": 0.17407331558468156,
      "grad_norm": 3.28125,
      "learning_rate": 9.915241468047476e-06,
      "loss": 0.5943,
      "step": 850
    },
    {
      "epoch": 0.17612123694450132,
      "grad_norm": 3.65625,
      "learning_rate": 9.912174219260332e-06,
      "loss": 0.5616,
      "step": 860
    },
    {
      "epoch": 0.17816915830432112,
      "grad_norm": 4.84375,
      "learning_rate": 9.90905294506782e-06,
      "loss": 0.5864,
      "step": 870
    },
    {
      "epoch": 0.18021707966414088,
      "grad_norm": 3.796875,
      "learning_rate": 9.905877679798552e-06,
      "loss": 0.6337,
      "step": 880
    },
    {
      "epoch": 0.18226500102396068,
      "grad_norm": 4.25,
      "learning_rate": 9.90264845837494e-06,
      "loss": 0.6289,
      "step": 890
    },
    {
      "epoch": 0.18431292238378047,
      "grad_norm": 3.328125,
      "learning_rate": 9.899365316312826e-06,
      "loss": 0.5829,
      "step": 900
    },
    {
      "epoch": 0.18636084374360024,
      "grad_norm": 4.90625,
      "learning_rate": 9.896028289721085e-06,
      "loss": 0.6398,
      "step": 910
    },
    {
      "epoch": 0.18840876510342003,
      "grad_norm": 3.765625,
      "learning_rate": 9.892637415301227e-06,
      "loss": 0.5661,
      "step": 920
    },
    {
      "epoch": 0.19045668646323982,
      "grad_norm": 3.765625,
      "learning_rate": 9.889192730346994e-06,
      "loss": 0.6144,
      "step": 930
    },
    {
      "epoch": 0.1925046078230596,
      "grad_norm": 4.375,
      "learning_rate": 9.885694272743955e-06,
      "loss": 0.5456,
      "step": 940
    },
    {
      "epoch": 0.19455252918287938,
      "grad_norm": 5.53125,
      "learning_rate": 9.882142080969082e-06,
      "loss": 0.6665,
      "step": 950
    },
    {
      "epoch": 0.19660045054269917,
      "grad_norm": 3.515625,
      "learning_rate": 9.878536194090326e-06,
      "loss": 0.6107,
      "step": 960
    },
    {
      "epoch": 0.19864837190251894,
      "grad_norm": 3.359375,
      "learning_rate": 9.874876651766198e-06,
      "loss": 0.5805,
      "step": 970
    },
    {
      "epoch": 0.20069629326233873,
      "grad_norm": 3.984375,
      "learning_rate": 9.871163494245324e-06,
      "loss": 0.6071,
      "step": 980
    },
    {
      "epoch": 0.2027442146221585,
      "grad_norm": 4.25,
      "learning_rate": 9.867396762366e-06,
      "loss": 0.6215,
      "step": 990
    },
    {
      "epoch": 0.2047921359819783,
      "grad_norm": 5.1875,
      "learning_rate": 9.863576497555752e-06,
      "loss": 0.6455,
      "step": 1000
    },
    {
      "epoch": 0.2047921359819783,
      "eval_loss": 0.5988336801528931,
      "eval_runtime": 49.5583,
      "eval_samples_per_second": 82.973,
      "eval_steps_per_second": 41.487,
      "step": 1000
    },
    {
      "epoch": 0.20684005734179808,
      "grad_norm": 3.609375,
      "learning_rate": 9.859702741830873e-06,
      "loss": 0.546,
      "step": 1010
    },
    {
      "epoch": 0.20888797870161785,
      "grad_norm": 4.125,
      "learning_rate": 9.855775537795965e-06,
      "loss": 0.6586,
      "step": 1020
    },
    {
      "epoch": 0.21093590006143764,
      "grad_norm": 4.625,
      "learning_rate": 9.851794928643465e-06,
      "loss": 0.6279,
      "step": 1030
    },
    {
      "epoch": 0.21298382142125744,
      "grad_norm": 3.296875,
      "learning_rate": 9.847760958153183e-06,
      "loss": 0.5474,
      "step": 1040
    },
    {
      "epoch": 0.2150317427810772,
      "grad_norm": 3.640625,
      "learning_rate": 9.8436736706918e-06,
      "loss": 0.58,
      "step": 1050
    },
    {
      "epoch": 0.217079664140897,
      "grad_norm": 3.515625,
      "learning_rate": 9.839533111212395e-06,
      "loss": 0.6043,
      "step": 1060
    },
    {
      "epoch": 0.21912758550071676,
      "grad_norm": 4.3125,
      "learning_rate": 9.835339325253954e-06,
      "loss": 0.6341,
      "step": 1070
    },
    {
      "epoch": 0.22117550686053655,
      "grad_norm": 4.0625,
      "learning_rate": 9.831092358940853e-06,
      "loss": 0.5638,
      "step": 1080
    },
    {
      "epoch": 0.22322342822035635,
      "grad_norm": 3.3125,
      "learning_rate": 9.82679225898236e-06,
      "loss": 0.6114,
      "step": 1090
    },
    {
      "epoch": 0.2252713495801761,
      "grad_norm": 4.90625,
      "learning_rate": 9.82243907267213e-06,
      "loss": 0.6526,
      "step": 1100
    },
    {
      "epoch": 0.2273192709399959,
      "grad_norm": 4.625,
      "learning_rate": 9.818032847887664e-06,
      "loss": 0.5896,
      "step": 1110
    },
    {
      "epoch": 0.2293671922998157,
      "grad_norm": 3.453125,
      "learning_rate": 9.813573633089804e-06,
      "loss": 0.5732,
      "step": 1120
    },
    {
      "epoch": 0.23141511365963546,
      "grad_norm": 3.71875,
      "learning_rate": 9.809061477322186e-06,
      "loss": 0.6223,
      "step": 1130
    },
    {
      "epoch": 0.23346303501945526,
      "grad_norm": 4.21875,
      "learning_rate": 9.804496430210704e-06,
      "loss": 0.6012,
      "step": 1140
    },
    {
      "epoch": 0.23551095637927505,
      "grad_norm": 4.0,
      "learning_rate": 9.799878541962968e-06,
      "loss": 0.6064,
      "step": 1150
    },
    {
      "epoch": 0.23755887773909481,
      "grad_norm": 3.828125,
      "learning_rate": 9.795207863367745e-06,
      "loss": 0.553,
      "step": 1160
    },
    {
      "epoch": 0.2396067990989146,
      "grad_norm": 4.53125,
      "learning_rate": 9.790484445794413e-06,
      "loss": 0.6526,
      "step": 1170
    },
    {
      "epoch": 0.24165472045873437,
      "grad_norm": 4.25,
      "learning_rate": 9.785708341192375e-06,
      "loss": 0.6117,
      "step": 1180
    },
    {
      "epoch": 0.24370264181855417,
      "grad_norm": 3.828125,
      "learning_rate": 9.78087960209051e-06,
      "loss": 0.5773,
      "step": 1190
    },
    {
      "epoch": 0.24575056317837396,
      "grad_norm": 4.71875,
      "learning_rate": 9.77599828159658e-06,
      "loss": 0.6004,
      "step": 1200
    },
    {
      "epoch": 0.24779848453819373,
      "grad_norm": 4.40625,
      "learning_rate": 9.771064433396652e-06,
      "loss": 0.5644,
      "step": 1210
    },
    {
      "epoch": 0.24984640589801352,
      "grad_norm": 3.75,
      "learning_rate": 9.76607811175451e-06,
      "loss": 0.5899,
      "step": 1220
    },
    {
      "epoch": 0.2518943272578333,
      "grad_norm": 4.78125,
      "learning_rate": 9.761039371511051e-06,
      "loss": 0.5757,
      "step": 1230
    },
    {
      "epoch": 0.2539422486176531,
      "grad_norm": 3.6875,
      "learning_rate": 9.755948268083689e-06,
      "loss": 0.5837,
      "step": 1240
    },
    {
      "epoch": 0.25599016997747287,
      "grad_norm": 4.21875,
      "learning_rate": 9.750804857465736e-06,
      "loss": 0.5767,
      "step": 1250
    },
    {
      "epoch": 0.25803809133729266,
      "grad_norm": 3.0625,
      "learning_rate": 9.745609196225804e-06,
      "loss": 0.6256,
      "step": 1260
    },
    {
      "epoch": 0.26008601269711246,
      "grad_norm": 3.9375,
      "learning_rate": 9.740361341507158e-06,
      "loss": 0.5644,
      "step": 1270
    },
    {
      "epoch": 0.2621339340569322,
      "grad_norm": 3.859375,
      "learning_rate": 9.735061351027114e-06,
      "loss": 0.6039,
      "step": 1280
    },
    {
      "epoch": 0.264181855416752,
      "grad_norm": 3.75,
      "learning_rate": 9.729709283076375e-06,
      "loss": 0.5902,
      "step": 1290
    },
    {
      "epoch": 0.2662297767765718,
      "grad_norm": 3.921875,
      "learning_rate": 9.724305196518425e-06,
      "loss": 0.5738,
      "step": 1300
    },
    {
      "epoch": 0.2682776981363916,
      "grad_norm": 3.75,
      "learning_rate": 9.718849150788848e-06,
      "loss": 0.5882,
      "step": 1310
    },
    {
      "epoch": 0.27032561949621137,
      "grad_norm": 3.65625,
      "learning_rate": 9.713341205894691e-06,
      "loss": 0.5549,
      "step": 1320
    },
    {
      "epoch": 0.2723735408560311,
      "grad_norm": 3.578125,
      "learning_rate": 9.707781422413811e-06,
      "loss": 0.5866,
      "step": 1330
    },
    {
      "epoch": 0.2744214622158509,
      "grad_norm": 3.671875,
      "learning_rate": 9.702169861494189e-06,
      "loss": 0.6095,
      "step": 1340
    },
    {
      "epoch": 0.2764693835756707,
      "grad_norm": 3.890625,
      "learning_rate": 9.696506584853271e-06,
      "loss": 0.5493,
      "step": 1350
    },
    {
      "epoch": 0.2785173049354905,
      "grad_norm": 4.125,
      "learning_rate": 9.690791654777293e-06,
      "loss": 0.6172,
      "step": 1360
    },
    {
      "epoch": 0.2805652262953103,
      "grad_norm": 3.046875,
      "learning_rate": 9.685025134120577e-06,
      "loss": 0.5721,
      "step": 1370
    },
    {
      "epoch": 0.28261314765513007,
      "grad_norm": 4.84375,
      "learning_rate": 9.679207086304864e-06,
      "loss": 0.5712,
      "step": 1380
    },
    {
      "epoch": 0.2846610690149498,
      "grad_norm": 3.890625,
      "learning_rate": 9.673337575318597e-06,
      "loss": 0.6429,
      "step": 1390
    },
    {
      "epoch": 0.2867089903747696,
      "grad_norm": 4.3125,
      "learning_rate": 9.66741666571623e-06,
      "loss": 0.6728,
      "step": 1400
    },
    {
      "epoch": 0.2887569117345894,
      "grad_norm": 4.375,
      "learning_rate": 9.661444422617505e-06,
      "loss": 0.6315,
      "step": 1410
    },
    {
      "epoch": 0.2908048330944092,
      "grad_norm": 4.3125,
      "learning_rate": 9.655420911706752e-06,
      "loss": 0.5544,
      "step": 1420
    },
    {
      "epoch": 0.292852754454229,
      "grad_norm": 3.796875,
      "learning_rate": 9.649346199232154e-06,
      "loss": 0.5458,
      "step": 1430
    },
    {
      "epoch": 0.2949006758140487,
      "grad_norm": 4.28125,
      "learning_rate": 9.643220352005023e-06,
      "loss": 0.5573,
      "step": 1440
    },
    {
      "epoch": 0.2969485971738685,
      "grad_norm": 3.75,
      "learning_rate": 9.637043437399063e-06,
      "loss": 0.6293,
      "step": 1450
    },
    {
      "epoch": 0.2989965185336883,
      "grad_norm": 3.296875,
      "learning_rate": 9.630815523349636e-06,
      "loss": 0.5829,
      "step": 1460
    },
    {
      "epoch": 0.3010444398935081,
      "grad_norm": 3.46875,
      "learning_rate": 9.624536678352998e-06,
      "loss": 0.5512,
      "step": 1470
    },
    {
      "epoch": 0.3030923612533279,
      "grad_norm": 3.65625,
      "learning_rate": 9.618206971465574e-06,
      "loss": 0.5756,
      "step": 1480
    },
    {
      "epoch": 0.30514028261314763,
      "grad_norm": 3.5625,
      "learning_rate": 9.611826472303165e-06,
      "loss": 0.5815,
      "step": 1490
    },
    {
      "epoch": 0.3071882039729674,
      "grad_norm": 3.515625,
      "learning_rate": 9.605395251040212e-06,
      "loss": 0.575,
      "step": 1500
    },
    {
      "epoch": 0.3092361253327872,
      "grad_norm": 3.484375,
      "learning_rate": 9.598913378409006e-06,
      "loss": 0.661,
      "step": 1510
    },
    {
      "epoch": 0.311284046692607,
      "grad_norm": 3.21875,
      "learning_rate": 9.592380925698911e-06,
      "loss": 0.5835,
      "step": 1520
    },
    {
      "epoch": 0.3133319680524268,
      "grad_norm": 4.53125,
      "learning_rate": 9.585797964755593e-06,
      "loss": 0.5867,
      "step": 1530
    },
    {
      "epoch": 0.3153798894122466,
      "grad_norm": 3.875,
      "learning_rate": 9.579164567980212e-06,
      "loss": 0.5765,
      "step": 1540
    },
    {
      "epoch": 0.31742781077206633,
      "grad_norm": 3.328125,
      "learning_rate": 9.572480808328643e-06,
      "loss": 0.5486,
      "step": 1550
    },
    {
      "epoch": 0.3194757321318861,
      "grad_norm": 3.65625,
      "learning_rate": 9.565746759310659e-06,
      "loss": 0.6179,
      "step": 1560
    },
    {
      "epoch": 0.3215236534917059,
      "grad_norm": 4.28125,
      "learning_rate": 9.558962494989133e-06,
      "loss": 0.6122,
      "step": 1570
    },
    {
      "epoch": 0.3235715748515257,
      "grad_norm": 4.0625,
      "learning_rate": 9.552128089979215e-06,
      "loss": 0.5539,
      "step": 1580
    },
    {
      "epoch": 0.3256194962113455,
      "grad_norm": 3.84375,
      "learning_rate": 9.545243619447524e-06,
      "loss": 0.5897,
      "step": 1590
    },
    {
      "epoch": 0.32766741757116524,
      "grad_norm": 4.15625,
      "learning_rate": 9.538309159111303e-06,
      "loss": 0.5231,
      "step": 1600
    },
    {
      "epoch": 0.32971533893098504,
      "grad_norm": 4.15625,
      "learning_rate": 9.531324785237603e-06,
      "loss": 0.6409,
      "step": 1610
    },
    {
      "epoch": 0.33176326029080483,
      "grad_norm": 3.109375,
      "learning_rate": 9.524290574642432e-06,
      "loss": 0.5647,
      "step": 1620
    },
    {
      "epoch": 0.3338111816506246,
      "grad_norm": 3.3125,
      "learning_rate": 9.517206604689924e-06,
      "loss": 0.6115,
      "step": 1630
    },
    {
      "epoch": 0.3358591030104444,
      "grad_norm": 4.21875,
      "learning_rate": 9.51007295329147e-06,
      "loss": 0.5784,
      "step": 1640
    },
    {
      "epoch": 0.3379070243702642,
      "grad_norm": 3.125,
      "learning_rate": 9.502889698904877e-06,
      "loss": 0.5442,
      "step": 1650
    },
    {
      "epoch": 0.33995494573008395,
      "grad_norm": 3.8125,
      "learning_rate": 9.495656920533499e-06,
      "loss": 0.5771,
      "step": 1660
    },
    {
      "epoch": 0.34200286708990374,
      "grad_norm": 3.78125,
      "learning_rate": 9.488374697725361e-06,
      "loss": 0.5675,
      "step": 1670
    },
    {
      "epoch": 0.34405078844972353,
      "grad_norm": 4.1875,
      "learning_rate": 9.481043110572301e-06,
      "loss": 0.6298,
      "step": 1680
    },
    {
      "epoch": 0.3460987098095433,
      "grad_norm": 5.0625,
      "learning_rate": 9.473662239709074e-06,
      "loss": 0.6504,
      "step": 1690
    },
    {
      "epoch": 0.3481466311693631,
      "grad_norm": 3.828125,
      "learning_rate": 9.46623216631247e-06,
      "loss": 0.606,
      "step": 1700
    },
    {
      "epoch": 0.35019455252918286,
      "grad_norm": 3.625,
      "learning_rate": 9.458752972100425e-06,
      "loss": 0.6153,
      "step": 1710
    },
    {
      "epoch": 0.35224247388900265,
      "grad_norm": 3.484375,
      "learning_rate": 9.451224739331113e-06,
      "loss": 0.5412,
      "step": 1720
    },
    {
      "epoch": 0.35429039524882244,
      "grad_norm": 3.5,
      "learning_rate": 9.443647550802052e-06,
      "loss": 0.5356,
      "step": 1730
    },
    {
      "epoch": 0.35633831660864224,
      "grad_norm": 5.1875,
      "learning_rate": 9.436021489849189e-06,
      "loss": 0.6106,
      "step": 1740
    },
    {
      "epoch": 0.35838623796846203,
      "grad_norm": 3.921875,
      "learning_rate": 9.42834664034598e-06,
      "loss": 0.5864,
      "step": 1750
    },
    {
      "epoch": 0.36043415932828177,
      "grad_norm": 4.03125,
      "learning_rate": 9.420623086702469e-06,
      "loss": 0.5616,
      "step": 1760
    },
    {
      "epoch": 0.36248208068810156,
      "grad_norm": 3.4375,
      "learning_rate": 9.41285091386437e-06,
      "loss": 0.5831,
      "step": 1770
    },
    {
      "epoch": 0.36453000204792135,
      "grad_norm": 3.640625,
      "learning_rate": 9.405030207312113e-06,
      "loss": 0.5746,
      "step": 1780
    },
    {
      "epoch": 0.36657792340774115,
      "grad_norm": 3.5,
      "learning_rate": 9.397161053059923e-06,
      "loss": 0.5429,
      "step": 1790
    },
    {
      "epoch": 0.36862584476756094,
      "grad_norm": 2.953125,
      "learning_rate": 9.389243537654857e-06,
      "loss": 0.5584,
      "step": 1800
    },
    {
      "epoch": 0.37067376612738073,
      "grad_norm": 3.90625,
      "learning_rate": 9.381277748175872e-06,
      "loss": 0.6123,
      "step": 1810
    },
    {
      "epoch": 0.37272168748720047,
      "grad_norm": 4.25,
      "learning_rate": 9.373263772232848e-06,
      "loss": 0.597,
      "step": 1820
    },
    {
      "epoch": 0.37476960884702026,
      "grad_norm": 4.0625,
      "learning_rate": 9.365201697965634e-06,
      "loss": 0.5578,
      "step": 1830
    },
    {
      "epoch": 0.37681753020684006,
      "grad_norm": 3.890625,
      "learning_rate": 9.357091614043078e-06,
      "loss": 0.6015,
      "step": 1840
    },
    {
      "epoch": 0.37886545156665985,
      "grad_norm": 3.515625,
      "learning_rate": 9.348933609662055e-06,
      "loss": 0.5694,
      "step": 1850
    },
    {
      "epoch": 0.38091337292647964,
      "grad_norm": 3.453125,
      "learning_rate": 9.340727774546476e-06,
      "loss": 0.5754,
      "step": 1860
    },
    {
      "epoch": 0.3829612942862994,
      "grad_norm": 3.609375,
      "learning_rate": 9.332474198946308e-06,
      "loss": 0.5598,
      "step": 1870
    },
    {
      "epoch": 0.3850092156461192,
      "grad_norm": 3.3125,
      "learning_rate": 9.324172973636583e-06,
      "loss": 0.5535,
      "step": 1880
    },
    {
      "epoch": 0.38705713700593897,
      "grad_norm": 3.890625,
      "learning_rate": 9.315824189916398e-06,
      "loss": 0.5996,
      "step": 1890
    },
    {
      "epoch": 0.38910505836575876,
      "grad_norm": 4.25,
      "learning_rate": 9.307427939607906e-06,
      "loss": 0.5701,
      "step": 1900
    },
    {
      "epoch": 0.39115297972557855,
      "grad_norm": 3.53125,
      "learning_rate": 9.298984315055316e-06,
      "loss": 0.5873,
      "step": 1910
    },
    {
      "epoch": 0.39320090108539835,
      "grad_norm": 3.515625,
      "learning_rate": 9.290493409123864e-06,
      "loss": 0.5693,
      "step": 1920
    },
    {
      "epoch": 0.3952488224452181,
      "grad_norm": 3.84375,
      "learning_rate": 9.281955315198805e-06,
      "loss": 0.5408,
      "step": 1930
    },
    {
      "epoch": 0.3972967438050379,
      "grad_norm": 3.203125,
      "learning_rate": 9.273370127184382e-06,
      "loss": 0.582,
      "step": 1940
    },
    {
      "epoch": 0.39934466516485767,
      "grad_norm": 3.28125,
      "learning_rate": 9.264737939502784e-06,
      "loss": 0.5886,
      "step": 1950
    },
    {
      "epoch": 0.40139258652467746,
      "grad_norm": 4.34375,
      "learning_rate": 9.256058847093122e-06,
      "loss": 0.5457,
      "step": 1960
    },
    {
      "epoch": 0.40344050788449726,
      "grad_norm": 4.0,
      "learning_rate": 9.24733294541037e-06,
      "loss": 0.5653,
      "step": 1970
    },
    {
      "epoch": 0.405488429244317,
      "grad_norm": 4.03125,
      "learning_rate": 9.238560330424333e-06,
      "loss": 0.6059,
      "step": 1980
    },
    {
      "epoch": 0.4075363506041368,
      "grad_norm": 4.1875,
      "learning_rate": 9.229741098618573e-06,
      "loss": 0.5621,
      "step": 1990
    },
    {
      "epoch": 0.4095842719639566,
      "grad_norm": 3.578125,
      "learning_rate": 9.220875346989363e-06,
      "loss": 0.5432,
      "step": 2000
    },
    {
      "epoch": 0.4095842719639566,
      "eval_loss": 0.5784266591072083,
      "eval_runtime": 49.5386,
      "eval_samples_per_second": 83.006,
      "eval_steps_per_second": 41.503,
      "step": 2000
    },
    {
      "epoch": 0.4116321933237764,
      "grad_norm": 3.09375,
      "learning_rate": 9.211963173044604e-06,
      "loss": 0.5743,
      "step": 2010
    },
    {
      "epoch": 0.41368011468359617,
      "grad_norm": 4.1875,
      "learning_rate": 9.203004674802774e-06,
      "loss": 0.613,
      "step": 2020
    },
    {
      "epoch": 0.41572803604341596,
      "grad_norm": 3.4375,
      "learning_rate": 9.193999950791825e-06,
      "loss": 0.5318,
      "step": 2030
    },
    {
      "epoch": 0.4177759574032357,
      "grad_norm": 3.1875,
      "learning_rate": 9.184949100048123e-06,
      "loss": 0.5718,
      "step": 2040
    },
    {
      "epoch": 0.4198238787630555,
      "grad_norm": 3.578125,
      "learning_rate": 9.175852222115346e-06,
      "loss": 0.5876,
      "step": 2050
    },
    {
      "epoch": 0.4218718001228753,
      "grad_norm": 4.15625,
      "learning_rate": 9.166709417043386e-06,
      "loss": 0.5436,
      "step": 2060
    },
    {
      "epoch": 0.4239197214826951,
      "grad_norm": 4.21875,
      "learning_rate": 9.15752078538726e-06,
      "loss": 0.5473,
      "step": 2070
    },
    {
      "epoch": 0.42596764284251487,
      "grad_norm": 3.5,
      "learning_rate": 9.148286428205994e-06,
      "loss": 0.6101,
      "step": 2080
    },
    {
      "epoch": 0.4280155642023346,
      "grad_norm": 4.625,
      "learning_rate": 9.139006447061518e-06,
      "loss": 0.5197,
      "step": 2090
    },
    {
      "epoch": 0.4300634855621544,
      "grad_norm": 4.25,
      "learning_rate": 9.129680944017544e-06,
      "loss": 0.5946,
      "step": 2100
    },
    {
      "epoch": 0.4321114069219742,
      "grad_norm": 4.5,
      "learning_rate": 9.120310021638451e-06,
      "loss": 0.5866,
      "step": 2110
    },
    {
      "epoch": 0.434159328281794,
      "grad_norm": 3.65625,
      "learning_rate": 9.110893782988148e-06,
      "loss": 0.5906,
      "step": 2120
    },
    {
      "epoch": 0.4362072496416138,
      "grad_norm": 3.28125,
      "learning_rate": 9.101432331628943e-06,
      "loss": 0.5084,
      "step": 2130
    },
    {
      "epoch": 0.4382551710014335,
      "grad_norm": 3.875,
      "learning_rate": 9.09192577162041e-06,
      "loss": 0.5464,
      "step": 2140
    },
    {
      "epoch": 0.4403030923612533,
      "grad_norm": 4.0,
      "learning_rate": 9.082374207518234e-06,
      "loss": 0.5649,
      "step": 2150
    },
    {
      "epoch": 0.4423510137210731,
      "grad_norm": 3.515625,
      "learning_rate": 9.072777744373073e-06,
      "loss": 0.5508,
      "step": 2160
    },
    {
      "epoch": 0.4443989350808929,
      "grad_norm": 3.265625,
      "learning_rate": 9.063136487729397e-06,
      "loss": 0.5607,
      "step": 2170
    },
    {
      "epoch": 0.4464468564407127,
      "grad_norm": 3.546875,
      "learning_rate": 9.05345054362432e-06,
      "loss": 0.5629,
      "step": 2180
    },
    {
      "epoch": 0.4484947778005325,
      "grad_norm": 4.09375,
      "learning_rate": 9.043720018586447e-06,
      "loss": 0.574,
      "step": 2190
    },
    {
      "epoch": 0.4505426991603522,
      "grad_norm": 4.125,
      "learning_rate": 9.033945019634693e-06,
      "loss": 0.5668,
      "step": 2200
    },
    {
      "epoch": 0.452590620520172,
      "grad_norm": 3.609375,
      "learning_rate": 9.024125654277111e-06,
      "loss": 0.5291,
      "step": 2210
    },
    {
      "epoch": 0.4546385418799918,
      "grad_norm": 4.1875,
      "learning_rate": 9.014262030509704e-06,
      "loss": 0.6214,
      "step": 2220
    },
    {
      "epoch": 0.4566864632398116,
      "grad_norm": 4.125,
      "learning_rate": 9.004354256815247e-06,
      "loss": 0.6003,
      "step": 2230
    },
    {
      "epoch": 0.4587343845996314,
      "grad_norm": 3.375,
      "learning_rate": 8.994402442162083e-06,
      "loss": 0.5938,
      "step": 2240
    },
    {
      "epoch": 0.46078230595945113,
      "grad_norm": 3.3125,
      "learning_rate": 8.984406696002925e-06,
      "loss": 0.5887,
      "step": 2250
    },
    {
      "epoch": 0.4628302273192709,
      "grad_norm": 3.84375,
      "learning_rate": 8.97436712827367e-06,
      "loss": 0.6317,
      "step": 2260
    },
    {
      "epoch": 0.4648781486790907,
      "grad_norm": 3.59375,
      "learning_rate": 8.964283849392163e-06,
      "loss": 0.5547,
      "step": 2270
    },
    {
      "epoch": 0.4669260700389105,
      "grad_norm": 3.515625,
      "learning_rate": 8.954156970257001e-06,
      "loss": 0.6202,
      "step": 2280
    },
    {
      "epoch": 0.4689739913987303,
      "grad_norm": 3.9375,
      "learning_rate": 8.94398660224631e-06,
      "loss": 0.6002,
      "step": 2290
    },
    {
      "epoch": 0.4710219127585501,
      "grad_norm": 3.9375,
      "learning_rate": 8.933772857216517e-06,
      "loss": 0.5653,
      "step": 2300
    },
    {
      "epoch": 0.47306983411836984,
      "grad_norm": 3.203125,
      "learning_rate": 8.923515847501117e-06,
      "loss": 0.624,
      "step": 2310
    },
    {
      "epoch": 0.47511775547818963,
      "grad_norm": 3.5,
      "learning_rate": 8.913215685909448e-06,
      "loss": 0.5667,
      "step": 2320
    },
    {
      "epoch": 0.4771656768380094,
      "grad_norm": 3.8125,
      "learning_rate": 8.90287248572544e-06,
      "loss": 0.5623,
      "step": 2330
    },
    {
      "epoch": 0.4792135981978292,
      "grad_norm": 3.640625,
      "learning_rate": 8.892486360706377e-06,
      "loss": 0.5846,
      "step": 2340
    },
    {
      "epoch": 0.481261519557649,
      "grad_norm": 4.25,
      "learning_rate": 8.882057425081635e-06,
      "loss": 0.5705,
      "step": 2350
    },
    {
      "epoch": 0.48330944091746875,
      "grad_norm": 3.96875,
      "learning_rate": 8.871585793551435e-06,
      "loss": 0.5893,
      "step": 2360
    },
    {
      "epoch": 0.48535736227728854,
      "grad_norm": 4.84375,
      "learning_rate": 8.861071581285584e-06,
      "loss": 0.6106,
      "step": 2370
    },
    {
      "epoch": 0.48740528363710833,
      "grad_norm": 3.71875,
      "learning_rate": 8.850514903922198e-06,
      "loss": 0.5863,
      "step": 2380
    },
    {
      "epoch": 0.4894532049969281,
      "grad_norm": 3.4375,
      "learning_rate": 8.839915877566438e-06,
      "loss": 0.5634,
      "step": 2390
    },
    {
      "epoch": 0.4915011263567479,
      "grad_norm": 3.765625,
      "learning_rate": 8.829274618789228e-06,
      "loss": 0.5406,
      "step": 2400
    },
    {
      "epoch": 0.49354904771656766,
      "grad_norm": 3.671875,
      "learning_rate": 8.818591244625977e-06,
      "loss": 0.5016,
      "step": 2410
    },
    {
      "epoch": 0.49559696907638745,
      "grad_norm": 3.484375,
      "learning_rate": 8.807865872575294e-06,
      "loss": 0.6582,
      "step": 2420
    },
    {
      "epoch": 0.49764489043620724,
      "grad_norm": 3.515625,
      "learning_rate": 8.797098620597683e-06,
      "loss": 0.573,
      "step": 2430
    },
    {
      "epoch": 0.49969281179602704,
      "grad_norm": 4.0625,
      "learning_rate": 8.786289607114262e-06,
      "loss": 0.6161,
      "step": 2440
    },
    {
      "epoch": 0.5017407331558468,
      "grad_norm": 3.953125,
      "learning_rate": 8.775438951005453e-06,
      "loss": 0.5801,
      "step": 2450
    },
    {
      "epoch": 0.5037886545156666,
      "grad_norm": 3.046875,
      "learning_rate": 8.764546771609672e-06,
      "loss": 0.5537,
      "step": 2460
    },
    {
      "epoch": 0.5058365758754864,
      "grad_norm": 4.03125,
      "learning_rate": 8.753613188722018e-06,
      "loss": 0.5411,
      "step": 2470
    },
    {
      "epoch": 0.5078844972353062,
      "grad_norm": 3.828125,
      "learning_rate": 8.74263832259296e-06,
      "loss": 0.531,
      "step": 2480
    },
    {
      "epoch": 0.509932418595126,
      "grad_norm": 3.78125,
      "learning_rate": 8.73162229392701e-06,
      "loss": 0.5652,
      "step": 2490
    },
    {
      "epoch": 0.5119803399549457,
      "grad_norm": 4.03125,
      "learning_rate": 8.7205652238814e-06,
      "loss": 0.6105,
      "step": 2500
    },
    {
      "epoch": 0.5140282613147655,
      "grad_norm": 4.125,
      "learning_rate": 8.709467234064738e-06,
      "loss": 0.5952,
      "step": 2510
    },
    {
      "epoch": 0.5160761826745853,
      "grad_norm": 4.125,
      "learning_rate": 8.698328446535683e-06,
      "loss": 0.5135,
      "step": 2520
    },
    {
      "epoch": 0.5181241040344051,
      "grad_norm": 4.28125,
      "learning_rate": 8.687148983801599e-06,
      "loss": 0.6145,
      "step": 2530
    },
    {
      "epoch": 0.5201720253942249,
      "grad_norm": 3.9375,
      "learning_rate": 8.675928968817206e-06,
      "loss": 0.5855,
      "step": 2540
    },
    {
      "epoch": 0.5222199467540446,
      "grad_norm": 3.875,
      "learning_rate": 8.664668524983232e-06,
      "loss": 0.6286,
      "step": 2550
    },
    {
      "epoch": 0.5242678681138644,
      "grad_norm": 3.0,
      "learning_rate": 8.653367776145045e-06,
      "loss": 0.5642,
      "step": 2560
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 3.0625,
      "learning_rate": 8.642026846591303e-06,
      "loss": 0.5474,
      "step": 2570
    },
    {
      "epoch": 0.528363710833504,
      "grad_norm": 3.734375,
      "learning_rate": 8.63064586105258e-06,
      "loss": 0.6154,
      "step": 2580
    },
    {
      "epoch": 0.5304116321933238,
      "grad_norm": 4.375,
      "learning_rate": 8.619224944700003e-06,
      "loss": 0.5802,
      "step": 2590
    },
    {
      "epoch": 0.5324595535531436,
      "grad_norm": 3.546875,
      "learning_rate": 8.607764223143854e-06,
      "loss": 0.5435,
      "step": 2600
    },
    {
      "epoch": 0.5345074749129634,
      "grad_norm": 3.015625,
      "learning_rate": 8.596263822432222e-06,
      "loss": 0.585,
      "step": 2610
    },
    {
      "epoch": 0.5365553962727831,
      "grad_norm": 3.171875,
      "learning_rate": 8.584723869049586e-06,
      "loss": 0.5417,
      "step": 2620
    },
    {
      "epoch": 0.5386033176326029,
      "grad_norm": 4.1875,
      "learning_rate": 8.573144489915437e-06,
      "loss": 0.6276,
      "step": 2630
    },
    {
      "epoch": 0.5406512389924227,
      "grad_norm": 3.5625,
      "learning_rate": 8.561525812382882e-06,
      "loss": 0.597,
      "step": 2640
    },
    {
      "epoch": 0.5426991603522425,
      "grad_norm": 3.265625,
      "learning_rate": 8.549867964237246e-06,
      "loss": 0.5478,
      "step": 2650
    },
    {
      "epoch": 0.5447470817120622,
      "grad_norm": 3.859375,
      "learning_rate": 8.538171073694656e-06,
      "loss": 0.5764,
      "step": 2660
    },
    {
      "epoch": 0.546795003071882,
      "grad_norm": 3.78125,
      "learning_rate": 8.526435269400644e-06,
      "loss": 0.5692,
      "step": 2670
    },
    {
      "epoch": 0.5488429244317018,
      "grad_norm": 3.359375,
      "learning_rate": 8.514660680428723e-06,
      "loss": 0.5599,
      "step": 2680
    },
    {
      "epoch": 0.5508908457915216,
      "grad_norm": 3.671875,
      "learning_rate": 8.50284743627897e-06,
      "loss": 0.6127,
      "step": 2690
    },
    {
      "epoch": 0.5529387671513414,
      "grad_norm": 4.1875,
      "learning_rate": 8.490995666876602e-06,
      "loss": 0.654,
      "step": 2700
    },
    {
      "epoch": 0.5549866885111612,
      "grad_norm": 5.1875,
      "learning_rate": 8.47910550257055e-06,
      "loss": 0.5806,
      "step": 2710
    },
    {
      "epoch": 0.557034609870981,
      "grad_norm": 3.984375,
      "learning_rate": 8.467177074132017e-06,
      "loss": 0.5577,
      "step": 2720
    },
    {
      "epoch": 0.5590825312308008,
      "grad_norm": 2.984375,
      "learning_rate": 8.455210512753047e-06,
      "loss": 0.5119,
      "step": 2730
    },
    {
      "epoch": 0.5611304525906206,
      "grad_norm": 3.59375,
      "learning_rate": 8.443205950045086e-06,
      "loss": 0.5889,
      "step": 2740
    },
    {
      "epoch": 0.5631783739504403,
      "grad_norm": 3.3125,
      "learning_rate": 8.431163518037523e-06,
      "loss": 0.5612,
      "step": 2750
    },
    {
      "epoch": 0.5652262953102601,
      "grad_norm": 3.21875,
      "learning_rate": 8.41908334917624e-06,
      "loss": 0.5988,
      "step": 2760
    },
    {
      "epoch": 0.5672742166700798,
      "grad_norm": 3.9375,
      "learning_rate": 8.40696557632217e-06,
      "loss": 0.5484,
      "step": 2770
    },
    {
      "epoch": 0.5693221380298996,
      "grad_norm": 4.34375,
      "learning_rate": 8.394810332749814e-06,
      "loss": 0.5484,
      "step": 2780
    },
    {
      "epoch": 0.5713700593897194,
      "grad_norm": 4.9375,
      "learning_rate": 8.382617752145794e-06,
      "loss": 0.5683,
      "step": 2790
    },
    {
      "epoch": 0.5734179807495392,
      "grad_norm": 5.03125,
      "learning_rate": 8.370387968607369e-06,
      "loss": 0.5616,
      "step": 2800
    },
    {
      "epoch": 0.575465902109359,
      "grad_norm": 3.515625,
      "learning_rate": 8.358121116640971e-06,
      "loss": 0.5814,
      "step": 2810
    },
    {
      "epoch": 0.5775138234691788,
      "grad_norm": 3.65625,
      "learning_rate": 8.345817331160714e-06,
      "loss": 0.5634,
      "step": 2820
    },
    {
      "epoch": 0.5795617448289986,
      "grad_norm": 3.6875,
      "learning_rate": 8.333476747486922e-06,
      "loss": 0.6005,
      "step": 2830
    },
    {
      "epoch": 0.5816096661888184,
      "grad_norm": 3.390625,
      "learning_rate": 8.321099501344634e-06,
      "loss": 0.5763,
      "step": 2840
    },
    {
      "epoch": 0.5836575875486382,
      "grad_norm": 3.46875,
      "learning_rate": 8.308685728862111e-06,
      "loss": 0.5533,
      "step": 2850
    },
    {
      "epoch": 0.585705508908458,
      "grad_norm": 4.09375,
      "learning_rate": 8.296235566569343e-06,
      "loss": 0.5745,
      "step": 2860
    },
    {
      "epoch": 0.5877534302682776,
      "grad_norm": 3.640625,
      "learning_rate": 8.283749151396542e-06,
      "loss": 0.5044,
      "step": 2870
    },
    {
      "epoch": 0.5898013516280974,
      "grad_norm": 3.453125,
      "learning_rate": 8.27122662067264e-06,
      "loss": 0.5412,
      "step": 2880
    },
    {
      "epoch": 0.5918492729879172,
      "grad_norm": 3.765625,
      "learning_rate": 8.258668112123781e-06,
      "loss": 0.5889,
      "step": 2890
    },
    {
      "epoch": 0.593897194347737,
      "grad_norm": 4.375,
      "learning_rate": 8.246073763871796e-06,
      "loss": 0.5874,
      "step": 2900
    },
    {
      "epoch": 0.5959451157075568,
      "grad_norm": 3.265625,
      "learning_rate": 8.233443714432698e-06,
      "loss": 0.5976,
      "step": 2910
    },
    {
      "epoch": 0.5979930370673766,
      "grad_norm": 4.21875,
      "learning_rate": 8.220778102715145e-06,
      "loss": 0.6134,
      "step": 2920
    },
    {
      "epoch": 0.6000409584271964,
      "grad_norm": 4.03125,
      "learning_rate": 8.20807706801892e-06,
      "loss": 0.5816,
      "step": 2930
    },
    {
      "epoch": 0.6020888797870162,
      "grad_norm": 3.578125,
      "learning_rate": 8.195340750033401e-06,
      "loss": 0.5833,
      "step": 2940
    },
    {
      "epoch": 0.604136801146836,
      "grad_norm": 3.9375,
      "learning_rate": 8.182569288836015e-06,
      "loss": 0.5768,
      "step": 2950
    },
    {
      "epoch": 0.6061847225066558,
      "grad_norm": 3.65625,
      "learning_rate": 8.16976282489071e-06,
      "loss": 0.578,
      "step": 2960
    },
    {
      "epoch": 0.6082326438664756,
      "grad_norm": 3.21875,
      "learning_rate": 8.1569214990464e-06,
      "loss": 0.461,
      "step": 2970
    },
    {
      "epoch": 0.6102805652262953,
      "grad_norm": 3.09375,
      "learning_rate": 8.144045452535418e-06,
      "loss": 0.5564,
      "step": 2980
    },
    {
      "epoch": 0.612328486586115,
      "grad_norm": 3.828125,
      "learning_rate": 8.131134826971967e-06,
      "loss": 0.5592,
      "step": 2990
    },
    {
      "epoch": 0.6143764079459348,
      "grad_norm": 4.4375,
      "learning_rate": 8.118189764350557e-06,
      "loss": 0.6086,
      "step": 3000
    },
    {
      "epoch": 0.6143764079459348,
      "eval_loss": 0.564007043838501,
      "eval_runtime": 49.3837,
      "eval_samples_per_second": 83.266,
      "eval_steps_per_second": 41.633,
      "step": 3000
    },
    {
      "epoch": 0.6164243293057546,
      "grad_norm": 3.890625,
      "learning_rate": 8.105210407044452e-06,
      "loss": 0.6133,
      "step": 3010
    },
    {
      "epoch": 0.6184722506655744,
      "grad_norm": 3.734375,
      "learning_rate": 8.09219689780409e-06,
      "loss": 0.5601,
      "step": 3020
    },
    {
      "epoch": 0.6205201720253942,
      "grad_norm": 3.828125,
      "learning_rate": 8.079149379755525e-06,
      "loss": 0.569,
      "step": 3030
    },
    {
      "epoch": 0.622568093385214,
      "grad_norm": 4.375,
      "learning_rate": 8.066067996398853e-06,
      "loss": 0.5513,
      "step": 3040
    },
    {
      "epoch": 0.6246160147450338,
      "grad_norm": 4.4375,
      "learning_rate": 8.052952891606617e-06,
      "loss": 0.5664,
      "step": 3050
    },
    {
      "epoch": 0.6266639361048536,
      "grad_norm": 4.09375,
      "learning_rate": 8.039804209622255e-06,
      "loss": 0.5716,
      "step": 3060
    },
    {
      "epoch": 0.6287118574646734,
      "grad_norm": 3.6875,
      "learning_rate": 8.026622095058483e-06,
      "loss": 0.5099,
      "step": 3070
    },
    {
      "epoch": 0.6307597788244932,
      "grad_norm": 4.125,
      "learning_rate": 8.01340669289572e-06,
      "loss": 0.5344,
      "step": 3080
    },
    {
      "epoch": 0.6328077001843129,
      "grad_norm": 3.78125,
      "learning_rate": 8.000158148480497e-06,
      "loss": 0.5317,
      "step": 3090
    },
    {
      "epoch": 0.6348556215441327,
      "grad_norm": 3.25,
      "learning_rate": 7.986876607523842e-06,
      "loss": 0.5768,
      "step": 3100
    },
    {
      "epoch": 0.6369035429039525,
      "grad_norm": 6.375,
      "learning_rate": 7.973562216099694e-06,
      "loss": 0.5755,
      "step": 3110
    },
    {
      "epoch": 0.6389514642637723,
      "grad_norm": 3.78125,
      "learning_rate": 7.960215120643294e-06,
      "loss": 0.5833,
      "step": 3120
    },
    {
      "epoch": 0.640999385623592,
      "grad_norm": 4.1875,
      "learning_rate": 7.946835467949561e-06,
      "loss": 0.5504,
      "step": 3130
    },
    {
      "epoch": 0.6430473069834118,
      "grad_norm": 4.0,
      "learning_rate": 7.933423405171496e-06,
      "loss": 0.5595,
      "step": 3140
    },
    {
      "epoch": 0.6450952283432316,
      "grad_norm": 3.984375,
      "learning_rate": 7.91997907981855e-06,
      "loss": 0.5771,
      "step": 3150
    },
    {
      "epoch": 0.6471431497030514,
      "grad_norm": 3.953125,
      "learning_rate": 7.906502639755008e-06,
      "loss": 0.5923,
      "step": 3160
    },
    {
      "epoch": 0.6491910710628712,
      "grad_norm": 3.0,
      "learning_rate": 7.892994233198361e-06,
      "loss": 0.5906,
      "step": 3170
    },
    {
      "epoch": 0.651238992422691,
      "grad_norm": 3.671875,
      "learning_rate": 7.879454008717675e-06,
      "loss": 0.5744,
      "step": 3180
    },
    {
      "epoch": 0.6532869137825108,
      "grad_norm": 3.78125,
      "learning_rate": 7.865882115231959e-06,
      "loss": 0.5737,
      "step": 3190
    },
    {
      "epoch": 0.6553348351423305,
      "grad_norm": 4.1875,
      "learning_rate": 7.852278702008527e-06,
      "loss": 0.5927,
      "step": 3200
    },
    {
      "epoch": 0.6573827565021503,
      "grad_norm": 3.625,
      "learning_rate": 7.838643918661354e-06,
      "loss": 0.5658,
      "step": 3210
    },
    {
      "epoch": 0.6594306778619701,
      "grad_norm": 3.046875,
      "learning_rate": 7.824977915149432e-06,
      "loss": 0.5321,
      "step": 3220
    },
    {
      "epoch": 0.6614785992217899,
      "grad_norm": 3.390625,
      "learning_rate": 7.811280841775124e-06,
      "loss": 0.5156,
      "step": 3230
    },
    {
      "epoch": 0.6635265205816097,
      "grad_norm": 3.578125,
      "learning_rate": 7.797552849182502e-06,
      "loss": 0.5588,
      "step": 3240
    },
    {
      "epoch": 0.6655744419414295,
      "grad_norm": 3.90625,
      "learning_rate": 7.783794088355698e-06,
      "loss": 0.6302,
      "step": 3250
    },
    {
      "epoch": 0.6676223633012492,
      "grad_norm": 3.1875,
      "learning_rate": 7.770004710617247e-06,
      "loss": 0.5842,
      "step": 3260
    },
    {
      "epoch": 0.669670284661069,
      "grad_norm": 3.015625,
      "learning_rate": 7.756184867626407e-06,
      "loss": 0.5592,
      "step": 3270
    },
    {
      "epoch": 0.6717182060208888,
      "grad_norm": 3.625,
      "learning_rate": 7.742334711377509e-06,
      "loss": 0.5674,
      "step": 3280
    },
    {
      "epoch": 0.6737661273807086,
      "grad_norm": 3.90625,
      "learning_rate": 7.728454394198271e-06,
      "loss": 0.6026,
      "step": 3290
    },
    {
      "epoch": 0.6758140487405284,
      "grad_norm": 3.265625,
      "learning_rate": 7.714544068748137e-06,
      "loss": 0.5144,
      "step": 3300
    },
    {
      "epoch": 0.6778619701003481,
      "grad_norm": 3.625,
      "learning_rate": 7.700603888016583e-06,
      "loss": 0.5396,
      "step": 3310
    },
    {
      "epoch": 0.6799098914601679,
      "grad_norm": 3.796875,
      "learning_rate": 7.686634005321442e-06,
      "loss": 0.5889,
      "step": 3320
    },
    {
      "epoch": 0.6819578128199877,
      "grad_norm": 2.921875,
      "learning_rate": 7.672634574307223e-06,
      "loss": 0.5414,
      "step": 3330
    },
    {
      "epoch": 0.6840057341798075,
      "grad_norm": 2.78125,
      "learning_rate": 7.658605748943407e-06,
      "loss": 0.4579,
      "step": 3340
    },
    {
      "epoch": 0.6860536555396273,
      "grad_norm": 4.5625,
      "learning_rate": 7.644547683522767e-06,
      "loss": 0.5635,
      "step": 3350
    },
    {
      "epoch": 0.6881015768994471,
      "grad_norm": 3.59375,
      "learning_rate": 7.630460532659667e-06,
      "loss": 0.555,
      "step": 3360
    },
    {
      "epoch": 0.6901494982592669,
      "grad_norm": 4.375,
      "learning_rate": 7.616344451288357e-06,
      "loss": 0.5808,
      "step": 3370
    },
    {
      "epoch": 0.6921974196190867,
      "grad_norm": 3.28125,
      "learning_rate": 7.602199594661273e-06,
      "loss": 0.5623,
      "step": 3380
    },
    {
      "epoch": 0.6942453409789064,
      "grad_norm": 3.46875,
      "learning_rate": 7.588026118347334e-06,
      "loss": 0.5326,
      "step": 3390
    },
    {
      "epoch": 0.6962932623387262,
      "grad_norm": 2.84375,
      "learning_rate": 7.5738241782302156e-06,
      "loss": 0.5501,
      "step": 3400
    },
    {
      "epoch": 0.698341183698546,
      "grad_norm": 3.96875,
      "learning_rate": 7.559593930506658e-06,
      "loss": 0.5593,
      "step": 3410
    },
    {
      "epoch": 0.7003891050583657,
      "grad_norm": 2.90625,
      "learning_rate": 7.545335531684725e-06,
      "loss": 0.5492,
      "step": 3420
    },
    {
      "epoch": 0.7024370264181855,
      "grad_norm": 4.96875,
      "learning_rate": 7.531049138582105e-06,
      "loss": 0.5853,
      "step": 3430
    },
    {
      "epoch": 0.7044849477780053,
      "grad_norm": 3.59375,
      "learning_rate": 7.516734908324363e-06,
      "loss": 0.5875,
      "step": 3440
    },
    {
      "epoch": 0.7065328691378251,
      "grad_norm": 3.453125,
      "learning_rate": 7.5023929983432344e-06,
      "loss": 0.5459,
      "step": 3450
    },
    {
      "epoch": 0.7085807904976449,
      "grad_norm": 3.171875,
      "learning_rate": 7.488023566374878e-06,
      "loss": 0.4829,
      "step": 3460
    },
    {
      "epoch": 0.7106287118574647,
      "grad_norm": 4.71875,
      "learning_rate": 7.4736267704581485e-06,
      "loss": 0.6025,
      "step": 3470
    },
    {
      "epoch": 0.7126766332172845,
      "grad_norm": 2.796875,
      "learning_rate": 7.459202768932856e-06,
      "loss": 0.5715,
      "step": 3480
    },
    {
      "epoch": 0.7147245545771043,
      "grad_norm": 3.21875,
      "learning_rate": 7.444751720438026e-06,
      "loss": 0.5352,
      "step": 3490
    },
    {
      "epoch": 0.7167724759369241,
      "grad_norm": 3.953125,
      "learning_rate": 7.430273783910153e-06,
      "loss": 0.554,
      "step": 3500
    },
    {
      "epoch": 0.7188203972967439,
      "grad_norm": 3.375,
      "learning_rate": 7.415769118581456e-06,
      "loss": 0.4954,
      "step": 3510
    },
    {
      "epoch": 0.7208683186565635,
      "grad_norm": 3.859375,
      "learning_rate": 7.40123788397812e-06,
      "loss": 0.5202,
      "step": 3520
    },
    {
      "epoch": 0.7229162400163833,
      "grad_norm": 3.828125,
      "learning_rate": 7.386680239918548e-06,
      "loss": 0.5294,
      "step": 3530
    },
    {
      "epoch": 0.7249641613762031,
      "grad_norm": 4.09375,
      "learning_rate": 7.372096346511602e-06,
      "loss": 0.5507,
      "step": 3540
    },
    {
      "epoch": 0.7270120827360229,
      "grad_norm": 3.0625,
      "learning_rate": 7.357486364154842e-06,
      "loss": 0.5977,
      "step": 3550
    },
    {
      "epoch": 0.7290600040958427,
      "grad_norm": 2.34375,
      "learning_rate": 7.3428504535327585e-06,
      "loss": 0.5502,
      "step": 3560
    },
    {
      "epoch": 0.7311079254556625,
      "grad_norm": 3.453125,
      "learning_rate": 7.328188775615009e-06,
      "loss": 0.5709,
      "step": 3570
    },
    {
      "epoch": 0.7331558468154823,
      "grad_norm": 3.265625,
      "learning_rate": 7.313501491654649e-06,
      "loss": 0.5813,
      "step": 3580
    },
    {
      "epoch": 0.7352037681753021,
      "grad_norm": 3.375,
      "learning_rate": 7.298788763186353e-06,
      "loss": 0.5048,
      "step": 3590
    },
    {
      "epoch": 0.7372516895351219,
      "grad_norm": 3.3125,
      "learning_rate": 7.284050752024643e-06,
      "loss": 0.5842,
      "step": 3600
    },
    {
      "epoch": 0.7392996108949417,
      "grad_norm": 3.515625,
      "learning_rate": 7.269287620262105e-06,
      "loss": 0.5087,
      "step": 3610
    },
    {
      "epoch": 0.7413475322547615,
      "grad_norm": 3.578125,
      "learning_rate": 7.25449953026761e-06,
      "loss": 0.5671,
      "step": 3620
    },
    {
      "epoch": 0.7433954536145811,
      "grad_norm": 3.90625,
      "learning_rate": 7.239686644684525e-06,
      "loss": 0.5711,
      "step": 3630
    },
    {
      "epoch": 0.7454433749744009,
      "grad_norm": 3.703125,
      "learning_rate": 7.224849126428925e-06,
      "loss": 0.5572,
      "step": 3640
    },
    {
      "epoch": 0.7474912963342207,
      "grad_norm": 4.5,
      "learning_rate": 7.209987138687806e-06,
      "loss": 0.543,
      "step": 3650
    },
    {
      "epoch": 0.7495392176940405,
      "grad_norm": 3.5,
      "learning_rate": 7.195100844917281e-06,
      "loss": 0.5518,
      "step": 3660
    },
    {
      "epoch": 0.7515871390538603,
      "grad_norm": 2.875,
      "learning_rate": 7.180190408840788e-06,
      "loss": 0.5479,
      "step": 3670
    },
    {
      "epoch": 0.7536350604136801,
      "grad_norm": 3.953125,
      "learning_rate": 7.165255994447288e-06,
      "loss": 0.5755,
      "step": 3680
    },
    {
      "epoch": 0.7556829817734999,
      "grad_norm": 4.0,
      "learning_rate": 7.150297765989467e-06,
      "loss": 0.6021,
      "step": 3690
    },
    {
      "epoch": 0.7577309031333197,
      "grad_norm": 3.734375,
      "learning_rate": 7.135315887981918e-06,
      "loss": 0.5618,
      "step": 3700
    },
    {
      "epoch": 0.7597788244931395,
      "grad_norm": 3.375,
      "learning_rate": 7.120310525199341e-06,
      "loss": 0.5302,
      "step": 3710
    },
    {
      "epoch": 0.7618267458529593,
      "grad_norm": 3.015625,
      "learning_rate": 7.105281842674729e-06,
      "loss": 0.4933,
      "step": 3720
    },
    {
      "epoch": 0.7638746672127791,
      "grad_norm": 3.140625,
      "learning_rate": 7.090230005697552e-06,
      "loss": 0.588,
      "step": 3730
    },
    {
      "epoch": 0.7659225885725988,
      "grad_norm": 3.546875,
      "learning_rate": 7.075155179811931e-06,
      "loss": 0.5628,
      "step": 3740
    },
    {
      "epoch": 0.7679705099324186,
      "grad_norm": 3.265625,
      "learning_rate": 7.0600575308148375e-06,
      "loss": 0.5523,
      "step": 3750
    },
    {
      "epoch": 0.7700184312922383,
      "grad_norm": 3.359375,
      "learning_rate": 7.044937224754249e-06,
      "loss": 0.5648,
      "step": 3760
    },
    {
      "epoch": 0.7720663526520581,
      "grad_norm": 3.46875,
      "learning_rate": 7.029794427927337e-06,
      "loss": 0.5495,
      "step": 3770
    },
    {
      "epoch": 0.7741142740118779,
      "grad_norm": 3.65625,
      "learning_rate": 7.0146293068786255e-06,
      "loss": 0.6386,
      "step": 3780
    },
    {
      "epoch": 0.7761621953716977,
      "grad_norm": 3.828125,
      "learning_rate": 6.999442028398171e-06,
      "loss": 0.5634,
      "step": 3790
    },
    {
      "epoch": 0.7782101167315175,
      "grad_norm": 4.34375,
      "learning_rate": 6.984232759519725e-06,
      "loss": 0.4956,
      "step": 3800
    },
    {
      "epoch": 0.7802580380913373,
      "grad_norm": 3.84375,
      "learning_rate": 6.969001667518891e-06,
      "loss": 0.5745,
      "step": 3810
    },
    {
      "epoch": 0.7823059594511571,
      "grad_norm": 3.171875,
      "learning_rate": 6.9537489199112915e-06,
      "loss": 0.5519,
      "step": 3820
    },
    {
      "epoch": 0.7843538808109769,
      "grad_norm": 3.1875,
      "learning_rate": 6.938474684450725e-06,
      "loss": 0.5338,
      "step": 3830
    },
    {
      "epoch": 0.7864018021707967,
      "grad_norm": 3.859375,
      "learning_rate": 6.923179129127312e-06,
      "loss": 0.641,
      "step": 3840
    },
    {
      "epoch": 0.7884497235306164,
      "grad_norm": 3.25,
      "learning_rate": 6.907862422165661e-06,
      "loss": 0.4959,
      "step": 3850
    },
    {
      "epoch": 0.7904976448904362,
      "grad_norm": 4.09375,
      "learning_rate": 6.892524732023013e-06,
      "loss": 0.5589,
      "step": 3860
    },
    {
      "epoch": 0.792545566250256,
      "grad_norm": 4.28125,
      "learning_rate": 6.877166227387382e-06,
      "loss": 0.6053,
      "step": 3870
    },
    {
      "epoch": 0.7945934876100758,
      "grad_norm": 3.125,
      "learning_rate": 6.86178707717571e-06,
      "loss": 0.5434,
      "step": 3880
    },
    {
      "epoch": 0.7966414089698955,
      "grad_norm": 4.0625,
      "learning_rate": 6.8463874505320004e-06,
      "loss": 0.5558,
      "step": 3890
    },
    {
      "epoch": 0.7986893303297153,
      "grad_norm": 4.15625,
      "learning_rate": 6.8309675168254664e-06,
      "loss": 0.5358,
      "step": 3900
    },
    {
      "epoch": 0.8007372516895351,
      "grad_norm": 3.921875,
      "learning_rate": 6.81552744564866e-06,
      "loss": 0.5228,
      "step": 3910
    },
    {
      "epoch": 0.8027851730493549,
      "grad_norm": 3.328125,
      "learning_rate": 6.800067406815612e-06,
      "loss": 0.5956,
      "step": 3920
    },
    {
      "epoch": 0.8048330944091747,
      "grad_norm": 3.140625,
      "learning_rate": 6.784587570359965e-06,
      "loss": 0.5785,
      "step": 3930
    },
    {
      "epoch": 0.8068810157689945,
      "grad_norm": 4.21875,
      "learning_rate": 6.769088106533097e-06,
      "loss": 0.5481,
      "step": 3940
    },
    {
      "epoch": 0.8089289371288143,
      "grad_norm": 4.21875,
      "learning_rate": 6.753569185802256e-06,
      "loss": 0.5554,
      "step": 3950
    },
    {
      "epoch": 0.810976858488634,
      "grad_norm": 4.125,
      "learning_rate": 6.73803097884868e-06,
      "loss": 0.5738,
      "step": 3960
    },
    {
      "epoch": 0.8130247798484538,
      "grad_norm": 2.984375,
      "learning_rate": 6.722473656565724e-06,
      "loss": 0.5275,
      "step": 3970
    },
    {
      "epoch": 0.8150727012082736,
      "grad_norm": 4.0625,
      "learning_rate": 6.706897390056979e-06,
      "loss": 0.597,
      "step": 3980
    },
    {
      "epoch": 0.8171206225680934,
      "grad_norm": 4.40625,
      "learning_rate": 6.691302350634386e-06,
      "loss": 0.5808,
      "step": 3990
    },
    {
      "epoch": 0.8191685439279132,
      "grad_norm": 2.9375,
      "learning_rate": 6.675688709816362e-06,
      "loss": 0.5217,
      "step": 4000
    },
    {
      "epoch": 0.8191685439279132,
      "eval_loss": 0.5564442873001099,
      "eval_runtime": 48.7272,
      "eval_samples_per_second": 84.388,
      "eval_steps_per_second": 42.194,
      "step": 4000
    },
    {
      "epoch": 0.821216465287733,
      "grad_norm": 3.203125,
      "learning_rate": 6.660056639325898e-06,
      "loss": 0.5558,
      "step": 4010
    },
    {
      "epoch": 0.8232643866475527,
      "grad_norm": 3.5,
      "learning_rate": 6.644406311088687e-06,
      "loss": 0.5654,
      "step": 4020
    },
    {
      "epoch": 0.8253123080073725,
      "grad_norm": 3.640625,
      "learning_rate": 6.628737897231226e-06,
      "loss": 0.5095,
      "step": 4030
    },
    {
      "epoch": 0.8273602293671923,
      "grad_norm": 2.9375,
      "learning_rate": 6.613051570078914e-06,
      "loss": 0.5459,
      "step": 4040
    },
    {
      "epoch": 0.8294081507270121,
      "grad_norm": 4.96875,
      "learning_rate": 6.597347502154177e-06,
      "loss": 0.5534,
      "step": 4050
    },
    {
      "epoch": 0.8314560720868319,
      "grad_norm": 3.875,
      "learning_rate": 6.581625866174548e-06,
      "loss": 0.598,
      "step": 4060
    },
    {
      "epoch": 0.8335039934466516,
      "grad_norm": 3.84375,
      "learning_rate": 6.565886835050785e-06,
      "loss": 0.5429,
      "step": 4070
    },
    {
      "epoch": 0.8355519148064714,
      "grad_norm": 3.4375,
      "learning_rate": 6.550130581884958e-06,
      "loss": 0.5431,
      "step": 4080
    },
    {
      "epoch": 0.8375998361662912,
      "grad_norm": 3.65625,
      "learning_rate": 6.534357279968554e-06,
      "loss": 0.5966,
      "step": 4090
    },
    {
      "epoch": 0.839647757526111,
      "grad_norm": 3.71875,
      "learning_rate": 6.5185671027805655e-06,
      "loss": 0.5669,
      "step": 4100
    },
    {
      "epoch": 0.8416956788859308,
      "grad_norm": 3.28125,
      "learning_rate": 6.5027602239855805e-06,
      "loss": 0.5071,
      "step": 4110
    },
    {
      "epoch": 0.8437436002457506,
      "grad_norm": 3.75,
      "learning_rate": 6.486936817431878e-06,
      "loss": 0.5571,
      "step": 4120
    },
    {
      "epoch": 0.8457915216055704,
      "grad_norm": 4.03125,
      "learning_rate": 6.471097057149516e-06,
      "loss": 0.6112,
      "step": 4130
    },
    {
      "epoch": 0.8478394429653902,
      "grad_norm": 3.21875,
      "learning_rate": 6.45524111734841e-06,
      "loss": 0.5302,
      "step": 4140
    },
    {
      "epoch": 0.84988736432521,
      "grad_norm": 4.4375,
      "learning_rate": 6.439369172416428e-06,
      "loss": 0.5725,
      "step": 4150
    },
    {
      "epoch": 0.8519352856850297,
      "grad_norm": 3.5625,
      "learning_rate": 6.4234813969174615e-06,
      "loss": 0.5502,
      "step": 4160
    },
    {
      "epoch": 0.8539832070448494,
      "grad_norm": 3.453125,
      "learning_rate": 6.407577965589515e-06,
      "loss": 0.5261,
      "step": 4170
    },
    {
      "epoch": 0.8560311284046692,
      "grad_norm": 3.46875,
      "learning_rate": 6.391659053342778e-06,
      "loss": 0.5446,
      "step": 4180
    },
    {
      "epoch": 0.858079049764489,
      "grad_norm": 3.65625,
      "learning_rate": 6.3757248352577015e-06,
      "loss": 0.6036,
      "step": 4190
    },
    {
      "epoch": 0.8601269711243088,
      "grad_norm": 3.203125,
      "learning_rate": 6.359775486583077e-06,
      "loss": 0.5096,
      "step": 4200
    },
    {
      "epoch": 0.8621748924841286,
      "grad_norm": 4.0,
      "learning_rate": 6.343811182734108e-06,
      "loss": 0.5413,
      "step": 4210
    },
    {
      "epoch": 0.8642228138439484,
      "grad_norm": 3.953125,
      "learning_rate": 6.327832099290473e-06,
      "loss": 0.5094,
      "step": 4220
    },
    {
      "epoch": 0.8662707352037682,
      "grad_norm": 3.90625,
      "learning_rate": 6.311838411994404e-06,
      "loss": 0.5392,
      "step": 4230
    },
    {
      "epoch": 0.868318656563588,
      "grad_norm": 3.484375,
      "learning_rate": 6.295830296748753e-06,
      "loss": 0.6021,
      "step": 4240
    },
    {
      "epoch": 0.8703665779234078,
      "grad_norm": 3.171875,
      "learning_rate": 6.279807929615047e-06,
      "loss": 0.5416,
      "step": 4250
    },
    {
      "epoch": 0.8724144992832276,
      "grad_norm": 3.796875,
      "learning_rate": 6.263771486811567e-06,
      "loss": 0.5066,
      "step": 4260
    },
    {
      "epoch": 0.8744624206430474,
      "grad_norm": 3.890625,
      "learning_rate": 6.2477211447113975e-06,
      "loss": 0.5029,
      "step": 4270
    },
    {
      "epoch": 0.876510342002867,
      "grad_norm": 4.03125,
      "learning_rate": 6.231657079840491e-06,
      "loss": 0.5448,
      "step": 4280
    },
    {
      "epoch": 0.8785582633626868,
      "grad_norm": 3.4375,
      "learning_rate": 6.215579468875729e-06,
      "loss": 0.5299,
      "step": 4290
    },
    {
      "epoch": 0.8806061847225066,
      "grad_norm": 3.53125,
      "learning_rate": 6.199488488642975e-06,
      "loss": 0.5806,
      "step": 4300
    },
    {
      "epoch": 0.8826541060823264,
      "grad_norm": 3.375,
      "learning_rate": 6.18338431611513e-06,
      "loss": 0.5305,
      "step": 4310
    },
    {
      "epoch": 0.8847020274421462,
      "grad_norm": 4.15625,
      "learning_rate": 6.167267128410191e-06,
      "loss": 0.5216,
      "step": 4320
    },
    {
      "epoch": 0.886749948801966,
      "grad_norm": 3.375,
      "learning_rate": 6.151137102789297e-06,
      "loss": 0.5515,
      "step": 4330
    },
    {
      "epoch": 0.8887978701617858,
      "grad_norm": 3.0625,
      "learning_rate": 6.1349944166547825e-06,
      "loss": 0.5276,
      "step": 4340
    },
    {
      "epoch": 0.8908457915216056,
      "grad_norm": 3.171875,
      "learning_rate": 6.118839247548226e-06,
      "loss": 0.5427,
      "step": 4350
    },
    {
      "epoch": 0.8928937128814254,
      "grad_norm": 3.171875,
      "learning_rate": 6.102671773148494e-06,
      "loss": 0.5531,
      "step": 4360
    },
    {
      "epoch": 0.8949416342412452,
      "grad_norm": 3.4375,
      "learning_rate": 6.086492171269794e-06,
      "loss": 0.509,
      "step": 4370
    },
    {
      "epoch": 0.896989555601065,
      "grad_norm": 4.5625,
      "learning_rate": 6.070300619859715e-06,
      "loss": 0.5731,
      "step": 4380
    },
    {
      "epoch": 0.8990374769608847,
      "grad_norm": 3.1875,
      "learning_rate": 6.054097296997266e-06,
      "loss": 0.5113,
      "step": 4390
    },
    {
      "epoch": 0.9010853983207044,
      "grad_norm": 5.03125,
      "learning_rate": 6.037882380890924e-06,
      "loss": 0.5583,
      "step": 4400
    },
    {
      "epoch": 0.9031333196805242,
      "grad_norm": 3.421875,
      "learning_rate": 6.021656049876672e-06,
      "loss": 0.5341,
      "step": 4410
    },
    {
      "epoch": 0.905181241040344,
      "grad_norm": 4.0,
      "learning_rate": 6.005418482416036e-06,
      "loss": 0.5671,
      "step": 4420
    },
    {
      "epoch": 0.9072291624001638,
      "grad_norm": 3.96875,
      "learning_rate": 5.989169857094126e-06,
      "loss": 0.5846,
      "step": 4430
    },
    {
      "epoch": 0.9092770837599836,
      "grad_norm": 3.203125,
      "learning_rate": 5.972910352617667e-06,
      "loss": 0.552,
      "step": 4440
    },
    {
      "epoch": 0.9113250051198034,
      "grad_norm": 3.78125,
      "learning_rate": 5.956640147813035e-06,
      "loss": 0.5681,
      "step": 4450
    },
    {
      "epoch": 0.9133729264796232,
      "grad_norm": 4.0,
      "learning_rate": 5.940359421624292e-06,
      "loss": 0.5831,
      "step": 4460
    },
    {
      "epoch": 0.915420847839443,
      "grad_norm": 3.21875,
      "learning_rate": 5.92406835311122e-06,
      "loss": 0.5516,
      "step": 4470
    },
    {
      "epoch": 0.9174687691992628,
      "grad_norm": 3.703125,
      "learning_rate": 5.907767121447343e-06,
      "loss": 0.5779,
      "step": 4480
    },
    {
      "epoch": 0.9195166905590826,
      "grad_norm": 4.09375,
      "learning_rate": 5.891455905917968e-06,
      "loss": 0.6099,
      "step": 4490
    },
    {
      "epoch": 0.9215646119189023,
      "grad_norm": 3.34375,
      "learning_rate": 5.8751348859182025e-06,
      "loss": 0.5506,
      "step": 4500
    },
    {
      "epoch": 0.9236125332787221,
      "grad_norm": 4.03125,
      "learning_rate": 5.858804240950988e-06,
      "loss": 0.533,
      "step": 4510
    },
    {
      "epoch": 0.9256604546385419,
      "grad_norm": 4.03125,
      "learning_rate": 5.842464150625125e-06,
      "loss": 0.5653,
      "step": 4520
    },
    {
      "epoch": 0.9277083759983616,
      "grad_norm": 3.484375,
      "learning_rate": 5.826114794653294e-06,
      "loss": 0.5573,
      "step": 4530
    },
    {
      "epoch": 0.9297562973581814,
      "grad_norm": 3.6875,
      "learning_rate": 5.809756352850083e-06,
      "loss": 0.5227,
      "step": 4540
    },
    {
      "epoch": 0.9318042187180012,
      "grad_norm": 3.734375,
      "learning_rate": 5.793389005130011e-06,
      "loss": 0.538,
      "step": 4550
    },
    {
      "epoch": 0.933852140077821,
      "grad_norm": 3.765625,
      "learning_rate": 5.7770129315055435e-06,
      "loss": 0.514,
      "step": 4560
    },
    {
      "epoch": 0.9359000614376408,
      "grad_norm": 3.75,
      "learning_rate": 5.760628312085114e-06,
      "loss": 0.5736,
      "step": 4570
    },
    {
      "epoch": 0.9379479827974606,
      "grad_norm": 4.28125,
      "learning_rate": 5.744235327071151e-06,
      "loss": 0.5473,
      "step": 4580
    },
    {
      "epoch": 0.9399959041572804,
      "grad_norm": 4.1875,
      "learning_rate": 5.727834156758082e-06,
      "loss": 0.5422,
      "step": 4590
    },
    {
      "epoch": 0.9420438255171002,
      "grad_norm": 3.203125,
      "learning_rate": 5.711424981530367e-06,
      "loss": 0.5134,
      "step": 4600
    },
    {
      "epoch": 0.9440917468769199,
      "grad_norm": 3.453125,
      "learning_rate": 5.6950079818605e-06,
      "loss": 0.5406,
      "step": 4610
    },
    {
      "epoch": 0.9461396682367397,
      "grad_norm": 4.5,
      "learning_rate": 5.678583338307033e-06,
      "loss": 0.6124,
      "step": 4620
    },
    {
      "epoch": 0.9481875895965595,
      "grad_norm": 3.59375,
      "learning_rate": 5.662151231512588e-06,
      "loss": 0.5095,
      "step": 4630
    },
    {
      "epoch": 0.9502355109563793,
      "grad_norm": 5.15625,
      "learning_rate": 5.645711842201869e-06,
      "loss": 0.5337,
      "step": 4640
    },
    {
      "epoch": 0.952283432316199,
      "grad_norm": 3.4375,
      "learning_rate": 5.629265351179673e-06,
      "loss": 0.5252,
      "step": 4650
    },
    {
      "epoch": 0.9543313536760188,
      "grad_norm": 3.390625,
      "learning_rate": 5.612811939328907e-06,
      "loss": 0.5631,
      "step": 4660
    },
    {
      "epoch": 0.9563792750358386,
      "grad_norm": 3.78125,
      "learning_rate": 5.596351787608592e-06,
      "loss": 0.5531,
      "step": 4670
    },
    {
      "epoch": 0.9584271963956584,
      "grad_norm": 3.671875,
      "learning_rate": 5.579885077051879e-06,
      "loss": 0.5673,
      "step": 4680
    },
    {
      "epoch": 0.9604751177554782,
      "grad_norm": 4.0625,
      "learning_rate": 5.5634119887640515e-06,
      "loss": 0.5721,
      "step": 4690
    },
    {
      "epoch": 0.962523039115298,
      "grad_norm": 3.875,
      "learning_rate": 5.546932703920539e-06,
      "loss": 0.577,
      "step": 4700
    },
    {
      "epoch": 0.9645709604751177,
      "grad_norm": 3.5625,
      "learning_rate": 5.530447403764924e-06,
      "loss": 0.4924,
      "step": 4710
    },
    {
      "epoch": 0.9666188818349375,
      "grad_norm": 3.875,
      "learning_rate": 5.513956269606944e-06,
      "loss": 0.5552,
      "step": 4720
    },
    {
      "epoch": 0.9686668031947573,
      "grad_norm": 3.8125,
      "learning_rate": 5.497459482820499e-06,
      "loss": 0.6076,
      "step": 4730
    },
    {
      "epoch": 0.9707147245545771,
      "grad_norm": 3.65625,
      "learning_rate": 5.480957224841666e-06,
      "loss": 0.5824,
      "step": 4740
    },
    {
      "epoch": 0.9727626459143969,
      "grad_norm": 4.59375,
      "learning_rate": 5.4644496771666845e-06,
      "loss": 0.5634,
      "step": 4750
    },
    {
      "epoch": 0.9748105672742167,
      "grad_norm": 4.3125,
      "learning_rate": 5.447937021349979e-06,
      "loss": 0.5331,
      "step": 4760
    },
    {
      "epoch": 0.9768584886340365,
      "grad_norm": 3.4375,
      "learning_rate": 5.431419439002155e-06,
      "loss": 0.5859,
      "step": 4770
    },
    {
      "epoch": 0.9789064099938563,
      "grad_norm": 2.765625,
      "learning_rate": 5.414897111787998e-06,
      "loss": 0.5416,
      "step": 4780
    },
    {
      "epoch": 0.980954331353676,
      "grad_norm": 3.3125,
      "learning_rate": 5.3983702214244805e-06,
      "loss": 0.5383,
      "step": 4790
    },
    {
      "epoch": 0.9830022527134958,
      "grad_norm": 2.859375,
      "learning_rate": 5.381838949678759e-06,
      "loss": 0.5247,
      "step": 4800
    },
    {
      "epoch": 0.9850501740733156,
      "grad_norm": 4.21875,
      "learning_rate": 5.365303478366184e-06,
      "loss": 0.5983,
      "step": 4810
    },
    {
      "epoch": 0.9870980954331353,
      "grad_norm": 3.96875,
      "learning_rate": 5.348763989348285e-06,
      "loss": 0.5884,
      "step": 4820
    },
    {
      "epoch": 0.9891460167929551,
      "grad_norm": 3.5,
      "learning_rate": 5.332220664530788e-06,
      "loss": 0.5408,
      "step": 4830
    },
    {
      "epoch": 0.9911939381527749,
      "grad_norm": 3.421875,
      "learning_rate": 5.315673685861602e-06,
      "loss": 0.4796,
      "step": 4840
    },
    {
      "epoch": 0.9932418595125947,
      "grad_norm": 4.0625,
      "learning_rate": 5.29912323532882e-06,
      "loss": 0.5175,
      "step": 4850
    },
    {
      "epoch": 0.9952897808724145,
      "grad_norm": 3.515625,
      "learning_rate": 5.282569494958723e-06,
      "loss": 0.5127,
      "step": 4860
    },
    {
      "epoch": 0.9973377022322343,
      "grad_norm": 3.796875,
      "learning_rate": 5.266012646813774e-06,
      "loss": 0.5912,
      "step": 4870
    },
    {
      "epoch": 0.9993856235920541,
      "grad_norm": 4.21875,
      "learning_rate": 5.249452872990616e-06,
      "loss": 0.5986,
      "step": 4880
    },
    {
      "epoch": 1.0014335449518739,
      "grad_norm": 4.21875,
      "learning_rate": 5.2328903556180666e-06,
      "loss": 0.5311,
      "step": 4890
    },
    {
      "epoch": 1.0034814663116935,
      "grad_norm": 3.15625,
      "learning_rate": 5.216325276855122e-06,
      "loss": 0.5112,
      "step": 4900
    },
    {
      "epoch": 1.0055293876715135,
      "grad_norm": 3.53125,
      "learning_rate": 5.1997578188889465e-06,
      "loss": 0.4965,
      "step": 4910
    },
    {
      "epoch": 1.0075773090313331,
      "grad_norm": 3.46875,
      "learning_rate": 5.183188163932873e-06,
      "loss": 0.4898,
      "step": 4920
    },
    {
      "epoch": 1.009625230391153,
      "grad_norm": 3.359375,
      "learning_rate": 5.166616494224393e-06,
      "loss": 0.5006,
      "step": 4930
    },
    {
      "epoch": 1.0116731517509727,
      "grad_norm": 3.625,
      "learning_rate": 5.150042992023166e-06,
      "loss": 0.5103,
      "step": 4940
    },
    {
      "epoch": 1.0137210731107926,
      "grad_norm": 3.609375,
      "learning_rate": 5.133467839608998e-06,
      "loss": 0.5185,
      "step": 4950
    },
    {
      "epoch": 1.0157689944706123,
      "grad_norm": 3.140625,
      "learning_rate": 5.116891219279846e-06,
      "loss": 0.517,
      "step": 4960
    },
    {
      "epoch": 1.0178169158304322,
      "grad_norm": 4.25,
      "learning_rate": 5.1003133133498115e-06,
      "loss": 0.5011,
      "step": 4970
    },
    {
      "epoch": 1.019864837190252,
      "grad_norm": 3.578125,
      "learning_rate": 5.083734304147139e-06,
      "loss": 0.5249,
      "step": 4980
    },
    {
      "epoch": 1.0219127585500716,
      "grad_norm": 3.875,
      "learning_rate": 5.067154374012201e-06,
      "loss": 0.5371,
      "step": 4990
    },
    {
      "epoch": 1.0239606799098915,
      "grad_norm": 3.21875,
      "learning_rate": 5.050573705295504e-06,
      "loss": 0.5436,
      "step": 5000
    },
    {
      "epoch": 1.0239606799098915,
      "eval_loss": 0.5540497303009033,
      "eval_runtime": 48.9146,
      "eval_samples_per_second": 84.065,
      "eval_steps_per_second": 42.032,
      "step": 5000
    },
    {
      "epoch": 1.0260086012697112,
      "grad_norm": 3.75,
      "learning_rate": 5.033992480355675e-06,
      "loss": 0.5154,
      "step": 5010
    },
    {
      "epoch": 1.028056522629531,
      "grad_norm": 3.4375,
      "learning_rate": 5.01741088155746e-06,
      "loss": 0.4631,
      "step": 5020
    },
    {
      "epoch": 1.0301044439893507,
      "grad_norm": 3.75,
      "learning_rate": 5.000829091269713e-06,
      "loss": 0.4978,
      "step": 5030
    },
    {
      "epoch": 1.0321523653491707,
      "grad_norm": 3.828125,
      "learning_rate": 4.984247291863399e-06,
      "loss": 0.5375,
      "step": 5040
    },
    {
      "epoch": 1.0342002867089903,
      "grad_norm": 3.796875,
      "learning_rate": 4.967665665709582e-06,
      "loss": 0.4977,
      "step": 5050
    },
    {
      "epoch": 1.0362482080688102,
      "grad_norm": 3.390625,
      "learning_rate": 4.951084395177416e-06,
      "loss": 0.4551,
      "step": 5060
    },
    {
      "epoch": 1.03829612942863,
      "grad_norm": 4.125,
      "learning_rate": 4.93450366263215e-06,
      "loss": 0.4873,
      "step": 5070
    },
    {
      "epoch": 1.0403440507884498,
      "grad_norm": 3.90625,
      "learning_rate": 4.917923650433116e-06,
      "loss": 0.4607,
      "step": 5080
    },
    {
      "epoch": 1.0423919721482695,
      "grad_norm": 4.34375,
      "learning_rate": 4.901344540931715e-06,
      "loss": 0.5109,
      "step": 5090
    },
    {
      "epoch": 1.0444398935080892,
      "grad_norm": 4.25,
      "learning_rate": 4.884766516469433e-06,
      "loss": 0.4653,
      "step": 5100
    },
    {
      "epoch": 1.046487814867909,
      "grad_norm": 3.625,
      "learning_rate": 4.868189759375807e-06,
      "loss": 0.5088,
      "step": 5110
    },
    {
      "epoch": 1.0485357362277288,
      "grad_norm": 3.921875,
      "learning_rate": 4.851614451966451e-06,
      "loss": 0.498,
      "step": 5120
    },
    {
      "epoch": 1.0505836575875487,
      "grad_norm": 3.53125,
      "learning_rate": 4.835040776541022e-06,
      "loss": 0.5035,
      "step": 5130
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 3.3125,
      "learning_rate": 4.818468915381236e-06,
      "loss": 0.4899,
      "step": 5140
    },
    {
      "epoch": 1.0546795003071883,
      "grad_norm": 3.734375,
      "learning_rate": 4.801899050748854e-06,
      "loss": 0.524,
      "step": 5150
    },
    {
      "epoch": 1.056727421667008,
      "grad_norm": 4.125,
      "learning_rate": 4.785331364883673e-06,
      "loss": 0.5365,
      "step": 5160
    },
    {
      "epoch": 1.0587753430268279,
      "grad_norm": 3.375,
      "learning_rate": 4.768766040001536e-06,
      "loss": 0.5169,
      "step": 5170
    },
    {
      "epoch": 1.0608232643866475,
      "grad_norm": 4.78125,
      "learning_rate": 4.752203258292312e-06,
      "loss": 0.5385,
      "step": 5180
    },
    {
      "epoch": 1.0628711857464674,
      "grad_norm": 3.234375,
      "learning_rate": 4.735643201917903e-06,
      "loss": 0.4745,
      "step": 5190
    },
    {
      "epoch": 1.0649191071062871,
      "grad_norm": 3.84375,
      "learning_rate": 4.7190860530102385e-06,
      "loss": 0.5175,
      "step": 5200
    },
    {
      "epoch": 1.0669670284661068,
      "grad_norm": 3.9375,
      "learning_rate": 4.702531993669265e-06,
      "loss": 0.5022,
      "step": 5210
    },
    {
      "epoch": 1.0690149498259267,
      "grad_norm": 3.921875,
      "learning_rate": 4.685981205960958e-06,
      "loss": 0.4885,
      "step": 5220
    },
    {
      "epoch": 1.0710628711857464,
      "grad_norm": 3.703125,
      "learning_rate": 4.669433871915301e-06,
      "loss": 0.5109,
      "step": 5230
    },
    {
      "epoch": 1.0731107925455663,
      "grad_norm": 3.78125,
      "learning_rate": 4.6528901735243e-06,
      "loss": 0.4801,
      "step": 5240
    },
    {
      "epoch": 1.075158713905386,
      "grad_norm": 4.125,
      "learning_rate": 4.636350292739974e-06,
      "loss": 0.5051,
      "step": 5250
    },
    {
      "epoch": 1.0772066352652059,
      "grad_norm": 3.71875,
      "learning_rate": 4.6198144114723506e-06,
      "loss": 0.488,
      "step": 5260
    },
    {
      "epoch": 1.0792545566250256,
      "grad_norm": 3.875,
      "learning_rate": 4.603282711587478e-06,
      "loss": 0.5006,
      "step": 5270
    },
    {
      "epoch": 1.0813024779848455,
      "grad_norm": 4.25,
      "learning_rate": 4.586755374905405e-06,
      "loss": 0.4973,
      "step": 5280
    },
    {
      "epoch": 1.0833503993446651,
      "grad_norm": 3.8125,
      "learning_rate": 4.570232583198206e-06,
      "loss": 0.5142,
      "step": 5290
    },
    {
      "epoch": 1.0853983207044848,
      "grad_norm": 3.625,
      "learning_rate": 4.553714518187955e-06,
      "loss": 0.4319,
      "step": 5300
    },
    {
      "epoch": 1.0874462420643047,
      "grad_norm": 4.03125,
      "learning_rate": 4.537201361544751e-06,
      "loss": 0.49,
      "step": 5310
    },
    {
      "epoch": 1.0894941634241244,
      "grad_norm": 3.609375,
      "learning_rate": 4.5206932948847075e-06,
      "loss": 0.5164,
      "step": 5320
    },
    {
      "epoch": 1.0915420847839443,
      "grad_norm": 4.09375,
      "learning_rate": 4.50419049976795e-06,
      "loss": 0.5184,
      "step": 5330
    },
    {
      "epoch": 1.093590006143764,
      "grad_norm": 3.4375,
      "learning_rate": 4.487693157696637e-06,
      "loss": 0.4799,
      "step": 5340
    },
    {
      "epoch": 1.095637927503584,
      "grad_norm": 3.625,
      "learning_rate": 4.471201450112942e-06,
      "loss": 0.4984,
      "step": 5350
    },
    {
      "epoch": 1.0976858488634036,
      "grad_norm": 4.21875,
      "learning_rate": 4.454715558397076e-06,
      "loss": 0.5458,
      "step": 5360
    },
    {
      "epoch": 1.0997337702232235,
      "grad_norm": 3.5,
      "learning_rate": 4.438235663865288e-06,
      "loss": 0.4502,
      "step": 5370
    },
    {
      "epoch": 1.1017816915830432,
      "grad_norm": 3.125,
      "learning_rate": 4.421761947767856e-06,
      "loss": 0.4981,
      "step": 5380
    },
    {
      "epoch": 1.103829612942863,
      "grad_norm": 3.40625,
      "learning_rate": 4.405294591287122e-06,
      "loss": 0.4796,
      "step": 5390
    },
    {
      "epoch": 1.1058775343026828,
      "grad_norm": 3.359375,
      "learning_rate": 4.388833775535469e-06,
      "loss": 0.4936,
      "step": 5400
    },
    {
      "epoch": 1.1079254556625027,
      "grad_norm": 3.265625,
      "learning_rate": 4.372379681553352e-06,
      "loss": 0.4897,
      "step": 5410
    },
    {
      "epoch": 1.1099733770223223,
      "grad_norm": 3.328125,
      "learning_rate": 4.3559324903072985e-06,
      "loss": 0.506,
      "step": 5420
    },
    {
      "epoch": 1.112021298382142,
      "grad_norm": 3.65625,
      "learning_rate": 4.3394923826879096e-06,
      "loss": 0.5046,
      "step": 5430
    },
    {
      "epoch": 1.114069219741962,
      "grad_norm": 3.734375,
      "learning_rate": 4.32305953950789e-06,
      "loss": 0.5027,
      "step": 5440
    },
    {
      "epoch": 1.1161171411017816,
      "grad_norm": 3.671875,
      "learning_rate": 4.306634141500037e-06,
      "loss": 0.5006,
      "step": 5450
    },
    {
      "epoch": 1.1181650624616015,
      "grad_norm": 4.15625,
      "learning_rate": 4.290216369315274e-06,
      "loss": 0.5329,
      "step": 5460
    },
    {
      "epoch": 1.1202129838214212,
      "grad_norm": 4.1875,
      "learning_rate": 4.273806403520644e-06,
      "loss": 0.4763,
      "step": 5470
    },
    {
      "epoch": 1.122260905181241,
      "grad_norm": 3.6875,
      "learning_rate": 4.257404424597342e-06,
      "loss": 0.5092,
      "step": 5480
    },
    {
      "epoch": 1.1243088265410608,
      "grad_norm": 3.03125,
      "learning_rate": 4.241010612938719e-06,
      "loss": 0.4719,
      "step": 5490
    },
    {
      "epoch": 1.1263567479008807,
      "grad_norm": 3.21875,
      "learning_rate": 4.224625148848292e-06,
      "loss": 0.4901,
      "step": 5500
    },
    {
      "epoch": 1.1284046692607004,
      "grad_norm": 3.296875,
      "learning_rate": 4.208248212537783e-06,
      "loss": 0.4323,
      "step": 5510
    },
    {
      "epoch": 1.13045259062052,
      "grad_norm": 4.40625,
      "learning_rate": 4.19187998412511e-06,
      "loss": 0.4783,
      "step": 5520
    },
    {
      "epoch": 1.13250051198034,
      "grad_norm": 3.65625,
      "learning_rate": 4.175520643632428e-06,
      "loss": 0.5215,
      "step": 5530
    },
    {
      "epoch": 1.1345484333401596,
      "grad_norm": 5.09375,
      "learning_rate": 4.15917037098414e-06,
      "loss": 0.4825,
      "step": 5540
    },
    {
      "epoch": 1.1365963546999795,
      "grad_norm": 4.0625,
      "learning_rate": 4.142829346004911e-06,
      "loss": 0.4754,
      "step": 5550
    },
    {
      "epoch": 1.1386442760597992,
      "grad_norm": 3.96875,
      "learning_rate": 4.126497748417708e-06,
      "loss": 0.4967,
      "step": 5560
    },
    {
      "epoch": 1.1406921974196191,
      "grad_norm": 3.4375,
      "learning_rate": 4.110175757841802e-06,
      "loss": 0.5174,
      "step": 5570
    },
    {
      "epoch": 1.1427401187794388,
      "grad_norm": 4.25,
      "learning_rate": 4.093863553790813e-06,
      "loss": 0.5289,
      "step": 5580
    },
    {
      "epoch": 1.1447880401392587,
      "grad_norm": 4.09375,
      "learning_rate": 4.077561315670721e-06,
      "loss": 0.5478,
      "step": 5590
    },
    {
      "epoch": 1.1468359614990784,
      "grad_norm": 3.59375,
      "learning_rate": 4.061269222777898e-06,
      "loss": 0.4968,
      "step": 5600
    },
    {
      "epoch": 1.1488838828588983,
      "grad_norm": 4.3125,
      "learning_rate": 4.044987454297142e-06,
      "loss": 0.4841,
      "step": 5610
    },
    {
      "epoch": 1.150931804218718,
      "grad_norm": 3.625,
      "learning_rate": 4.028716189299691e-06,
      "loss": 0.4725,
      "step": 5620
    },
    {
      "epoch": 1.152979725578538,
      "grad_norm": 4.4375,
      "learning_rate": 4.01245560674127e-06,
      "loss": 0.5094,
      "step": 5630
    },
    {
      "epoch": 1.1550276469383576,
      "grad_norm": 3.265625,
      "learning_rate": 3.996205885460112e-06,
      "loss": 0.5184,
      "step": 5640
    },
    {
      "epoch": 1.1570755682981773,
      "grad_norm": 3.9375,
      "learning_rate": 3.979967204174998e-06,
      "loss": 0.4978,
      "step": 5650
    },
    {
      "epoch": 1.1591234896579972,
      "grad_norm": 4.09375,
      "learning_rate": 3.963739741483285e-06,
      "loss": 0.4841,
      "step": 5660
    },
    {
      "epoch": 1.1611714110178168,
      "grad_norm": 3.9375,
      "learning_rate": 3.947523675858945e-06,
      "loss": 0.5245,
      "step": 5670
    },
    {
      "epoch": 1.1632193323776367,
      "grad_norm": 3.703125,
      "learning_rate": 3.931319185650606e-06,
      "loss": 0.4784,
      "step": 5680
    },
    {
      "epoch": 1.1652672537374564,
      "grad_norm": 3.703125,
      "learning_rate": 3.915126449079582e-06,
      "loss": 0.5205,
      "step": 5690
    },
    {
      "epoch": 1.1673151750972763,
      "grad_norm": 4.6875,
      "learning_rate": 3.898945644237919e-06,
      "loss": 0.5475,
      "step": 5700
    },
    {
      "epoch": 1.169363096457096,
      "grad_norm": 3.96875,
      "learning_rate": 3.882776949086436e-06,
      "loss": 0.5308,
      "step": 5710
    },
    {
      "epoch": 1.171411017816916,
      "grad_norm": 4.28125,
      "learning_rate": 3.866620541452764e-06,
      "loss": 0.516,
      "step": 5720
    },
    {
      "epoch": 1.1734589391767356,
      "grad_norm": 3.125,
      "learning_rate": 3.850476599029394e-06,
      "loss": 0.4846,
      "step": 5730
    },
    {
      "epoch": 1.1755068605365553,
      "grad_norm": 4.0625,
      "learning_rate": 3.834345299371719e-06,
      "loss": 0.481,
      "step": 5740
    },
    {
      "epoch": 1.1775547818963752,
      "grad_norm": 4.0,
      "learning_rate": 3.818226819896089e-06,
      "loss": 0.4901,
      "step": 5750
    },
    {
      "epoch": 1.1796027032561949,
      "grad_norm": 3.953125,
      "learning_rate": 3.802121337877848e-06,
      "loss": 0.5132,
      "step": 5760
    },
    {
      "epoch": 1.1816506246160148,
      "grad_norm": 4.15625,
      "learning_rate": 3.7860290304493953e-06,
      "loss": 0.4929,
      "step": 5770
    },
    {
      "epoch": 1.1836985459758345,
      "grad_norm": 4.84375,
      "learning_rate": 3.7699500745982287e-06,
      "loss": 0.5142,
      "step": 5780
    },
    {
      "epoch": 1.1857464673356544,
      "grad_norm": 3.5,
      "learning_rate": 3.7538846471650038e-06,
      "loss": 0.5093,
      "step": 5790
    },
    {
      "epoch": 1.187794388695474,
      "grad_norm": 2.859375,
      "learning_rate": 3.737832924841587e-06,
      "loss": 0.4884,
      "step": 5800
    },
    {
      "epoch": 1.189842310055294,
      "grad_norm": 3.625,
      "learning_rate": 3.72179508416911e-06,
      "loss": 0.4842,
      "step": 5810
    },
    {
      "epoch": 1.1918902314151136,
      "grad_norm": 4.09375,
      "learning_rate": 3.7057713015360365e-06,
      "loss": 0.505,
      "step": 5820
    },
    {
      "epoch": 1.1939381527749335,
      "grad_norm": 4.0625,
      "learning_rate": 3.6897617531762086e-06,
      "loss": 0.5513,
      "step": 5830
    },
    {
      "epoch": 1.1959860741347532,
      "grad_norm": 3.46875,
      "learning_rate": 3.6737666151669206e-06,
      "loss": 0.5029,
      "step": 5840
    },
    {
      "epoch": 1.1980339954945731,
      "grad_norm": 4.53125,
      "learning_rate": 3.657786063426977e-06,
      "loss": 0.4956,
      "step": 5850
    },
    {
      "epoch": 1.2000819168543928,
      "grad_norm": 3.1875,
      "learning_rate": 3.6418202737147566e-06,
      "loss": 0.4914,
      "step": 5860
    },
    {
      "epoch": 1.2021298382142125,
      "grad_norm": 3.671875,
      "learning_rate": 3.6258694216262845e-06,
      "loss": 0.5057,
      "step": 5870
    },
    {
      "epoch": 1.2041777595740324,
      "grad_norm": 3.796875,
      "learning_rate": 3.609933682593299e-06,
      "loss": 0.4909,
      "step": 5880
    },
    {
      "epoch": 1.206225680933852,
      "grad_norm": 3.609375,
      "learning_rate": 3.5940132318813137e-06,
      "loss": 0.5094,
      "step": 5890
    },
    {
      "epoch": 1.208273602293672,
      "grad_norm": 3.328125,
      "learning_rate": 3.578108244587705e-06,
      "loss": 0.5162,
      "step": 5900
    },
    {
      "epoch": 1.2103215236534917,
      "grad_norm": 3.625,
      "learning_rate": 3.562218895639775e-06,
      "loss": 0.5171,
      "step": 5910
    },
    {
      "epoch": 1.2123694450133116,
      "grad_norm": 3.875,
      "learning_rate": 3.5463453597928306e-06,
      "loss": 0.5411,
      "step": 5920
    },
    {
      "epoch": 1.2144173663731312,
      "grad_norm": 3.96875,
      "learning_rate": 3.5304878116282654e-06,
      "loss": 0.4972,
      "step": 5930
    },
    {
      "epoch": 1.2164652877329512,
      "grad_norm": 4.375,
      "learning_rate": 3.51464642555163e-06,
      "loss": 0.5189,
      "step": 5940
    },
    {
      "epoch": 1.2185132090927708,
      "grad_norm": 3.78125,
      "learning_rate": 3.498821375790728e-06,
      "loss": 0.5105,
      "step": 5950
    },
    {
      "epoch": 1.2205611304525905,
      "grad_norm": 4.59375,
      "learning_rate": 3.4830128363936835e-06,
      "loss": 0.4979,
      "step": 5960
    },
    {
      "epoch": 1.2226090518124104,
      "grad_norm": 4.65625,
      "learning_rate": 3.467220981227042e-06,
      "loss": 0.5311,
      "step": 5970
    },
    {
      "epoch": 1.22465697317223,
      "grad_norm": 3.359375,
      "learning_rate": 3.451445983973848e-06,
      "loss": 0.4564,
      "step": 5980
    },
    {
      "epoch": 1.22670489453205,
      "grad_norm": 4.125,
      "learning_rate": 3.4356880181317377e-06,
      "loss": 0.5113,
      "step": 5990
    },
    {
      "epoch": 1.2287528158918697,
      "grad_norm": 4.1875,
      "learning_rate": 3.419947257011036e-06,
      "loss": 0.5245,
      "step": 6000
    },
    {
      "epoch": 1.2287528158918697,
      "eval_loss": 0.5531573295593262,
      "eval_runtime": 48.5342,
      "eval_samples_per_second": 84.724,
      "eval_steps_per_second": 42.362,
      "step": 6000
    },
    {
      "epoch": 1.2308007372516896,
      "grad_norm": 4.34375,
      "learning_rate": 3.404223873732837e-06,
      "loss": 0.5338,
      "step": 6010
    },
    {
      "epoch": 1.2328486586115093,
      "grad_norm": 3.90625,
      "learning_rate": 3.3885180412271183e-06,
      "loss": 0.5234,
      "step": 6020
    },
    {
      "epoch": 1.2348965799713292,
      "grad_norm": 3.328125,
      "learning_rate": 3.372829932230821e-06,
      "loss": 0.5337,
      "step": 6030
    },
    {
      "epoch": 1.2369445013311489,
      "grad_norm": 3.96875,
      "learning_rate": 3.357159719285964e-06,
      "loss": 0.5321,
      "step": 6040
    },
    {
      "epoch": 1.2389924226909685,
      "grad_norm": 4.15625,
      "learning_rate": 3.3415075747377414e-06,
      "loss": 0.5118,
      "step": 6050
    },
    {
      "epoch": 1.2410403440507884,
      "grad_norm": 3.453125,
      "learning_rate": 3.325873670732619e-06,
      "loss": 0.5142,
      "step": 6060
    },
    {
      "epoch": 1.2430882654106084,
      "grad_norm": 4.1875,
      "learning_rate": 3.3102581792164566e-06,
      "loss": 0.4996,
      "step": 6070
    },
    {
      "epoch": 1.245136186770428,
      "grad_norm": 4.9375,
      "learning_rate": 3.294661271932601e-06,
      "loss": 0.5165,
      "step": 6080
    },
    {
      "epoch": 1.2471841081302477,
      "grad_norm": 4.5,
      "learning_rate": 3.2790831204200113e-06,
      "loss": 0.5101,
      "step": 6090
    },
    {
      "epoch": 1.2492320294900676,
      "grad_norm": 3.21875,
      "learning_rate": 3.2635238960113634e-06,
      "loss": 0.5147,
      "step": 6100
    },
    {
      "epoch": 1.2512799508498873,
      "grad_norm": 3.75,
      "learning_rate": 3.2479837698311646e-06,
      "loss": 0.5224,
      "step": 6110
    },
    {
      "epoch": 1.2533278722097072,
      "grad_norm": 3.375,
      "learning_rate": 3.23246291279388e-06,
      "loss": 0.4954,
      "step": 6120
    },
    {
      "epoch": 1.2553757935695269,
      "grad_norm": 4.3125,
      "learning_rate": 3.2169614956020423e-06,
      "loss": 0.4957,
      "step": 6130
    },
    {
      "epoch": 1.2574237149293468,
      "grad_norm": 3.390625,
      "learning_rate": 3.2014796887443854e-06,
      "loss": 0.4907,
      "step": 6140
    },
    {
      "epoch": 1.2594716362891665,
      "grad_norm": 3.46875,
      "learning_rate": 3.1860176624939566e-06,
      "loss": 0.4758,
      "step": 6150
    },
    {
      "epoch": 1.2615195576489864,
      "grad_norm": 3.65625,
      "learning_rate": 3.1705755869062553e-06,
      "loss": 0.4639,
      "step": 6160
    },
    {
      "epoch": 1.263567479008806,
      "grad_norm": 3.6875,
      "learning_rate": 3.1551536318173613e-06,
      "loss": 0.5144,
      "step": 6170
    },
    {
      "epoch": 1.2656154003686257,
      "grad_norm": 4.25,
      "learning_rate": 3.139751966842054e-06,
      "loss": 0.5526,
      "step": 6180
    },
    {
      "epoch": 1.2676633217284456,
      "grad_norm": 4.625,
      "learning_rate": 3.124370761371968e-06,
      "loss": 0.4946,
      "step": 6190
    },
    {
      "epoch": 1.2697112430882653,
      "grad_norm": 3.984375,
      "learning_rate": 3.1090101845737084e-06,
      "loss": 0.4808,
      "step": 6200
    },
    {
      "epoch": 1.2717591644480852,
      "grad_norm": 3.875,
      "learning_rate": 3.0936704053870083e-06,
      "loss": 0.5139,
      "step": 6210
    },
    {
      "epoch": 1.273807085807905,
      "grad_norm": 3.625,
      "learning_rate": 3.0783515925228626e-06,
      "loss": 0.4892,
      "step": 6220
    },
    {
      "epoch": 1.2758550071677248,
      "grad_norm": 4.53125,
      "learning_rate": 3.0630539144616646e-06,
      "loss": 0.4363,
      "step": 6230
    },
    {
      "epoch": 1.2779029285275445,
      "grad_norm": 3.828125,
      "learning_rate": 3.0477775394513743e-06,
      "loss": 0.4816,
      "step": 6240
    },
    {
      "epoch": 1.2799508498873644,
      "grad_norm": 4.09375,
      "learning_rate": 3.0325226355056425e-06,
      "loss": 0.5106,
      "step": 6250
    },
    {
      "epoch": 1.281998771247184,
      "grad_norm": 3.9375,
      "learning_rate": 3.0172893704019846e-06,
      "loss": 0.4663,
      "step": 6260
    },
    {
      "epoch": 1.2840466926070038,
      "grad_norm": 3.90625,
      "learning_rate": 3.0020779116799236e-06,
      "loss": 0.5006,
      "step": 6270
    },
    {
      "epoch": 1.2860946139668237,
      "grad_norm": 3.28125,
      "learning_rate": 2.9868884266391464e-06,
      "loss": 0.511,
      "step": 6280
    },
    {
      "epoch": 1.2881425353266436,
      "grad_norm": 4.65625,
      "learning_rate": 2.9717210823376742e-06,
      "loss": 0.4877,
      "step": 6290
    },
    {
      "epoch": 1.2901904566864633,
      "grad_norm": 3.90625,
      "learning_rate": 2.9565760455900106e-06,
      "loss": 0.4974,
      "step": 6300
    },
    {
      "epoch": 1.292238378046283,
      "grad_norm": 3.09375,
      "learning_rate": 2.941453482965323e-06,
      "loss": 0.5354,
      "step": 6310
    },
    {
      "epoch": 1.2942862994061028,
      "grad_norm": 4.125,
      "learning_rate": 2.926353560785594e-06,
      "loss": 0.5125,
      "step": 6320
    },
    {
      "epoch": 1.2963342207659225,
      "grad_norm": 3.828125,
      "learning_rate": 2.9112764451238074e-06,
      "loss": 0.5341,
      "step": 6330
    },
    {
      "epoch": 1.2983821421257424,
      "grad_norm": 3.609375,
      "learning_rate": 2.8962223018021116e-06,
      "loss": 0.4633,
      "step": 6340
    },
    {
      "epoch": 1.3004300634855621,
      "grad_norm": 3.828125,
      "learning_rate": 2.8811912963900013e-06,
      "loss": 0.4744,
      "step": 6350
    },
    {
      "epoch": 1.302477984845382,
      "grad_norm": 3.9375,
      "learning_rate": 2.8661835942024915e-06,
      "loss": 0.5308,
      "step": 6360
    },
    {
      "epoch": 1.3045259062052017,
      "grad_norm": 3.65625,
      "learning_rate": 2.8511993602983023e-06,
      "loss": 0.5155,
      "step": 6370
    },
    {
      "epoch": 1.3065738275650216,
      "grad_norm": 3.578125,
      "learning_rate": 2.836238759478045e-06,
      "loss": 0.4424,
      "step": 6380
    },
    {
      "epoch": 1.3086217489248413,
      "grad_norm": 3.984375,
      "learning_rate": 2.821301956282408e-06,
      "loss": 0.5507,
      "step": 6390
    },
    {
      "epoch": 1.310669670284661,
      "grad_norm": 3.234375,
      "learning_rate": 2.806389114990345e-06,
      "loss": 0.5209,
      "step": 6400
    },
    {
      "epoch": 1.3127175916444809,
      "grad_norm": 4.0625,
      "learning_rate": 2.7915003996172724e-06,
      "loss": 0.5002,
      "step": 6410
    },
    {
      "epoch": 1.3147655130043006,
      "grad_norm": 3.46875,
      "learning_rate": 2.776635973913262e-06,
      "loss": 0.4994,
      "step": 6420
    },
    {
      "epoch": 1.3168134343641205,
      "grad_norm": 3.109375,
      "learning_rate": 2.761796001361241e-06,
      "loss": 0.4868,
      "step": 6430
    },
    {
      "epoch": 1.3188613557239401,
      "grad_norm": 3.703125,
      "learning_rate": 2.746980645175199e-06,
      "loss": 0.4633,
      "step": 6440
    },
    {
      "epoch": 1.32090927708376,
      "grad_norm": 3.9375,
      "learning_rate": 2.732190068298378e-06,
      "loss": 0.5256,
      "step": 6450
    },
    {
      "epoch": 1.3229571984435797,
      "grad_norm": 3.75,
      "learning_rate": 2.7174244334015025e-06,
      "loss": 0.5183,
      "step": 6460
    },
    {
      "epoch": 1.3250051198033996,
      "grad_norm": 4.15625,
      "learning_rate": 2.7026839028809704e-06,
      "loss": 0.5399,
      "step": 6470
    },
    {
      "epoch": 1.3270530411632193,
      "grad_norm": 4.125,
      "learning_rate": 2.6879686388570782e-06,
      "loss": 0.5054,
      "step": 6480
    },
    {
      "epoch": 1.329100962523039,
      "grad_norm": 3.984375,
      "learning_rate": 2.6732788031722325e-06,
      "loss": 0.4731,
      "step": 6490
    },
    {
      "epoch": 1.331148883882859,
      "grad_norm": 4.59375,
      "learning_rate": 2.658614557389174e-06,
      "loss": 0.4896,
      "step": 6500
    },
    {
      "epoch": 1.3331968052426788,
      "grad_norm": 3.796875,
      "learning_rate": 2.6439760627892e-06,
      "loss": 0.5294,
      "step": 6510
    },
    {
      "epoch": 1.3352447266024985,
      "grad_norm": 3.421875,
      "learning_rate": 2.6293634803703853e-06,
      "loss": 0.5028,
      "step": 6520
    },
    {
      "epoch": 1.3372926479623182,
      "grad_norm": 3.96875,
      "learning_rate": 2.61477697084582e-06,
      "loss": 0.4852,
      "step": 6530
    },
    {
      "epoch": 1.339340569322138,
      "grad_norm": 4.34375,
      "learning_rate": 2.600216694641835e-06,
      "loss": 0.4643,
      "step": 6540
    },
    {
      "epoch": 1.3413884906819578,
      "grad_norm": 4.15625,
      "learning_rate": 2.5856828118962385e-06,
      "loss": 0.4863,
      "step": 6550
    },
    {
      "epoch": 1.3434364120417777,
      "grad_norm": 3.453125,
      "learning_rate": 2.571175482456565e-06,
      "loss": 0.5265,
      "step": 6560
    },
    {
      "epoch": 1.3454843334015973,
      "grad_norm": 3.453125,
      "learning_rate": 2.5566948658782947e-06,
      "loss": 0.512,
      "step": 6570
    },
    {
      "epoch": 1.347532254761417,
      "grad_norm": 3.65625,
      "learning_rate": 2.5422411214231272e-06,
      "loss": 0.453,
      "step": 6580
    },
    {
      "epoch": 1.349580176121237,
      "grad_norm": 4.25,
      "learning_rate": 2.5278144080572013e-06,
      "loss": 0.4624,
      "step": 6590
    },
    {
      "epoch": 1.3516280974810568,
      "grad_norm": 3.84375,
      "learning_rate": 2.513414884449373e-06,
      "loss": 0.5362,
      "step": 6600
    },
    {
      "epoch": 1.3536760188408765,
      "grad_norm": 3.6875,
      "learning_rate": 2.4990427089694517e-06,
      "loss": 0.466,
      "step": 6610
    },
    {
      "epoch": 1.3557239402006962,
      "grad_norm": 3.828125,
      "learning_rate": 2.484698039686465e-06,
      "loss": 0.513,
      "step": 6620
    },
    {
      "epoch": 1.357771861560516,
      "grad_norm": 4.0625,
      "learning_rate": 2.4703810343669204e-06,
      "loss": 0.4675,
      "step": 6630
    },
    {
      "epoch": 1.3598197829203358,
      "grad_norm": 4.28125,
      "learning_rate": 2.4560918504730712e-06,
      "loss": 0.461,
      "step": 6640
    },
    {
      "epoch": 1.3618677042801557,
      "grad_norm": 3.6875,
      "learning_rate": 2.4418306451611816e-06,
      "loss": 0.4595,
      "step": 6650
    },
    {
      "epoch": 1.3639156256399754,
      "grad_norm": 3.875,
      "learning_rate": 2.427597575279801e-06,
      "loss": 0.5385,
      "step": 6660
    },
    {
      "epoch": 1.3659635469997953,
      "grad_norm": 4.03125,
      "learning_rate": 2.413392797368034e-06,
      "loss": 0.4778,
      "step": 6670
    },
    {
      "epoch": 1.368011468359615,
      "grad_norm": 4.9375,
      "learning_rate": 2.3992164676538336e-06,
      "loss": 0.5586,
      "step": 6680
    },
    {
      "epoch": 1.3700593897194349,
      "grad_norm": 3.671875,
      "learning_rate": 2.385068742052255e-06,
      "loss": 0.4678,
      "step": 6690
    },
    {
      "epoch": 1.3721073110792545,
      "grad_norm": 3.921875,
      "learning_rate": 2.3709497761637744e-06,
      "loss": 0.5287,
      "step": 6700
    },
    {
      "epoch": 1.3741552324390742,
      "grad_norm": 4.59375,
      "learning_rate": 2.3568597252725466e-06,
      "loss": 0.5866,
      "step": 6710
    },
    {
      "epoch": 1.3762031537988941,
      "grad_norm": 3.75,
      "learning_rate": 2.3427987443447237e-06,
      "loss": 0.5044,
      "step": 6720
    },
    {
      "epoch": 1.378251075158714,
      "grad_norm": 4.15625,
      "learning_rate": 2.3287669880267317e-06,
      "loss": 0.5611,
      "step": 6730
    },
    {
      "epoch": 1.3802989965185337,
      "grad_norm": 3.65625,
      "learning_rate": 2.3147646106435773e-06,
      "loss": 0.4826,
      "step": 6740
    },
    {
      "epoch": 1.3823469178783534,
      "grad_norm": 3.5,
      "learning_rate": 2.300791766197151e-06,
      "loss": 0.4959,
      "step": 6750
    },
    {
      "epoch": 1.3843948392381733,
      "grad_norm": 3.84375,
      "learning_rate": 2.2868486083645325e-06,
      "loss": 0.4908,
      "step": 6760
    },
    {
      "epoch": 1.386442760597993,
      "grad_norm": 3.75,
      "learning_rate": 2.272935290496297e-06,
      "loss": 0.5306,
      "step": 6770
    },
    {
      "epoch": 1.3884906819578129,
      "grad_norm": 4.0625,
      "learning_rate": 2.259051965614839e-06,
      "loss": 0.5183,
      "step": 6780
    },
    {
      "epoch": 1.3905386033176326,
      "grad_norm": 4.25,
      "learning_rate": 2.2451987864126712e-06,
      "loss": 0.5608,
      "step": 6790
    },
    {
      "epoch": 1.3925865246774523,
      "grad_norm": 4.40625,
      "learning_rate": 2.231375905250769e-06,
      "loss": 0.5404,
      "step": 6800
    },
    {
      "epoch": 1.3946344460372722,
      "grad_norm": 4.375,
      "learning_rate": 2.2175834741568677e-06,
      "loss": 0.5083,
      "step": 6810
    },
    {
      "epoch": 1.396682367397092,
      "grad_norm": 3.265625,
      "learning_rate": 2.2038216448238185e-06,
      "loss": 0.4536,
      "step": 6820
    },
    {
      "epoch": 1.3987302887569117,
      "grad_norm": 3.328125,
      "learning_rate": 2.1900905686078945e-06,
      "loss": 0.4988,
      "step": 6830
    },
    {
      "epoch": 1.4007782101167314,
      "grad_norm": 3.421875,
      "learning_rate": 2.1763903965271466e-06,
      "loss": 0.5205,
      "step": 6840
    },
    {
      "epoch": 1.4028261314765513,
      "grad_norm": 4.3125,
      "learning_rate": 2.162721279259729e-06,
      "loss": 0.5292,
      "step": 6850
    },
    {
      "epoch": 1.404874052836371,
      "grad_norm": 3.765625,
      "learning_rate": 2.1490833671422484e-06,
      "loss": 0.484,
      "step": 6860
    },
    {
      "epoch": 1.406921974196191,
      "grad_norm": 3.8125,
      "learning_rate": 2.1354768101681077e-06,
      "loss": 0.462,
      "step": 6870
    },
    {
      "epoch": 1.4089698955560106,
      "grad_norm": 3.75,
      "learning_rate": 2.1219017579858586e-06,
      "loss": 0.5473,
      "step": 6880
    },
    {
      "epoch": 1.4110178169158305,
      "grad_norm": 4.25,
      "learning_rate": 2.108358359897553e-06,
      "loss": 0.4682,
      "step": 6890
    },
    {
      "epoch": 1.4130657382756502,
      "grad_norm": 3.765625,
      "learning_rate": 2.0948467648571085e-06,
      "loss": 0.516,
      "step": 6900
    },
    {
      "epoch": 1.41511365963547,
      "grad_norm": 4.375,
      "learning_rate": 2.0813671214686533e-06,
      "loss": 0.5238,
      "step": 6910
    },
    {
      "epoch": 1.4171615809952898,
      "grad_norm": 4.15625,
      "learning_rate": 2.067919577984916e-06,
      "loss": 0.4904,
      "step": 6920
    },
    {
      "epoch": 1.4192095023551095,
      "grad_norm": 4.375,
      "learning_rate": 2.054504282305569e-06,
      "loss": 0.543,
      "step": 6930
    },
    {
      "epoch": 1.4212574237149294,
      "grad_norm": 3.78125,
      "learning_rate": 2.041121381975624e-06,
      "loss": 0.4894,
      "step": 6940
    },
    {
      "epoch": 1.4233053450747493,
      "grad_norm": 3.25,
      "learning_rate": 2.027771024183798e-06,
      "loss": 0.4765,
      "step": 6950
    },
    {
      "epoch": 1.425353266434569,
      "grad_norm": 3.28125,
      "learning_rate": 2.0144533557608925e-06,
      "loss": 0.4883,
      "step": 6960
    },
    {
      "epoch": 1.4274011877943886,
      "grad_norm": 3.609375,
      "learning_rate": 2.0011685231781876e-06,
      "loss": 0.5183,
      "step": 6970
    },
    {
      "epoch": 1.4294491091542085,
      "grad_norm": 3.9375,
      "learning_rate": 1.987916672545822e-06,
      "loss": 0.5087,
      "step": 6980
    },
    {
      "epoch": 1.4314970305140282,
      "grad_norm": 3.984375,
      "learning_rate": 1.9746979496111936e-06,
      "loss": 0.4799,
      "step": 6990
    },
    {
      "epoch": 1.4335449518738481,
      "grad_norm": 4.78125,
      "learning_rate": 1.96151249975735e-06,
      "loss": 0.4663,
      "step": 7000
    },
    {
      "epoch": 1.4335449518738481,
      "eval_loss": 0.5521626472473145,
      "eval_runtime": 50.1391,
      "eval_samples_per_second": 82.012,
      "eval_steps_per_second": 41.006,
      "step": 7000
    },
    {
      "epoch": 1.4355928732336678,
      "grad_norm": 4.03125,
      "learning_rate": 1.948360468001393e-06,
      "loss": 0.5628,
      "step": 7010
    },
    {
      "epoch": 1.4376407945934875,
      "grad_norm": 4.375,
      "learning_rate": 1.935241998992889e-06,
      "loss": 0.4884,
      "step": 7020
    },
    {
      "epoch": 1.4396887159533074,
      "grad_norm": 4.1875,
      "learning_rate": 1.9221572370122606e-06,
      "loss": 0.5088,
      "step": 7030
    },
    {
      "epoch": 1.4417366373131273,
      "grad_norm": 3.90625,
      "learning_rate": 1.9091063259692255e-06,
      "loss": 0.4645,
      "step": 7040
    },
    {
      "epoch": 1.443784558672947,
      "grad_norm": 4.0625,
      "learning_rate": 1.896089409401185e-06,
      "loss": 0.5207,
      "step": 7050
    },
    {
      "epoch": 1.4458324800327667,
      "grad_norm": 5.15625,
      "learning_rate": 1.8831066304716738e-06,
      "loss": 0.5466,
      "step": 7060
    },
    {
      "epoch": 1.4478804013925866,
      "grad_norm": 4.125,
      "learning_rate": 1.8701581319687634e-06,
      "loss": 0.5313,
      "step": 7070
    },
    {
      "epoch": 1.4499283227524062,
      "grad_norm": 4.1875,
      "learning_rate": 1.8572440563035016e-06,
      "loss": 0.523,
      "step": 7080
    },
    {
      "epoch": 1.4519762441122261,
      "grad_norm": 3.9375,
      "learning_rate": 1.8443645455083465e-06,
      "loss": 0.4837,
      "step": 7090
    },
    {
      "epoch": 1.4540241654720458,
      "grad_norm": 3.765625,
      "learning_rate": 1.8315197412356006e-06,
      "loss": 0.4952,
      "step": 7100
    },
    {
      "epoch": 1.4560720868318657,
      "grad_norm": 3.515625,
      "learning_rate": 1.8187097847558532e-06,
      "loss": 0.4919,
      "step": 7110
    },
    {
      "epoch": 1.4581200081916854,
      "grad_norm": 4.09375,
      "learning_rate": 1.8059348169564366e-06,
      "loss": 0.4318,
      "step": 7120
    },
    {
      "epoch": 1.4601679295515053,
      "grad_norm": 3.828125,
      "learning_rate": 1.793194978339855e-06,
      "loss": 0.4651,
      "step": 7130
    },
    {
      "epoch": 1.462215850911325,
      "grad_norm": 3.5625,
      "learning_rate": 1.7804904090222664e-06,
      "loss": 0.4786,
      "step": 7140
    },
    {
      "epoch": 1.4642637722711447,
      "grad_norm": 3.890625,
      "learning_rate": 1.767821248731913e-06,
      "loss": 0.4668,
      "step": 7150
    },
    {
      "epoch": 1.4663116936309646,
      "grad_norm": 4.125,
      "learning_rate": 1.7551876368076154e-06,
      "loss": 0.5105,
      "step": 7160
    },
    {
      "epoch": 1.4683596149907845,
      "grad_norm": 4.40625,
      "learning_rate": 1.7425897121972068e-06,
      "loss": 0.501,
      "step": 7170
    },
    {
      "epoch": 1.4704075363506042,
      "grad_norm": 3.09375,
      "learning_rate": 1.7300276134560367e-06,
      "loss": 0.5251,
      "step": 7180
    },
    {
      "epoch": 1.4724554577104239,
      "grad_norm": 4.15625,
      "learning_rate": 1.717501478745423e-06,
      "loss": 0.5332,
      "step": 7190
    },
    {
      "epoch": 1.4745033790702438,
      "grad_norm": 4.28125,
      "learning_rate": 1.7050114458311446e-06,
      "loss": 0.4896,
      "step": 7200
    },
    {
      "epoch": 1.4765513004300634,
      "grad_norm": 4.5625,
      "learning_rate": 1.6925576520819225e-06,
      "loss": 0.5556,
      "step": 7210
    },
    {
      "epoch": 1.4785992217898833,
      "grad_norm": 4.625,
      "learning_rate": 1.6801402344679102e-06,
      "loss": 0.5056,
      "step": 7220
    },
    {
      "epoch": 1.480647143149703,
      "grad_norm": 3.296875,
      "learning_rate": 1.6677593295591848e-06,
      "loss": 0.5639,
      "step": 7230
    },
    {
      "epoch": 1.4826950645095227,
      "grad_norm": 4.28125,
      "learning_rate": 1.6554150735242535e-06,
      "loss": 0.5681,
      "step": 7240
    },
    {
      "epoch": 1.4847429858693426,
      "grad_norm": 4.4375,
      "learning_rate": 1.6431076021285381e-06,
      "loss": 0.4984,
      "step": 7250
    },
    {
      "epoch": 1.4867909072291625,
      "grad_norm": 3.46875,
      "learning_rate": 1.6308370507329057e-06,
      "loss": 0.453,
      "step": 7260
    },
    {
      "epoch": 1.4888388285889822,
      "grad_norm": 4.40625,
      "learning_rate": 1.6186035542921546e-06,
      "loss": 0.5102,
      "step": 7270
    },
    {
      "epoch": 1.4908867499488019,
      "grad_norm": 4.09375,
      "learning_rate": 1.6064072473535546e-06,
      "loss": 0.499,
      "step": 7280
    },
    {
      "epoch": 1.4929346713086218,
      "grad_norm": 4.65625,
      "learning_rate": 1.5942482640553486e-06,
      "loss": 0.5314,
      "step": 7290
    },
    {
      "epoch": 1.4949825926684415,
      "grad_norm": 3.84375,
      "learning_rate": 1.5821267381252858e-06,
      "loss": 0.4847,
      "step": 7300
    },
    {
      "epoch": 1.4970305140282614,
      "grad_norm": 3.40625,
      "learning_rate": 1.570042802879148e-06,
      "loss": 0.5141,
      "step": 7310
    },
    {
      "epoch": 1.499078435388081,
      "grad_norm": 4.46875,
      "learning_rate": 1.5579965912192873e-06,
      "loss": 0.4914,
      "step": 7320
    },
    {
      "epoch": 1.5011263567479007,
      "grad_norm": 3.453125,
      "learning_rate": 1.5459882356331596e-06,
      "loss": 0.5131,
      "step": 7330
    },
    {
      "epoch": 1.5031742781077206,
      "grad_norm": 3.21875,
      "learning_rate": 1.53401786819187e-06,
      "loss": 0.4868,
      "step": 7340
    },
    {
      "epoch": 1.5052221994675405,
      "grad_norm": 3.6875,
      "learning_rate": 1.5220856205487183e-06,
      "loss": 0.485,
      "step": 7350
    },
    {
      "epoch": 1.5072701208273602,
      "grad_norm": 3.890625,
      "learning_rate": 1.510191623937759e-06,
      "loss": 0.4816,
      "step": 7360
    },
    {
      "epoch": 1.50931804218718,
      "grad_norm": 3.546875,
      "learning_rate": 1.498336009172341e-06,
      "loss": 0.5506,
      "step": 7370
    },
    {
      "epoch": 1.5113659635469998,
      "grad_norm": 4.1875,
      "learning_rate": 1.4865189066436909e-06,
      "loss": 0.4692,
      "step": 7380
    },
    {
      "epoch": 1.5134138849068197,
      "grad_norm": 4.5,
      "learning_rate": 1.4747404463194553e-06,
      "loss": 0.5319,
      "step": 7390
    },
    {
      "epoch": 1.5154618062666394,
      "grad_norm": 3.546875,
      "learning_rate": 1.4630007577422949e-06,
      "loss": 0.4945,
      "step": 7400
    },
    {
      "epoch": 1.517509727626459,
      "grad_norm": 3.375,
      "learning_rate": 1.451299970028442e-06,
      "loss": 0.5442,
      "step": 7410
    },
    {
      "epoch": 1.5195576489862788,
      "grad_norm": 4.3125,
      "learning_rate": 1.4396382118662877e-06,
      "loss": 0.4956,
      "step": 7420
    },
    {
      "epoch": 1.5216055703460987,
      "grad_norm": 4.3125,
      "learning_rate": 1.4280156115149667e-06,
      "loss": 0.531,
      "step": 7430
    },
    {
      "epoch": 1.5236534917059186,
      "grad_norm": 3.671875,
      "learning_rate": 1.4164322968029442e-06,
      "loss": 0.4683,
      "step": 7440
    },
    {
      "epoch": 1.5257014130657383,
      "grad_norm": 3.5,
      "learning_rate": 1.4048883951266135e-06,
      "loss": 0.5077,
      "step": 7450
    },
    {
      "epoch": 1.527749334425558,
      "grad_norm": 3.765625,
      "learning_rate": 1.3933840334488903e-06,
      "loss": 0.4806,
      "step": 7460
    },
    {
      "epoch": 1.5297972557853778,
      "grad_norm": 4.71875,
      "learning_rate": 1.3819193382978207e-06,
      "loss": 0.452,
      "step": 7470
    },
    {
      "epoch": 1.5318451771451977,
      "grad_norm": 3.703125,
      "learning_rate": 1.3704944357651872e-06,
      "loss": 0.4668,
      "step": 7480
    },
    {
      "epoch": 1.5338930985050174,
      "grad_norm": 3.78125,
      "learning_rate": 1.3591094515051223e-06,
      "loss": 0.4809,
      "step": 7490
    },
    {
      "epoch": 1.535941019864837,
      "grad_norm": 3.6875,
      "learning_rate": 1.347764510732727e-06,
      "loss": 0.5493,
      "step": 7500
    },
    {
      "epoch": 1.537988941224657,
      "grad_norm": 3.375,
      "learning_rate": 1.3364597382226908e-06,
      "loss": 0.5105,
      "step": 7510
    },
    {
      "epoch": 1.540036862584477,
      "grad_norm": 3.90625,
      "learning_rate": 1.325195258307928e-06,
      "loss": 0.5163,
      "step": 7520
    },
    {
      "epoch": 1.5420847839442966,
      "grad_norm": 3.375,
      "learning_rate": 1.313971194878198e-06,
      "loss": 0.495,
      "step": 7530
    },
    {
      "epoch": 1.5441327053041163,
      "grad_norm": 3.890625,
      "learning_rate": 1.3027876713787524e-06,
      "loss": 0.4762,
      "step": 7540
    },
    {
      "epoch": 1.546180626663936,
      "grad_norm": 3.78125,
      "learning_rate": 1.2916448108089713e-06,
      "loss": 0.4488,
      "step": 7550
    },
    {
      "epoch": 1.5482285480237559,
      "grad_norm": 3.59375,
      "learning_rate": 1.2805427357210154e-06,
      "loss": 0.4543,
      "step": 7560
    },
    {
      "epoch": 1.5502764693835758,
      "grad_norm": 3.890625,
      "learning_rate": 1.2694815682184747e-06,
      "loss": 0.5251,
      "step": 7570
    },
    {
      "epoch": 1.5523243907433955,
      "grad_norm": 4.34375,
      "learning_rate": 1.2584614299550257e-06,
      "loss": 0.5251,
      "step": 7580
    },
    {
      "epoch": 1.5543723121032151,
      "grad_norm": 3.359375,
      "learning_rate": 1.2474824421330962e-06,
      "loss": 0.4684,
      "step": 7590
    },
    {
      "epoch": 1.556420233463035,
      "grad_norm": 4.15625,
      "learning_rate": 1.2365447255025293e-06,
      "loss": 0.4835,
      "step": 7600
    },
    {
      "epoch": 1.558468154822855,
      "grad_norm": 4.40625,
      "learning_rate": 1.2256484003592572e-06,
      "loss": 0.4605,
      "step": 7610
    },
    {
      "epoch": 1.5605160761826746,
      "grad_norm": 4.3125,
      "learning_rate": 1.2147935865439747e-06,
      "loss": 0.4669,
      "step": 7620
    },
    {
      "epoch": 1.5625639975424943,
      "grad_norm": 3.703125,
      "learning_rate": 1.2039804034408315e-06,
      "loss": 0.4901,
      "step": 7630
    },
    {
      "epoch": 1.564611918902314,
      "grad_norm": 3.484375,
      "learning_rate": 1.1932089699760995e-06,
      "loss": 0.5078,
      "step": 7640
    },
    {
      "epoch": 1.566659840262134,
      "grad_norm": 4.28125,
      "learning_rate": 1.1824794046168892e-06,
      "loss": 0.4499,
      "step": 7650
    },
    {
      "epoch": 1.5687077616219538,
      "grad_norm": 4.71875,
      "learning_rate": 1.1717918253698273e-06,
      "loss": 0.5427,
      "step": 7660
    },
    {
      "epoch": 1.5707556829817735,
      "grad_norm": 3.296875,
      "learning_rate": 1.1611463497797676e-06,
      "loss": 0.4713,
      "step": 7670
    },
    {
      "epoch": 1.5728036043415932,
      "grad_norm": 4.15625,
      "learning_rate": 1.1505430949284962e-06,
      "loss": 0.511,
      "step": 7680
    },
    {
      "epoch": 1.574851525701413,
      "grad_norm": 4.03125,
      "learning_rate": 1.1399821774334457e-06,
      "loss": 0.5032,
      "step": 7690
    },
    {
      "epoch": 1.576899447061233,
      "grad_norm": 3.890625,
      "learning_rate": 1.1294637134464104e-06,
      "loss": 0.4683,
      "step": 7700
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 4.21875,
      "learning_rate": 1.1189878186522684e-06,
      "loss": 0.4678,
      "step": 7710
    },
    {
      "epoch": 1.5809952897808723,
      "grad_norm": 4.40625,
      "learning_rate": 1.1085546082677123e-06,
      "loss": 0.4456,
      "step": 7720
    },
    {
      "epoch": 1.5830432111406922,
      "grad_norm": 4.0625,
      "learning_rate": 1.0981641970399786e-06,
      "loss": 0.4764,
      "step": 7730
    },
    {
      "epoch": 1.5850911325005121,
      "grad_norm": 4.03125,
      "learning_rate": 1.0878166992455874e-06,
      "loss": 0.4918,
      "step": 7740
    },
    {
      "epoch": 1.5871390538603318,
      "grad_norm": 3.65625,
      "learning_rate": 1.0775122286890894e-06,
      "loss": 0.4753,
      "step": 7750
    },
    {
      "epoch": 1.5891869752201515,
      "grad_norm": 3.96875,
      "learning_rate": 1.0672508987018016e-06,
      "loss": 0.4841,
      "step": 7760
    },
    {
      "epoch": 1.5912348965799712,
      "grad_norm": 4.21875,
      "learning_rate": 1.0570328221405796e-06,
      "loss": 0.5167,
      "step": 7770
    },
    {
      "epoch": 1.593282817939791,
      "grad_norm": 3.578125,
      "learning_rate": 1.046858111386556e-06,
      "loss": 0.5181,
      "step": 7780
    },
    {
      "epoch": 1.595330739299611,
      "grad_norm": 3.828125,
      "learning_rate": 1.0367268783439249e-06,
      "loss": 0.4854,
      "step": 7790
    },
    {
      "epoch": 1.5973786606594307,
      "grad_norm": 3.46875,
      "learning_rate": 1.0266392344386939e-06,
      "loss": 0.4486,
      "step": 7800
    },
    {
      "epoch": 1.5994265820192504,
      "grad_norm": 4.5,
      "learning_rate": 1.0165952906174675e-06,
      "loss": 0.4954,
      "step": 7810
    },
    {
      "epoch": 1.6014745033790703,
      "grad_norm": 4.03125,
      "learning_rate": 1.006595157346225e-06,
      "loss": 0.4548,
      "step": 7820
    },
    {
      "epoch": 1.6035224247388902,
      "grad_norm": 3.875,
      "learning_rate": 9.966389446091068e-07,
      "loss": 0.5134,
      "step": 7830
    },
    {
      "epoch": 1.6055703460987099,
      "grad_norm": 3.703125,
      "learning_rate": 9.867267619072013e-07,
      "loss": 0.4977,
      "step": 7840
    },
    {
      "epoch": 1.6076182674585295,
      "grad_norm": 4.09375,
      "learning_rate": 9.768587182573442e-07,
      "loss": 0.4621,
      "step": 7850
    },
    {
      "epoch": 1.6096661888183492,
      "grad_norm": 3.859375,
      "learning_rate": 9.67034922190917e-07,
      "loss": 0.5124,
      "step": 7860
    },
    {
      "epoch": 1.6117141101781691,
      "grad_norm": 4.65625,
      "learning_rate": 9.572554817526592e-07,
      "loss": 0.5,
      "step": 7870
    },
    {
      "epoch": 1.613762031537989,
      "grad_norm": 4.15625,
      "learning_rate": 9.475205044994651e-07,
      "loss": 0.44,
      "step": 7880
    },
    {
      "epoch": 1.6158099528978087,
      "grad_norm": 3.453125,
      "learning_rate": 9.378300974992238e-07,
      "loss": 0.5015,
      "step": 7890
    },
    {
      "epoch": 1.6178578742576284,
      "grad_norm": 3.8125,
      "learning_rate": 9.281843673296165e-07,
      "loss": 0.486,
      "step": 7900
    },
    {
      "epoch": 1.6199057956174483,
      "grad_norm": 3.5625,
      "learning_rate": 9.185834200769662e-07,
      "loss": 0.4988,
      "step": 7910
    },
    {
      "epoch": 1.6219537169772682,
      "grad_norm": 3.765625,
      "learning_rate": 9.090273613350564e-07,
      "loss": 0.5192,
      "step": 7920
    },
    {
      "epoch": 1.6240016383370879,
      "grad_norm": 4.0,
      "learning_rate": 8.995162962039761e-07,
      "loss": 0.512,
      "step": 7930
    },
    {
      "epoch": 1.6260495596969076,
      "grad_norm": 4.1875,
      "learning_rate": 8.900503292889628e-07,
      "loss": 0.5336,
      "step": 7940
    },
    {
      "epoch": 1.6280974810567275,
      "grad_norm": 3.828125,
      "learning_rate": 8.806295646992508e-07,
      "loss": 0.4872,
      "step": 7950
    },
    {
      "epoch": 1.6301454024165472,
      "grad_norm": 3.28125,
      "learning_rate": 8.712541060469271e-07,
      "loss": 0.5295,
      "step": 7960
    },
    {
      "epoch": 1.632193323776367,
      "grad_norm": 4.03125,
      "learning_rate": 8.61924056445796e-07,
      "loss": 0.4899,
      "step": 7970
    },
    {
      "epoch": 1.6342412451361867,
      "grad_norm": 3.5625,
      "learning_rate": 8.526395185102321e-07,
      "loss": 0.4615,
      "step": 7980
    },
    {
      "epoch": 1.6362891664960064,
      "grad_norm": 3.734375,
      "learning_rate": 8.434005943540718e-07,
      "loss": 0.4868,
      "step": 7990
    },
    {
      "epoch": 1.6383370878558263,
      "grad_norm": 4.3125,
      "learning_rate": 8.342073855894673e-07,
      "loss": 0.5378,
      "step": 8000
    },
    {
      "epoch": 1.6383370878558263,
      "eval_loss": 0.5521250367164612,
      "eval_runtime": 50.5238,
      "eval_samples_per_second": 81.387,
      "eval_steps_per_second": 40.694,
      "step": 8000
    },
    {
      "epoch": 1.6403850092156462,
      "grad_norm": 4.03125,
      "learning_rate": 8.250599933257919e-07,
      "loss": 0.4721,
      "step": 8010
    },
    {
      "epoch": 1.642432930575466,
      "grad_norm": 3.71875,
      "learning_rate": 8.15958518168506e-07,
      "loss": 0.4715,
      "step": 8020
    },
    {
      "epoch": 1.6444808519352856,
      "grad_norm": 3.859375,
      "learning_rate": 8.069030602180705e-07,
      "loss": 0.5022,
      "step": 8030
    },
    {
      "epoch": 1.6465287732951055,
      "grad_norm": 4.34375,
      "learning_rate": 7.978937190688318e-07,
      "loss": 0.5166,
      "step": 8040
    },
    {
      "epoch": 1.6485766946549254,
      "grad_norm": 3.796875,
      "learning_rate": 7.889305938079328e-07,
      "loss": 0.5324,
      "step": 8050
    },
    {
      "epoch": 1.650624616014745,
      "grad_norm": 3.078125,
      "learning_rate": 7.800137830142212e-07,
      "loss": 0.4688,
      "step": 8060
    },
    {
      "epoch": 1.6526725373745648,
      "grad_norm": 4.03125,
      "learning_rate": 7.711433847571664e-07,
      "loss": 0.5302,
      "step": 8070
    },
    {
      "epoch": 1.6547204587343844,
      "grad_norm": 3.8125,
      "learning_rate": 7.623194965957786e-07,
      "loss": 0.4856,
      "step": 8080
    },
    {
      "epoch": 1.6567683800942044,
      "grad_norm": 3.984375,
      "learning_rate": 7.535422155775423e-07,
      "loss": 0.5046,
      "step": 8090
    },
    {
      "epoch": 1.6588163014540243,
      "grad_norm": 3.921875,
      "learning_rate": 7.44811638237336e-07,
      "loss": 0.5231,
      "step": 8100
    },
    {
      "epoch": 1.660864222813844,
      "grad_norm": 3.421875,
      "learning_rate": 7.361278605963884e-07,
      "loss": 0.5012,
      "step": 8110
    },
    {
      "epoch": 1.6629121441736636,
      "grad_norm": 3.703125,
      "learning_rate": 7.274909781612033e-07,
      "loss": 0.4458,
      "step": 8120
    },
    {
      "epoch": 1.6649600655334835,
      "grad_norm": 4.15625,
      "learning_rate": 7.18901085922526e-07,
      "loss": 0.504,
      "step": 8130
    },
    {
      "epoch": 1.6670079868933034,
      "grad_norm": 3.53125,
      "learning_rate": 7.103582783542867e-07,
      "loss": 0.4965,
      "step": 8140
    },
    {
      "epoch": 1.669055908253123,
      "grad_norm": 3.96875,
      "learning_rate": 7.018626494125674e-07,
      "loss": 0.5263,
      "step": 8150
    },
    {
      "epoch": 1.6711038296129428,
      "grad_norm": 4.53125,
      "learning_rate": 6.934142925345661e-07,
      "loss": 0.5062,
      "step": 8160
    },
    {
      "epoch": 1.6731517509727627,
      "grad_norm": 3.078125,
      "learning_rate": 6.850133006375704e-07,
      "loss": 0.4966,
      "step": 8170
    },
    {
      "epoch": 1.6751996723325824,
      "grad_norm": 4.21875,
      "learning_rate": 6.766597661179352e-07,
      "loss": 0.4877,
      "step": 8180
    },
    {
      "epoch": 1.6772475936924023,
      "grad_norm": 4.1875,
      "learning_rate": 6.683537808500673e-07,
      "loss": 0.5252,
      "step": 8190
    },
    {
      "epoch": 1.679295515052222,
      "grad_norm": 3.984375,
      "learning_rate": 6.600954361854117e-07,
      "loss": 0.5023,
      "step": 8200
    },
    {
      "epoch": 1.6813434364120416,
      "grad_norm": 4.65625,
      "learning_rate": 6.518848229514541e-07,
      "loss": 0.5164,
      "step": 8210
    },
    {
      "epoch": 1.6833913577718616,
      "grad_norm": 3.921875,
      "learning_rate": 6.437220314507098e-07,
      "loss": 0.4991,
      "step": 8220
    },
    {
      "epoch": 1.6854392791316815,
      "grad_norm": 3.828125,
      "learning_rate": 6.356071514597467e-07,
      "loss": 0.4688,
      "step": 8230
    },
    {
      "epoch": 1.6874872004915011,
      "grad_norm": 4.3125,
      "learning_rate": 6.275402722281798e-07,
      "loss": 0.5088,
      "step": 8240
    },
    {
      "epoch": 1.6895351218513208,
      "grad_norm": 3.703125,
      "learning_rate": 6.195214824777068e-07,
      "loss": 0.496,
      "step": 8250
    },
    {
      "epoch": 1.6915830432111407,
      "grad_norm": 3.875,
      "learning_rate": 6.115508704011208e-07,
      "loss": 0.5199,
      "step": 8260
    },
    {
      "epoch": 1.6936309645709606,
      "grad_norm": 3.296875,
      "learning_rate": 6.036285236613437e-07,
      "loss": 0.4849,
      "step": 8270
    },
    {
      "epoch": 1.6956788859307803,
      "grad_norm": 3.453125,
      "learning_rate": 5.957545293904632e-07,
      "loss": 0.489,
      "step": 8280
    },
    {
      "epoch": 1.6977268072906,
      "grad_norm": 4.25,
      "learning_rate": 5.879289741887739e-07,
      "loss": 0.4736,
      "step": 8290
    },
    {
      "epoch": 1.6997747286504197,
      "grad_norm": 4.65625,
      "learning_rate": 5.801519441238229e-07,
      "loss": 0.4806,
      "step": 8300
    },
    {
      "epoch": 1.7018226500102396,
      "grad_norm": 3.359375,
      "learning_rate": 5.72423524729469e-07,
      "loss": 0.4315,
      "step": 8310
    },
    {
      "epoch": 1.7038705713700595,
      "grad_norm": 4.03125,
      "learning_rate": 5.647438010049305e-07,
      "loss": 0.5183,
      "step": 8320
    },
    {
      "epoch": 1.7059184927298792,
      "grad_norm": 3.328125,
      "learning_rate": 5.571128574138667e-07,
      "loss": 0.4632,
      "step": 8330
    },
    {
      "epoch": 1.7079664140896988,
      "grad_norm": 4.0,
      "learning_rate": 5.495307778834319e-07,
      "loss": 0.5044,
      "step": 8340
    },
    {
      "epoch": 1.7100143354495188,
      "grad_norm": 3.375,
      "learning_rate": 5.419976458033666e-07,
      "loss": 0.5054,
      "step": 8350
    },
    {
      "epoch": 1.7120622568093387,
      "grad_norm": 3.90625,
      "learning_rate": 5.345135440250687e-07,
      "loss": 0.5582,
      "step": 8360
    },
    {
      "epoch": 1.7141101781691583,
      "grad_norm": 3.9375,
      "learning_rate": 5.270785548606927e-07,
      "loss": 0.5177,
      "step": 8370
    },
    {
      "epoch": 1.716158099528978,
      "grad_norm": 3.921875,
      "learning_rate": 5.196927600822366e-07,
      "loss": 0.5106,
      "step": 8380
    },
    {
      "epoch": 1.718206020888798,
      "grad_norm": 3.40625,
      "learning_rate": 5.123562409206457e-07,
      "loss": 0.4848,
      "step": 8390
    },
    {
      "epoch": 1.7202539422486176,
      "grad_norm": 3.796875,
      "learning_rate": 5.050690780649193e-07,
      "loss": 0.5181,
      "step": 8400
    },
    {
      "epoch": 1.7223018636084375,
      "grad_norm": 4.3125,
      "learning_rate": 4.978313516612232e-07,
      "loss": 0.4963,
      "step": 8410
    },
    {
      "epoch": 1.7243497849682572,
      "grad_norm": 3.1875,
      "learning_rate": 4.906431413120056e-07,
      "loss": 0.5136,
      "step": 8420
    },
    {
      "epoch": 1.7263977063280769,
      "grad_norm": 4.0,
      "learning_rate": 4.835045260751297e-07,
      "loss": 0.5049,
      "step": 8430
    },
    {
      "epoch": 1.7284456276878968,
      "grad_norm": 3.546875,
      "learning_rate": 4.7641558446298974e-07,
      "loss": 0.5492,
      "step": 8440
    },
    {
      "epoch": 1.7304935490477167,
      "grad_norm": 3.921875,
      "learning_rate": 4.6937639444166514e-07,
      "loss": 0.5012,
      "step": 8450
    },
    {
      "epoch": 1.7325414704075364,
      "grad_norm": 3.359375,
      "learning_rate": 4.6238703343004565e-07,
      "loss": 0.4954,
      "step": 8460
    },
    {
      "epoch": 1.734589391767356,
      "grad_norm": 4.1875,
      "learning_rate": 4.5544757829899535e-07,
      "loss": 0.4714,
      "step": 8470
    },
    {
      "epoch": 1.736637313127176,
      "grad_norm": 3.890625,
      "learning_rate": 4.485581053704957e-07,
      "loss": 0.485,
      "step": 8480
    },
    {
      "epoch": 1.7386852344869959,
      "grad_norm": 3.390625,
      "learning_rate": 4.417186904168125e-07,
      "loss": 0.4994,
      "step": 8490
    },
    {
      "epoch": 1.7407331558468155,
      "grad_norm": 3.453125,
      "learning_rate": 4.3492940865966006e-07,
      "loss": 0.4771,
      "step": 8500
    },
    {
      "epoch": 1.7427810772066352,
      "grad_norm": 3.578125,
      "learning_rate": 4.2819033476937386e-07,
      "loss": 0.4589,
      "step": 8510
    },
    {
      "epoch": 1.744828998566455,
      "grad_norm": 4.15625,
      "learning_rate": 4.2150154286409006e-07,
      "loss": 0.5249,
      "step": 8520
    },
    {
      "epoch": 1.7468769199262748,
      "grad_norm": 4.0,
      "learning_rate": 4.148631065089315e-07,
      "loss": 0.5176,
      "step": 8530
    },
    {
      "epoch": 1.7489248412860947,
      "grad_norm": 3.84375,
      "learning_rate": 4.0827509871519455e-07,
      "loss": 0.5037,
      "step": 8540
    },
    {
      "epoch": 1.7509727626459144,
      "grad_norm": 3.953125,
      "learning_rate": 4.0173759193955107e-07,
      "loss": 0.5298,
      "step": 8550
    },
    {
      "epoch": 1.753020684005734,
      "grad_norm": 3.703125,
      "learning_rate": 3.952506580832477e-07,
      "loss": 0.5212,
      "step": 8560
    },
    {
      "epoch": 1.755068605365554,
      "grad_norm": 4.15625,
      "learning_rate": 3.8881436849131925e-07,
      "loss": 0.5017,
      "step": 8570
    },
    {
      "epoch": 1.7571165267253739,
      "grad_norm": 3.265625,
      "learning_rate": 3.8242879395179677e-07,
      "loss": 0.5061,
      "step": 8580
    },
    {
      "epoch": 1.7591644480851936,
      "grad_norm": 3.921875,
      "learning_rate": 3.760940046949385e-07,
      "loss": 0.4965,
      "step": 8590
    },
    {
      "epoch": 1.7612123694450132,
      "grad_norm": 3.96875,
      "learning_rate": 3.69810070392449e-07,
      "loss": 0.5191,
      "step": 8600
    },
    {
      "epoch": 1.763260290804833,
      "grad_norm": 4.1875,
      "learning_rate": 3.63577060156719e-07,
      "loss": 0.5086,
      "step": 8610
    },
    {
      "epoch": 1.7653082121646528,
      "grad_norm": 4.40625,
      "learning_rate": 3.573950425400602e-07,
      "loss": 0.5109,
      "step": 8620
    },
    {
      "epoch": 1.7673561335244727,
      "grad_norm": 4.34375,
      "learning_rate": 3.512640855339561e-07,
      "loss": 0.4675,
      "step": 8630
    },
    {
      "epoch": 1.7694040548842924,
      "grad_norm": 3.828125,
      "learning_rate": 3.4518425656831e-07,
      "loss": 0.5042,
      "step": 8640
    },
    {
      "epoch": 1.771451976244112,
      "grad_norm": 4.09375,
      "learning_rate": 3.391556225107073e-07,
      "loss": 0.5347,
      "step": 8650
    },
    {
      "epoch": 1.773499897603932,
      "grad_norm": 4.15625,
      "learning_rate": 3.3317824966567713e-07,
      "loss": 0.4946,
      "step": 8660
    },
    {
      "epoch": 1.775547818963752,
      "grad_norm": 3.96875,
      "learning_rate": 3.272522037739634e-07,
      "loss": 0.4695,
      "step": 8670
    },
    {
      "epoch": 1.7775957403235716,
      "grad_norm": 4.625,
      "learning_rate": 3.213775500118038e-07,
      "loss": 0.5263,
      "step": 8680
    },
    {
      "epoch": 1.7796436616833913,
      "grad_norm": 4.1875,
      "learning_rate": 3.155543529902111e-07,
      "loss": 0.4979,
      "step": 8690
    },
    {
      "epoch": 1.7816915830432112,
      "grad_norm": 3.390625,
      "learning_rate": 3.0978267675426164e-07,
      "loss": 0.4523,
      "step": 8700
    },
    {
      "epoch": 1.783739504403031,
      "grad_norm": 4.0625,
      "learning_rate": 3.040625847823958e-07,
      "loss": 0.5158,
      "step": 8710
    },
    {
      "epoch": 1.7857874257628508,
      "grad_norm": 4.0625,
      "learning_rate": 2.983941399857138e-07,
      "loss": 0.475,
      "step": 8720
    },
    {
      "epoch": 1.7878353471226704,
      "grad_norm": 3.515625,
      "learning_rate": 2.9277740470728623e-07,
      "loss": 0.5373,
      "step": 8730
    },
    {
      "epoch": 1.7898832684824901,
      "grad_norm": 3.90625,
      "learning_rate": 2.8721244072146995e-07,
      "loss": 0.4803,
      "step": 8740
    },
    {
      "epoch": 1.79193118984231,
      "grad_norm": 3.90625,
      "learning_rate": 2.816993092332265e-07,
      "loss": 0.5106,
      "step": 8750
    },
    {
      "epoch": 1.79397911120213,
      "grad_norm": 3.828125,
      "learning_rate": 2.7623807087745035e-07,
      "loss": 0.5295,
      "step": 8760
    },
    {
      "epoch": 1.7960270325619496,
      "grad_norm": 3.609375,
      "learning_rate": 2.708287857183006e-07,
      "loss": 0.5699,
      "step": 8770
    },
    {
      "epoch": 1.7980749539217693,
      "grad_norm": 4.0,
      "learning_rate": 2.654715132485414e-07,
      "loss": 0.4657,
      "step": 8780
    },
    {
      "epoch": 1.8001228752815892,
      "grad_norm": 3.796875,
      "learning_rate": 2.6016631238888865e-07,
      "loss": 0.4737,
      "step": 8790
    },
    {
      "epoch": 1.8021707966414091,
      "grad_norm": 3.796875,
      "learning_rate": 2.549132414873584e-07,
      "loss": 0.5056,
      "step": 8800
    },
    {
      "epoch": 1.8042187180012288,
      "grad_norm": 3.015625,
      "learning_rate": 2.497123583186295e-07,
      "loss": 0.5065,
      "step": 8810
    },
    {
      "epoch": 1.8062666393610485,
      "grad_norm": 3.6875,
      "learning_rate": 2.4456372008340724e-07,
      "loss": 0.4631,
      "step": 8820
    },
    {
      "epoch": 1.8083145607208682,
      "grad_norm": 3.859375,
      "learning_rate": 2.394673834077882e-07,
      "loss": 0.5392,
      "step": 8830
    },
    {
      "epoch": 1.810362482080688,
      "grad_norm": 3.78125,
      "learning_rate": 2.3442340434264798e-07,
      "loss": 0.4679,
      "step": 8840
    },
    {
      "epoch": 1.812410403440508,
      "grad_norm": 4.5625,
      "learning_rate": 2.2943183836301676e-07,
      "loss": 0.518,
      "step": 8850
    },
    {
      "epoch": 1.8144583248003276,
      "grad_norm": 4.0625,
      "learning_rate": 2.2449274036747072e-07,
      "loss": 0.5188,
      "step": 8860
    },
    {
      "epoch": 1.8165062461601473,
      "grad_norm": 4.0,
      "learning_rate": 2.1960616467753104e-07,
      "loss": 0.4933,
      "step": 8870
    },
    {
      "epoch": 1.8185541675199672,
      "grad_norm": 3.84375,
      "learning_rate": 2.1477216503706267e-07,
      "loss": 0.5042,
      "step": 8880
    },
    {
      "epoch": 1.8206020888797871,
      "grad_norm": 3.8125,
      "learning_rate": 2.0999079461168692e-07,
      "loss": 0.4993,
      "step": 8890
    },
    {
      "epoch": 1.8226500102396068,
      "grad_norm": 4.15625,
      "learning_rate": 2.0526210598819373e-07,
      "loss": 0.5079,
      "step": 8900
    },
    {
      "epoch": 1.8246979315994265,
      "grad_norm": 3.578125,
      "learning_rate": 2.0058615117396486e-07,
      "loss": 0.49,
      "step": 8910
    },
    {
      "epoch": 1.8267458529592464,
      "grad_norm": 4.65625,
      "learning_rate": 1.9596298159640149e-07,
      "loss": 0.5116,
      "step": 8920
    },
    {
      "epoch": 1.8287937743190663,
      "grad_norm": 3.796875,
      "learning_rate": 1.9139264810235757e-07,
      "loss": 0.5092,
      "step": 8930
    },
    {
      "epoch": 1.830841695678886,
      "grad_norm": 3.859375,
      "learning_rate": 1.8687520095758583e-07,
      "loss": 0.5138,
      "step": 8940
    },
    {
      "epoch": 1.8328896170387057,
      "grad_norm": 3.828125,
      "learning_rate": 1.8241068984617415e-07,
      "loss": 0.5163,
      "step": 8950
    },
    {
      "epoch": 1.8349375383985254,
      "grad_norm": 3.921875,
      "learning_rate": 1.7799916387001183e-07,
      "loss": 0.4741,
      "step": 8960
    },
    {
      "epoch": 1.8369854597583453,
      "grad_norm": 3.484375,
      "learning_rate": 1.736406715482397e-07,
      "loss": 0.4613,
      "step": 8970
    },
    {
      "epoch": 1.8390333811181652,
      "grad_norm": 4.09375,
      "learning_rate": 1.6933526081672202e-07,
      "loss": 0.5445,
      "step": 8980
    },
    {
      "epoch": 1.8410813024779848,
      "grad_norm": 4.28125,
      "learning_rate": 1.650829790275177e-07,
      "loss": 0.4681,
      "step": 8990
    },
    {
      "epoch": 1.8431292238378045,
      "grad_norm": 4.65625,
      "learning_rate": 1.608838729483575e-07,
      "loss": 0.5008,
      "step": 9000
    },
    {
      "epoch": 1.8431292238378045,
      "eval_loss": 0.5520676374435425,
      "eval_runtime": 50.2055,
      "eval_samples_per_second": 81.903,
      "eval_steps_per_second": 40.952,
      "step": 9000
    },
    {
      "epoch": 1.8451771451976244,
      "grad_norm": 4.375,
      "learning_rate": 1.5673798876213398e-07,
      "loss": 0.4875,
      "step": 9010
    },
    {
      "epoch": 1.8472250665574443,
      "grad_norm": 4.25,
      "learning_rate": 1.5264537206638896e-07,
      "loss": 0.4904,
      "step": 9020
    },
    {
      "epoch": 1.849272987917264,
      "grad_norm": 3.234375,
      "learning_rate": 1.4860606787281518e-07,
      "loss": 0.5164,
      "step": 9030
    },
    {
      "epoch": 1.8513209092770837,
      "grad_norm": 4.34375,
      "learning_rate": 1.4462012060676045e-07,
      "loss": 0.4678,
      "step": 9040
    },
    {
      "epoch": 1.8533688306369034,
      "grad_norm": 3.796875,
      "learning_rate": 1.4068757410673762e-07,
      "loss": 0.4999,
      "step": 9050
    },
    {
      "epoch": 1.8554167519967233,
      "grad_norm": 3.375,
      "learning_rate": 1.368084716239465e-07,
      "loss": 0.4937,
      "step": 9060
    },
    {
      "epoch": 1.8574646733565432,
      "grad_norm": 4.0625,
      "learning_rate": 1.3298285582179104e-07,
      "loss": 0.5074,
      "step": 9070
    },
    {
      "epoch": 1.8595125947163629,
      "grad_norm": 3.921875,
      "learning_rate": 1.2921076877541794e-07,
      "loss": 0.4735,
      "step": 9080
    },
    {
      "epoch": 1.8615605160761826,
      "grad_norm": 4.75,
      "learning_rate": 1.2549225197124814e-07,
      "loss": 0.4848,
      "step": 9090
    },
    {
      "epoch": 1.8636084374360025,
      "grad_norm": 3.71875,
      "learning_rate": 1.2182734630652448e-07,
      "loss": 0.4821,
      "step": 9100
    },
    {
      "epoch": 1.8656563587958224,
      "grad_norm": 3.484375,
      "learning_rate": 1.1821609208885865e-07,
      "loss": 0.4889,
      "step": 9110
    },
    {
      "epoch": 1.867704280155642,
      "grad_norm": 4.4375,
      "learning_rate": 1.14658529035791e-07,
      "loss": 0.4905,
      "step": 9120
    },
    {
      "epoch": 1.8697522015154617,
      "grad_norm": 3.53125,
      "learning_rate": 1.1115469627434983e-07,
      "loss": 0.4923,
      "step": 9130
    },
    {
      "epoch": 1.8718001228752816,
      "grad_norm": 3.1875,
      "learning_rate": 1.0770463234062556e-07,
      "loss": 0.4689,
      "step": 9140
    },
    {
      "epoch": 1.8738480442351013,
      "grad_norm": 4.03125,
      "learning_rate": 1.0430837517934278e-07,
      "loss": 0.4526,
      "step": 9150
    },
    {
      "epoch": 1.8758959655949212,
      "grad_norm": 3.8125,
      "learning_rate": 1.0096596214344723e-07,
      "loss": 0.5039,
      "step": 9160
    },
    {
      "epoch": 1.877943886954741,
      "grad_norm": 3.796875,
      "learning_rate": 9.767742999368945e-08,
      "loss": 0.5123,
      "step": 9170
    },
    {
      "epoch": 1.8799918083145606,
      "grad_norm": 3.9375,
      "learning_rate": 9.444281489822737e-08,
      "loss": 0.5069,
      "step": 9180
    },
    {
      "epoch": 1.8820397296743805,
      "grad_norm": 3.375,
      "learning_rate": 9.126215243221992e-08,
      "loss": 0.5134,
      "step": 9190
    },
    {
      "epoch": 1.8840876510342004,
      "grad_norm": 3.96875,
      "learning_rate": 8.813547757744568e-08,
      "loss": 0.5314,
      "step": 9200
    },
    {
      "epoch": 1.88613557239402,
      "grad_norm": 3.71875,
      "learning_rate": 8.506282472190819e-08,
      "loss": 0.5024,
      "step": 9210
    },
    {
      "epoch": 1.8881834937538398,
      "grad_norm": 3.90625,
      "learning_rate": 8.204422765946741e-08,
      "loss": 0.4582,
      "step": 9220
    },
    {
      "epoch": 1.8902314151136597,
      "grad_norm": 3.578125,
      "learning_rate": 7.907971958945992e-08,
      "loss": 0.5259,
      "step": 9230
    },
    {
      "epoch": 1.8922793364734796,
      "grad_norm": 4.5625,
      "learning_rate": 7.616933311633823e-08,
      "loss": 0.4899,
      "step": 9240
    },
    {
      "epoch": 1.8943272578332992,
      "grad_norm": 4.40625,
      "learning_rate": 7.331310024931038e-08,
      "loss": 0.4689,
      "step": 9250
    },
    {
      "epoch": 1.896375179193119,
      "grad_norm": 4.625,
      "learning_rate": 7.051105240198975e-08,
      "loss": 0.5188,
      "step": 9260
    },
    {
      "epoch": 1.8984231005529386,
      "grad_norm": 3.796875,
      "learning_rate": 6.776322039204642e-08,
      "loss": 0.4636,
      "step": 9270
    },
    {
      "epoch": 1.9004710219127585,
      "grad_norm": 3.984375,
      "learning_rate": 6.506963444087355e-08,
      "loss": 0.4749,
      "step": 9280
    },
    {
      "epoch": 1.9025189432725784,
      "grad_norm": 4.34375,
      "learning_rate": 6.243032417324768e-08,
      "loss": 0.549,
      "step": 9290
    },
    {
      "epoch": 1.904566864632398,
      "grad_norm": 3.953125,
      "learning_rate": 5.984531861701004e-08,
      "loss": 0.4987,
      "step": 9300
    },
    {
      "epoch": 1.9066147859922178,
      "grad_norm": 4.0,
      "learning_rate": 5.7314646202742405e-08,
      "loss": 0.5011,
      "step": 9310
    },
    {
      "epoch": 1.9086627073520377,
      "grad_norm": 3.671875,
      "learning_rate": 5.483833476345624e-08,
      "loss": 0.5187,
      "step": 9320
    },
    {
      "epoch": 1.9107106287118576,
      "grad_norm": 4.0625,
      "learning_rate": 5.241641153428734e-08,
      "loss": 0.5182,
      "step": 9330
    },
    {
      "epoch": 1.9127585500716773,
      "grad_norm": 4.09375,
      "learning_rate": 5.004890315219446e-08,
      "loss": 0.5202,
      "step": 9340
    },
    {
      "epoch": 1.914806471431497,
      "grad_norm": 3.53125,
      "learning_rate": 4.7735835655667864e-08,
      "loss": 0.5006,
      "step": 9350
    },
    {
      "epoch": 1.9168543927913169,
      "grad_norm": 4.5625,
      "learning_rate": 4.547723448444286e-08,
      "loss": 0.4896,
      "step": 9360
    },
    {
      "epoch": 1.9189023141511365,
      "grad_norm": 4.5,
      "learning_rate": 4.3273124479218386e-08,
      "loss": 0.5277,
      "step": 9370
    },
    {
      "epoch": 1.9209502355109564,
      "grad_norm": 4.4375,
      "learning_rate": 4.112352988138557e-08,
      "loss": 0.4983,
      "step": 9380
    },
    {
      "epoch": 1.9229981568707761,
      "grad_norm": 4.40625,
      "learning_rate": 3.902847433276014e-08,
      "loss": 0.5211,
      "step": 9390
    },
    {
      "epoch": 1.9250460782305958,
      "grad_norm": 3.6875,
      "learning_rate": 3.698798087532485e-08,
      "loss": 0.4465,
      "step": 9400
    },
    {
      "epoch": 1.9270939995904157,
      "grad_norm": 3.09375,
      "learning_rate": 3.500207195096972e-08,
      "loss": 0.4819,
      "step": 9410
    },
    {
      "epoch": 1.9291419209502356,
      "grad_norm": 3.875,
      "learning_rate": 3.3070769401254424e-08,
      "loss": 0.5247,
      "step": 9420
    },
    {
      "epoch": 1.9311898423100553,
      "grad_norm": 4.125,
      "learning_rate": 3.1194094467159044e-08,
      "loss": 0.5271,
      "step": 9430
    },
    {
      "epoch": 1.933237763669875,
      "grad_norm": 3.546875,
      "learning_rate": 2.9372067788857594e-08,
      "loss": 0.4789,
      "step": 9440
    },
    {
      "epoch": 1.935285685029695,
      "grad_norm": 4.4375,
      "learning_rate": 2.760470940548543e-08,
      "loss": 0.4851,
      "step": 9450
    },
    {
      "epoch": 1.9373336063895148,
      "grad_norm": 4.8125,
      "learning_rate": 2.58920387549233e-08,
      "loss": 0.4704,
      "step": 9460
    },
    {
      "epoch": 1.9393815277493345,
      "grad_norm": 3.328125,
      "learning_rate": 2.4234074673580853e-08,
      "loss": 0.5505,
      "step": 9470
    },
    {
      "epoch": 1.9414294491091542,
      "grad_norm": 4.125,
      "learning_rate": 2.2630835396190155e-08,
      "loss": 0.5198,
      "step": 9480
    },
    {
      "epoch": 1.9434773704689738,
      "grad_norm": 3.203125,
      "learning_rate": 2.1082338555605265e-08,
      "loss": 0.4826,
      "step": 9490
    },
    {
      "epoch": 1.9455252918287937,
      "grad_norm": 3.578125,
      "learning_rate": 1.9588601182608524e-08,
      "loss": 0.5109,
      "step": 9500
    },
    {
      "epoch": 1.9475732131886137,
      "grad_norm": 3.6875,
      "learning_rate": 1.814963970572292e-08,
      "loss": 0.5035,
      "step": 9510
    },
    {
      "epoch": 1.9496211345484333,
      "grad_norm": 3.859375,
      "learning_rate": 1.676546995103223e-08,
      "loss": 0.513,
      "step": 9520
    },
    {
      "epoch": 1.951669055908253,
      "grad_norm": 3.734375,
      "learning_rate": 1.5436107142003943e-08,
      "loss": 0.5197,
      "step": 9530
    },
    {
      "epoch": 1.953716977268073,
      "grad_norm": 3.203125,
      "learning_rate": 1.4161565899327157e-08,
      "loss": 0.4731,
      "step": 9540
    },
    {
      "epoch": 1.9557648986278928,
      "grad_norm": 3.953125,
      "learning_rate": 1.2941860240746617e-08,
      "loss": 0.4463,
      "step": 9550
    },
    {
      "epoch": 1.9578128199877125,
      "grad_norm": 3.640625,
      "learning_rate": 1.1777003580911161e-08,
      "loss": 0.5008,
      "step": 9560
    },
    {
      "epoch": 1.9598607413475322,
      "grad_norm": 3.71875,
      "learning_rate": 1.0667008731225504e-08,
      "loss": 0.4966,
      "step": 9570
    },
    {
      "epoch": 1.961908662707352,
      "grad_norm": 3.953125,
      "learning_rate": 9.611887899710349e-09,
      "loss": 0.5175,
      "step": 9580
    },
    {
      "epoch": 1.9639565840671718,
      "grad_norm": 3.71875,
      "learning_rate": 8.61165269086639e-09,
      "loss": 0.479,
      "step": 9590
    },
    {
      "epoch": 1.9660045054269917,
      "grad_norm": 3.765625,
      "learning_rate": 7.666314105547745e-09,
      "loss": 0.5428,
      "step": 9600
    },
    {
      "epoch": 1.9680524267868114,
      "grad_norm": 4.3125,
      "learning_rate": 6.775882540841494e-09,
      "loss": 0.4741,
      "step": 9610
    },
    {
      "epoch": 1.970100348146631,
      "grad_norm": 4.03125,
      "learning_rate": 5.940367789951107e-09,
      "loss": 0.4811,
      "step": 9620
    },
    {
      "epoch": 1.972148269506451,
      "grad_norm": 3.609375,
      "learning_rate": 5.159779042092083e-09,
      "loss": 0.54,
      "step": 9630
    },
    {
      "epoch": 1.9741961908662709,
      "grad_norm": 4.0,
      "learning_rate": 4.434124882388146e-09,
      "loss": 0.4908,
      "step": 9640
    },
    {
      "epoch": 1.9762441122260905,
      "grad_norm": 3.546875,
      "learning_rate": 3.763413291776874e-09,
      "loss": 0.5134,
      "step": 9650
    },
    {
      "epoch": 1.9782920335859102,
      "grad_norm": 4.09375,
      "learning_rate": 3.1476516469247655e-09,
      "loss": 0.4972,
      "step": 9660
    },
    {
      "epoch": 1.9803399549457301,
      "grad_norm": 3.828125,
      "learning_rate": 2.586846720141756e-09,
      "loss": 0.5337,
      "step": 9670
    },
    {
      "epoch": 1.98238787630555,
      "grad_norm": 3.625,
      "learning_rate": 2.0810046793118266e-09,
      "loss": 0.5048,
      "step": 9680
    },
    {
      "epoch": 1.9844357976653697,
      "grad_norm": 3.515625,
      "learning_rate": 1.6301310878197307e-09,
      "loss": 0.472,
      "step": 9690
    },
    {
      "epoch": 1.9864837190251894,
      "grad_norm": 3.921875,
      "learning_rate": 1.2342309044943712e-09,
      "loss": 0.4877,
      "step": 9700
    },
    {
      "epoch": 1.988531640385009,
      "grad_norm": 3.46875,
      "learning_rate": 8.933084835521799e-10,
      "loss": 0.4863,
      "step": 9710
    },
    {
      "epoch": 1.990579561744829,
      "grad_norm": 4.03125,
      "learning_rate": 6.073675745482677e-10,
      "loss": 0.4992,
      "step": 9720
    },
    {
      "epoch": 1.9926274831046489,
      "grad_norm": 3.53125,
      "learning_rate": 3.764113223375665e-10,
      "loss": 0.5157,
      "step": 9730
    },
    {
      "epoch": 1.9946754044644686,
      "grad_norm": 4.3125,
      "learning_rate": 2.0044226703819225e-10,
      "loss": 0.4896,
      "step": 9740
    },
    {
      "epoch": 1.9967233258242882,
      "grad_norm": 3.234375,
      "learning_rate": 7.946234400590947e-11,
      "loss": 0.4628,
      "step": 9750
    },
    {
      "epoch": 1.9987712471841081,
      "grad_norm": 4.125,
      "learning_rate": 1.3472883809706461e-11,
      "loss": 0.4633,
      "step": 9760
    }
  ],
  "logging_steps": 10,
  "max_steps": 9766,
  "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": 6.580184799992873e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}