{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.361963190184049,
  "eval_steps": 500,
  "global_step": 1200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03067484662576687,
      "grad_norm": 2.916747570037842,
      "learning_rate": 1.6326530612244901e-07,
      "loss": 1.2014,
      "step": 5
    },
    {
      "epoch": 0.06134969325153374,
      "grad_norm": 2.8873817920684814,
      "learning_rate": 3.6734693877551025e-07,
      "loss": 1.1968,
      "step": 10
    },
    {
      "epoch": 0.09202453987730061,
      "grad_norm": 2.65090012550354,
      "learning_rate": 5.714285714285715e-07,
      "loss": 1.1791,
      "step": 15
    },
    {
      "epoch": 0.12269938650306748,
      "grad_norm": 2.3519155979156494,
      "learning_rate": 7.755102040816327e-07,
      "loss": 1.1753,
      "step": 20
    },
    {
      "epoch": 0.15337423312883436,
      "grad_norm": 1.5954474210739136,
      "learning_rate": 9.795918367346939e-07,
      "loss": 1.1336,
      "step": 25
    },
    {
      "epoch": 0.18404907975460122,
      "grad_norm": 1.1593321561813354,
      "learning_rate": 1.1836734693877552e-06,
      "loss": 1.1015,
      "step": 30
    },
    {
      "epoch": 0.2147239263803681,
      "grad_norm": 1.082676887512207,
      "learning_rate": 1.3877551020408165e-06,
      "loss": 1.054,
      "step": 35
    },
    {
      "epoch": 0.24539877300613497,
      "grad_norm": 1.2616770267486572,
      "learning_rate": 1.5918367346938775e-06,
      "loss": 1.0152,
      "step": 40
    },
    {
      "epoch": 0.27607361963190186,
      "grad_norm": 0.7849059700965881,
      "learning_rate": 1.7959183673469388e-06,
      "loss": 0.9693,
      "step": 45
    },
    {
      "epoch": 0.3067484662576687,
      "grad_norm": 0.5774770379066467,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.9483,
      "step": 50
    },
    {
      "epoch": 0.3374233128834356,
      "grad_norm": 0.5419191718101501,
      "learning_rate": 2.2040816326530616e-06,
      "loss": 0.928,
      "step": 55
    },
    {
      "epoch": 0.36809815950920244,
      "grad_norm": 0.5010412931442261,
      "learning_rate": 2.4081632653061225e-06,
      "loss": 0.8946,
      "step": 60
    },
    {
      "epoch": 0.3987730061349693,
      "grad_norm": 0.45867446064949036,
      "learning_rate": 2.6122448979591842e-06,
      "loss": 0.8818,
      "step": 65
    },
    {
      "epoch": 0.4294478527607362,
      "grad_norm": 0.41711899638175964,
      "learning_rate": 2.816326530612245e-06,
      "loss": 0.8657,
      "step": 70
    },
    {
      "epoch": 0.4601226993865031,
      "grad_norm": 0.3863469362258911,
      "learning_rate": 3.0204081632653064e-06,
      "loss": 0.8518,
      "step": 75
    },
    {
      "epoch": 0.49079754601226994,
      "grad_norm": 0.3639077842235565,
      "learning_rate": 3.2244897959183672e-06,
      "loss": 0.8378,
      "step": 80
    },
    {
      "epoch": 0.5214723926380368,
      "grad_norm": 0.3680429458618164,
      "learning_rate": 3.428571428571429e-06,
      "loss": 0.8328,
      "step": 85
    },
    {
      "epoch": 0.5521472392638037,
      "grad_norm": 0.3493054211139679,
      "learning_rate": 3.6326530612244903e-06,
      "loss": 0.8278,
      "step": 90
    },
    {
      "epoch": 0.5828220858895705,
      "grad_norm": 0.35972270369529724,
      "learning_rate": 3.836734693877551e-06,
      "loss": 0.8177,
      "step": 95
    },
    {
      "epoch": 0.6134969325153374,
      "grad_norm": 0.3597475588321686,
      "learning_rate": 4.040816326530612e-06,
      "loss": 0.8045,
      "step": 100
    },
    {
      "epoch": 0.6441717791411042,
      "grad_norm": 0.3763820230960846,
      "learning_rate": 4.244897959183674e-06,
      "loss": 0.8003,
      "step": 105
    },
    {
      "epoch": 0.6748466257668712,
      "grad_norm": 0.4080606997013092,
      "learning_rate": 4.448979591836735e-06,
      "loss": 0.7961,
      "step": 110
    },
    {
      "epoch": 0.7055214723926381,
      "grad_norm": 0.3602585196495056,
      "learning_rate": 4.653061224489796e-06,
      "loss": 0.7918,
      "step": 115
    },
    {
      "epoch": 0.7361963190184049,
      "grad_norm": 0.3696184456348419,
      "learning_rate": 4.857142857142858e-06,
      "loss": 0.7875,
      "step": 120
    },
    {
      "epoch": 0.7668711656441718,
      "grad_norm": 0.37600934505462646,
      "learning_rate": 5.061224489795918e-06,
      "loss": 0.7875,
      "step": 125
    },
    {
      "epoch": 0.7975460122699386,
      "grad_norm": 0.38016051054000854,
      "learning_rate": 5.26530612244898e-06,
      "loss": 0.7748,
      "step": 130
    },
    {
      "epoch": 0.8282208588957055,
      "grad_norm": 0.3912647068500519,
      "learning_rate": 5.4693877551020415e-06,
      "loss": 0.772,
      "step": 135
    },
    {
      "epoch": 0.8588957055214724,
      "grad_norm": 0.3811889886856079,
      "learning_rate": 5.673469387755103e-06,
      "loss": 0.7678,
      "step": 140
    },
    {
      "epoch": 0.8895705521472392,
      "grad_norm": 0.38242849707603455,
      "learning_rate": 5.877551020408164e-06,
      "loss": 0.761,
      "step": 145
    },
    {
      "epoch": 0.9202453987730062,
      "grad_norm": 0.4102751910686493,
      "learning_rate": 6.0816326530612245e-06,
      "loss": 0.763,
      "step": 150
    },
    {
      "epoch": 0.950920245398773,
      "grad_norm": 0.38672712445259094,
      "learning_rate": 6.285714285714286e-06,
      "loss": 0.7576,
      "step": 155
    },
    {
      "epoch": 0.9815950920245399,
      "grad_norm": 0.3906507194042206,
      "learning_rate": 6.489795918367348e-06,
      "loss": 0.751,
      "step": 160
    },
    {
      "epoch": 1.0122699386503067,
      "grad_norm": 0.4010821580886841,
      "learning_rate": 6.693877551020409e-06,
      "loss": 0.7451,
      "step": 165
    },
    {
      "epoch": 1.0429447852760736,
      "grad_norm": 0.454095721244812,
      "learning_rate": 6.8979591836734705e-06,
      "loss": 0.7373,
      "step": 170
    },
    {
      "epoch": 1.0736196319018405,
      "grad_norm": 0.44405630230903625,
      "learning_rate": 7.102040816326531e-06,
      "loss": 0.7361,
      "step": 175
    },
    {
      "epoch": 1.1042944785276074,
      "grad_norm": 0.42133402824401855,
      "learning_rate": 7.306122448979592e-06,
      "loss": 0.7309,
      "step": 180
    },
    {
      "epoch": 1.1349693251533743,
      "grad_norm": 0.4115142226219177,
      "learning_rate": 7.5102040816326536e-06,
      "loss": 0.7213,
      "step": 185
    },
    {
      "epoch": 1.165644171779141,
      "grad_norm": 0.48450756072998047,
      "learning_rate": 7.714285714285716e-06,
      "loss": 0.7311,
      "step": 190
    },
    {
      "epoch": 1.196319018404908,
      "grad_norm": 0.44604143500328064,
      "learning_rate": 7.918367346938776e-06,
      "loss": 0.7265,
      "step": 195
    },
    {
      "epoch": 1.2269938650306749,
      "grad_norm": 0.4400920271873474,
      "learning_rate": 8.122448979591837e-06,
      "loss": 0.7231,
      "step": 200
    },
    {
      "epoch": 1.2576687116564418,
      "grad_norm": 0.4529883861541748,
      "learning_rate": 8.326530612244899e-06,
      "loss": 0.7177,
      "step": 205
    },
    {
      "epoch": 1.2883435582822087,
      "grad_norm": 0.4646182954311371,
      "learning_rate": 8.530612244897961e-06,
      "loss": 0.7119,
      "step": 210
    },
    {
      "epoch": 1.3190184049079754,
      "grad_norm": 0.45873209834098816,
      "learning_rate": 8.734693877551021e-06,
      "loss": 0.7131,
      "step": 215
    },
    {
      "epoch": 1.3496932515337423,
      "grad_norm": 0.45401182770729065,
      "learning_rate": 8.938775510204082e-06,
      "loss": 0.7136,
      "step": 220
    },
    {
      "epoch": 1.3803680981595092,
      "grad_norm": 0.4406537711620331,
      "learning_rate": 9.142857142857144e-06,
      "loss": 0.7148,
      "step": 225
    },
    {
      "epoch": 1.4110429447852761,
      "grad_norm": 0.5117276310920715,
      "learning_rate": 9.346938775510204e-06,
      "loss": 0.7112,
      "step": 230
    },
    {
      "epoch": 1.441717791411043,
      "grad_norm": 0.4318268895149231,
      "learning_rate": 9.551020408163266e-06,
      "loss": 0.7065,
      "step": 235
    },
    {
      "epoch": 1.4723926380368098,
      "grad_norm": 0.4682416617870331,
      "learning_rate": 9.755102040816327e-06,
      "loss": 0.7019,
      "step": 240
    },
    {
      "epoch": 1.5030674846625767,
      "grad_norm": 0.48544424772262573,
      "learning_rate": 9.959183673469387e-06,
      "loss": 0.7084,
      "step": 245
    },
    {
      "epoch": 1.5337423312883436,
      "grad_norm": 0.45956820249557495,
      "learning_rate": 9.999918433243253e-06,
      "loss": 0.7045,
      "step": 250
    },
    {
      "epoch": 1.5644171779141103,
      "grad_norm": 0.4829379618167877,
      "learning_rate": 9.999587072854989e-06,
      "loss": 0.6992,
      "step": 255
    },
    {
      "epoch": 1.5950920245398774,
      "grad_norm": 0.4817286431789398,
      "learning_rate": 9.99900083779239e-06,
      "loss": 0.6949,
      "step": 260
    },
    {
      "epoch": 1.6257668711656441,
      "grad_norm": 0.41464245319366455,
      "learning_rate": 9.998159757941219e-06,
      "loss": 0.6943,
      "step": 265
    },
    {
      "epoch": 1.656441717791411,
      "grad_norm": 0.45459651947021484,
      "learning_rate": 9.997063876179007e-06,
      "loss": 0.6978,
      "step": 270
    },
    {
      "epoch": 1.687116564417178,
      "grad_norm": 0.4576292335987091,
      "learning_rate": 9.99571324837287e-06,
      "loss": 0.6944,
      "step": 275
    },
    {
      "epoch": 1.7177914110429446,
      "grad_norm": 0.4427788555622101,
      "learning_rate": 9.994107943376654e-06,
      "loss": 0.6914,
      "step": 280
    },
    {
      "epoch": 1.7484662576687118,
      "grad_norm": 0.4589153528213501,
      "learning_rate": 9.992248043027441e-06,
      "loss": 0.6818,
      "step": 285
    },
    {
      "epoch": 1.7791411042944785,
      "grad_norm": 0.4295940399169922,
      "learning_rate": 9.990133642141359e-06,
      "loss": 0.6902,
      "step": 290
    },
    {
      "epoch": 1.8098159509202454,
      "grad_norm": 0.4209536910057068,
      "learning_rate": 9.987764848508756e-06,
      "loss": 0.6856,
      "step": 295
    },
    {
      "epoch": 1.8404907975460123,
      "grad_norm": 0.4128643870353699,
      "learning_rate": 9.985141782888705e-06,
      "loss": 0.683,
      "step": 300
    },
    {
      "epoch": 1.871165644171779,
      "grad_norm": 0.44537627696990967,
      "learning_rate": 9.982264579002853e-06,
      "loss": 0.6837,
      "step": 305
    },
    {
      "epoch": 1.9018404907975461,
      "grad_norm": 0.40354472398757935,
      "learning_rate": 9.979133383528591e-06,
      "loss": 0.686,
      "step": 310
    },
    {
      "epoch": 1.9325153374233128,
      "grad_norm": 0.43714097142219543,
      "learning_rate": 9.975748356091589e-06,
      "loss": 0.6809,
      "step": 315
    },
    {
      "epoch": 1.9631901840490797,
      "grad_norm": 0.4338328540325165,
      "learning_rate": 9.972109669257645e-06,
      "loss": 0.6787,
      "step": 320
    },
    {
      "epoch": 1.9938650306748467,
      "grad_norm": 0.4852101802825928,
      "learning_rate": 9.968217508523913e-06,
      "loss": 0.6696,
      "step": 325
    },
    {
      "epoch": 2.0245398773006134,
      "grad_norm": 0.49633219838142395,
      "learning_rate": 9.964072072309412e-06,
      "loss": 0.652,
      "step": 330
    },
    {
      "epoch": 2.0552147239263805,
      "grad_norm": 0.48235881328582764,
      "learning_rate": 9.959673571944939e-06,
      "loss": 0.6525,
      "step": 335
    },
    {
      "epoch": 2.085889570552147,
      "grad_norm": 0.47604846954345703,
      "learning_rate": 9.955022231662282e-06,
      "loss": 0.6445,
      "step": 340
    },
    {
      "epoch": 2.116564417177914,
      "grad_norm": 0.4336640238761902,
      "learning_rate": 9.95011828858279e-06,
      "loss": 0.6557,
      "step": 345
    },
    {
      "epoch": 2.147239263803681,
      "grad_norm": 0.4802437722682953,
      "learning_rate": 9.944961992705288e-06,
      "loss": 0.6467,
      "step": 350
    },
    {
      "epoch": 2.1779141104294477,
      "grad_norm": 0.476906955242157,
      "learning_rate": 9.939553606893334e-06,
      "loss": 0.6463,
      "step": 355
    },
    {
      "epoch": 2.208588957055215,
      "grad_norm": 0.49822670221328735,
      "learning_rate": 9.933893406861808e-06,
      "loss": 0.6483,
      "step": 360
    },
    {
      "epoch": 2.2392638036809815,
      "grad_norm": 0.49377647042274475,
      "learning_rate": 9.927981681162873e-06,
      "loss": 0.6413,
      "step": 365
    },
    {
      "epoch": 2.2699386503067487,
      "grad_norm": 0.4360464811325073,
      "learning_rate": 9.921818731171249e-06,
      "loss": 0.6478,
      "step": 370
    },
    {
      "epoch": 2.3006134969325154,
      "grad_norm": 0.5180469155311584,
      "learning_rate": 9.915404871068855e-06,
      "loss": 0.6458,
      "step": 375
    },
    {
      "epoch": 2.331288343558282,
      "grad_norm": 0.5449960231781006,
      "learning_rate": 9.9087404278288e-06,
      "loss": 0.6445,
      "step": 380
    },
    {
      "epoch": 2.361963190184049,
      "grad_norm": 0.4497530460357666,
      "learning_rate": 9.901825741198697e-06,
      "loss": 0.6403,
      "step": 385
    },
    {
      "epoch": 2.392638036809816,
      "grad_norm": 0.4317333698272705,
      "learning_rate": 9.894661163683361e-06,
      "loss": 0.6434,
      "step": 390
    },
    {
      "epoch": 2.4233128834355826,
      "grad_norm": 0.47986018657684326,
      "learning_rate": 9.887247060526827e-06,
      "loss": 0.6422,
      "step": 395
    },
    {
      "epoch": 2.4539877300613497,
      "grad_norm": 0.4332720637321472,
      "learning_rate": 9.879583809693737e-06,
      "loss": 0.6438,
      "step": 400
    },
    {
      "epoch": 2.4846625766871164,
      "grad_norm": 0.4776013493537903,
      "learning_rate": 9.871671801850065e-06,
      "loss": 0.6393,
      "step": 405
    },
    {
      "epoch": 2.5153374233128836,
      "grad_norm": 0.48002681136131287,
      "learning_rate": 9.863511440343206e-06,
      "loss": 0.6385,
      "step": 410
    },
    {
      "epoch": 2.5460122699386503,
      "grad_norm": 0.4508747458457947,
      "learning_rate": 9.855103141181412e-06,
      "loss": 0.6362,
      "step": 415
    },
    {
      "epoch": 2.5766871165644174,
      "grad_norm": 0.5055172443389893,
      "learning_rate": 9.846447333012587e-06,
      "loss": 0.6388,
      "step": 420
    },
    {
      "epoch": 2.607361963190184,
      "grad_norm": 0.4450007975101471,
      "learning_rate": 9.837544457102428e-06,
      "loss": 0.6377,
      "step": 425
    },
    {
      "epoch": 2.638036809815951,
      "grad_norm": 0.4479919970035553,
      "learning_rate": 9.82839496731194e-06,
      "loss": 0.6372,
      "step": 430
    },
    {
      "epoch": 2.668711656441718,
      "grad_norm": 0.45310866832733154,
      "learning_rate": 9.818999330074288e-06,
      "loss": 0.6336,
      "step": 435
    },
    {
      "epoch": 2.6993865030674846,
      "grad_norm": 0.4643765091896057,
      "learning_rate": 9.809358024371025e-06,
      "loss": 0.6363,
      "step": 440
    },
    {
      "epoch": 2.7300613496932513,
      "grad_norm": 0.48741263151168823,
      "learning_rate": 9.799471541707672e-06,
      "loss": 0.6382,
      "step": 445
    },
    {
      "epoch": 2.7607361963190185,
      "grad_norm": 0.45080405473709106,
      "learning_rate": 9.789340386088663e-06,
      "loss": 0.6357,
      "step": 450
    },
    {
      "epoch": 2.791411042944785,
      "grad_norm": 0.5053460597991943,
      "learning_rate": 9.778965073991652e-06,
      "loss": 0.6315,
      "step": 455
    },
    {
      "epoch": 2.8220858895705523,
      "grad_norm": 0.523303747177124,
      "learning_rate": 9.768346134341174e-06,
      "loss": 0.6336,
      "step": 460
    },
    {
      "epoch": 2.852760736196319,
      "grad_norm": 0.4880037307739258,
      "learning_rate": 9.757484108481695e-06,
      "loss": 0.63,
      "step": 465
    },
    {
      "epoch": 2.883435582822086,
      "grad_norm": 0.4551119804382324,
      "learning_rate": 9.74637955015001e-06,
      "loss": 0.6315,
      "step": 470
    },
    {
      "epoch": 2.914110429447853,
      "grad_norm": 0.529259204864502,
      "learning_rate": 9.735033025447e-06,
      "loss": 0.6346,
      "step": 475
    },
    {
      "epoch": 2.9447852760736195,
      "grad_norm": 0.44139307737350464,
      "learning_rate": 9.723445112808802e-06,
      "loss": 0.6284,
      "step": 480
    },
    {
      "epoch": 2.9754601226993866,
      "grad_norm": 0.44775474071502686,
      "learning_rate": 9.71161640297729e-06,
      "loss": 0.6316,
      "step": 485
    },
    {
      "epoch": 3.0061349693251533,
      "grad_norm": 0.5552355051040649,
      "learning_rate": 9.699547498969978e-06,
      "loss": 0.6217,
      "step": 490
    },
    {
      "epoch": 3.03680981595092,
      "grad_norm": 0.5540191531181335,
      "learning_rate": 9.687239016049275e-06,
      "loss": 0.5991,
      "step": 495
    },
    {
      "epoch": 3.067484662576687,
      "grad_norm": 0.5161706209182739,
      "learning_rate": 9.674691581691114e-06,
      "loss": 0.5983,
      "step": 500
    },
    {
      "epoch": 3.098159509202454,
      "grad_norm": 0.564895510673523,
      "learning_rate": 9.661905835552974e-06,
      "loss": 0.599,
      "step": 505
    },
    {
      "epoch": 3.128834355828221,
      "grad_norm": 0.5886032581329346,
      "learning_rate": 9.648882429441258e-06,
      "loss": 0.5958,
      "step": 510
    },
    {
      "epoch": 3.1595092024539877,
      "grad_norm": 0.5232917666435242,
      "learning_rate": 9.635622027278076e-06,
      "loss": 0.5981,
      "step": 515
    },
    {
      "epoch": 3.190184049079755,
      "grad_norm": 0.5367550849914551,
      "learning_rate": 9.622125305067394e-06,
      "loss": 0.6009,
      "step": 520
    },
    {
      "epoch": 3.2208588957055215,
      "grad_norm": 0.5145596861839294,
      "learning_rate": 9.608392950860568e-06,
      "loss": 0.5964,
      "step": 525
    },
    {
      "epoch": 3.2515337423312882,
      "grad_norm": 0.48504817485809326,
      "learning_rate": 9.594425664721275e-06,
      "loss": 0.5954,
      "step": 530
    },
    {
      "epoch": 3.2822085889570554,
      "grad_norm": 0.44789251685142517,
      "learning_rate": 9.580224158689821e-06,
      "loss": 0.5963,
      "step": 535
    },
    {
      "epoch": 3.312883435582822,
      "grad_norm": 0.44385582208633423,
      "learning_rate": 9.565789156746843e-06,
      "loss": 0.5953,
      "step": 540
    },
    {
      "epoch": 3.3435582822085887,
      "grad_norm": 0.46638399362564087,
      "learning_rate": 9.551121394776395e-06,
      "loss": 0.5953,
      "step": 545
    },
    {
      "epoch": 3.374233128834356,
      "grad_norm": 0.47886648774147034,
      "learning_rate": 9.536221620528442e-06,
      "loss": 0.5984,
      "step": 550
    },
    {
      "epoch": 3.4049079754601226,
      "grad_norm": 0.43750545382499695,
      "learning_rate": 9.521090593580737e-06,
      "loss": 0.5966,
      "step": 555
    },
    {
      "epoch": 3.4355828220858897,
      "grad_norm": 0.4883286654949188,
      "learning_rate": 9.505729085300098e-06,
      "loss": 0.5976,
      "step": 560
    },
    {
      "epoch": 3.4662576687116564,
      "grad_norm": 0.5246576070785522,
      "learning_rate": 9.490137878803078e-06,
      "loss": 0.596,
      "step": 565
    },
    {
      "epoch": 3.4969325153374236,
      "grad_norm": 0.47524797916412354,
      "learning_rate": 9.47431776891606e-06,
      "loss": 0.5991,
      "step": 570
    },
    {
      "epoch": 3.5276073619631902,
      "grad_norm": 0.4653734266757965,
      "learning_rate": 9.458269562134717e-06,
      "loss": 0.594,
      "step": 575
    },
    {
      "epoch": 3.558282208588957,
      "grad_norm": 0.5341756343841553,
      "learning_rate": 9.441994076582907e-06,
      "loss": 0.5985,
      "step": 580
    },
    {
      "epoch": 3.588957055214724,
      "grad_norm": 0.5514233708381653,
      "learning_rate": 9.425492141970973e-06,
      "loss": 0.5992,
      "step": 585
    },
    {
      "epoch": 3.6196319018404908,
      "grad_norm": 0.4911518096923828,
      "learning_rate": 9.408764599553429e-06,
      "loss": 0.5935,
      "step": 590
    },
    {
      "epoch": 3.6503067484662575,
      "grad_norm": 0.4620342552661896,
      "learning_rate": 9.391812302086088e-06,
      "loss": 0.5977,
      "step": 595
    },
    {
      "epoch": 3.6809815950920246,
      "grad_norm": 0.44591575860977173,
      "learning_rate": 9.374636113782576e-06,
      "loss": 0.5928,
      "step": 600
    },
    {
      "epoch": 3.7116564417177913,
      "grad_norm": 0.40037840604782104,
      "learning_rate": 9.357236910270292e-06,
      "loss": 0.5883,
      "step": 605
    },
    {
      "epoch": 3.7423312883435584,
      "grad_norm": 0.4316963255405426,
      "learning_rate": 9.339615578545753e-06,
      "loss": 0.5916,
      "step": 610
    },
    {
      "epoch": 3.773006134969325,
      "grad_norm": 0.47416016459465027,
      "learning_rate": 9.321773016929382e-06,
      "loss": 0.5935,
      "step": 615
    },
    {
      "epoch": 3.8036809815950923,
      "grad_norm": 0.486749529838562,
      "learning_rate": 9.30371013501972e-06,
      "loss": 0.5976,
      "step": 620
    },
    {
      "epoch": 3.834355828220859,
      "grad_norm": 0.46578896045684814,
      "learning_rate": 9.285427853647038e-06,
      "loss": 0.5964,
      "step": 625
    },
    {
      "epoch": 3.8650306748466257,
      "grad_norm": 0.4229313135147095,
      "learning_rate": 9.26692710482641e-06,
      "loss": 0.5941,
      "step": 630
    },
    {
      "epoch": 3.895705521472393,
      "grad_norm": 0.45090341567993164,
      "learning_rate": 9.248208831710195e-06,
      "loss": 0.5922,
      "step": 635
    },
    {
      "epoch": 3.9263803680981595,
      "grad_norm": 0.4940469264984131,
      "learning_rate": 9.229273988539951e-06,
      "loss": 0.5957,
      "step": 640
    },
    {
      "epoch": 3.957055214723926,
      "grad_norm": 0.45583370327949524,
      "learning_rate": 9.210123540597792e-06,
      "loss": 0.592,
      "step": 645
    },
    {
      "epoch": 3.9877300613496933,
      "grad_norm": 0.43879297375679016,
      "learning_rate": 9.190758464157184e-06,
      "loss": 0.5936,
      "step": 650
    },
    {
      "epoch": 4.0184049079754605,
      "grad_norm": 0.5458897948265076,
      "learning_rate": 9.171179746433164e-06,
      "loss": 0.5735,
      "step": 655
    },
    {
      "epoch": 4.049079754601227,
      "grad_norm": 0.5021305084228516,
      "learning_rate": 9.151388385532022e-06,
      "loss": 0.5571,
      "step": 660
    },
    {
      "epoch": 4.079754601226994,
      "grad_norm": 0.47301429510116577,
      "learning_rate": 9.131385390400417e-06,
      "loss": 0.5594,
      "step": 665
    },
    {
      "epoch": 4.110429447852761,
      "grad_norm": 0.46195101737976074,
      "learning_rate": 9.111171780773938e-06,
      "loss": 0.556,
      "step": 670
    },
    {
      "epoch": 4.141104294478527,
      "grad_norm": 0.4663168489933014,
      "learning_rate": 9.090748587125118e-06,
      "loss": 0.5576,
      "step": 675
    },
    {
      "epoch": 4.171779141104294,
      "grad_norm": 0.5041781067848206,
      "learning_rate": 9.070116850610911e-06,
      "loss": 0.5546,
      "step": 680
    },
    {
      "epoch": 4.2024539877300615,
      "grad_norm": 0.571372389793396,
      "learning_rate": 9.049277623019603e-06,
      "loss": 0.5609,
      "step": 685
    },
    {
      "epoch": 4.233128834355828,
      "grad_norm": 0.4924178421497345,
      "learning_rate": 9.0282319667172e-06,
      "loss": 0.5569,
      "step": 690
    },
    {
      "epoch": 4.263803680981595,
      "grad_norm": 0.4700438678264618,
      "learning_rate": 9.006980954593262e-06,
      "loss": 0.5622,
      "step": 695
    },
    {
      "epoch": 4.294478527607362,
      "grad_norm": 0.46645787358283997,
      "learning_rate": 8.985525670006225e-06,
      "loss": 0.5561,
      "step": 700
    },
    {
      "epoch": 4.325153374233129,
      "grad_norm": 0.47866857051849365,
      "learning_rate": 8.963867206728147e-06,
      "loss": 0.5594,
      "step": 705
    },
    {
      "epoch": 4.355828220858895,
      "grad_norm": 0.4613681733608246,
      "learning_rate": 8.942006668888972e-06,
      "loss": 0.5556,
      "step": 710
    },
    {
      "epoch": 4.386503067484663,
      "grad_norm": 0.451261043548584,
      "learning_rate": 8.919945170920224e-06,
      "loss": 0.558,
      "step": 715
    },
    {
      "epoch": 4.41717791411043,
      "grad_norm": 0.5014335513114929,
      "learning_rate": 8.89768383749821e-06,
      "loss": 0.5622,
      "step": 720
    },
    {
      "epoch": 4.447852760736196,
      "grad_norm": 0.5099634528160095,
      "learning_rate": 8.875223803486674e-06,
      "loss": 0.566,
      "step": 725
    },
    {
      "epoch": 4.478527607361963,
      "grad_norm": 0.5409094095230103,
      "learning_rate": 8.852566213878947e-06,
      "loss": 0.5562,
      "step": 730
    },
    {
      "epoch": 4.50920245398773,
      "grad_norm": 0.5168077945709229,
      "learning_rate": 8.829712223739574e-06,
      "loss": 0.5589,
      "step": 735
    },
    {
      "epoch": 4.539877300613497,
      "grad_norm": 0.5725398063659668,
      "learning_rate": 8.80666299814543e-06,
      "loss": 0.5567,
      "step": 740
    },
    {
      "epoch": 4.570552147239264,
      "grad_norm": 0.45501619577407837,
      "learning_rate": 8.783419712126335e-06,
      "loss": 0.5615,
      "step": 745
    },
    {
      "epoch": 4.601226993865031,
      "grad_norm": 0.5540989637374878,
      "learning_rate": 8.759983550605132e-06,
      "loss": 0.5578,
      "step": 750
    },
    {
      "epoch": 4.631901840490798,
      "grad_norm": 0.5188400745391846,
      "learning_rate": 8.736355708337298e-06,
      "loss": 0.5605,
      "step": 755
    },
    {
      "epoch": 4.662576687116564,
      "grad_norm": 0.48551928997039795,
      "learning_rate": 8.71253738985003e-06,
      "loss": 0.5579,
      "step": 760
    },
    {
      "epoch": 4.693251533742331,
      "grad_norm": 0.44432246685028076,
      "learning_rate": 8.688529809380843e-06,
      "loss": 0.5595,
      "step": 765
    },
    {
      "epoch": 4.723926380368098,
      "grad_norm": 0.45983144640922546,
      "learning_rate": 8.66433419081566e-06,
      "loss": 0.5611,
      "step": 770
    },
    {
      "epoch": 4.754601226993865,
      "grad_norm": 0.48174920678138733,
      "learning_rate": 8.639951767626429e-06,
      "loss": 0.5637,
      "step": 775
    },
    {
      "epoch": 4.785276073619632,
      "grad_norm": 0.4685279130935669,
      "learning_rate": 8.615383782808238e-06,
      "loss": 0.5608,
      "step": 780
    },
    {
      "epoch": 4.815950920245399,
      "grad_norm": 0.5617627501487732,
      "learning_rate": 8.590631488815945e-06,
      "loss": 0.5596,
      "step": 785
    },
    {
      "epoch": 4.846625766871165,
      "grad_norm": 0.5751484036445618,
      "learning_rate": 8.565696147500338e-06,
      "loss": 0.5596,
      "step": 790
    },
    {
      "epoch": 4.877300613496932,
      "grad_norm": 0.6005717515945435,
      "learning_rate": 8.540579030043795e-06,
      "loss": 0.5614,
      "step": 795
    },
    {
      "epoch": 4.9079754601226995,
      "grad_norm": 0.5048331022262573,
      "learning_rate": 8.515281416895489e-06,
      "loss": 0.5585,
      "step": 800
    },
    {
      "epoch": 4.938650306748467,
      "grad_norm": 0.4952603578567505,
      "learning_rate": 8.48980459770611e-06,
      "loss": 0.5599,
      "step": 805
    },
    {
      "epoch": 4.969325153374233,
      "grad_norm": 0.49712324142456055,
      "learning_rate": 8.464149871262118e-06,
      "loss": 0.5614,
      "step": 810
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5496305227279663,
      "learning_rate": 8.43831854541953e-06,
      "loss": 0.559,
      "step": 815
    },
    {
      "epoch": 5.030674846625767,
      "grad_norm": 0.565785825252533,
      "learning_rate": 8.412311937037255e-06,
      "loss": 0.5256,
      "step": 820
    },
    {
      "epoch": 5.061349693251533,
      "grad_norm": 0.5519700050354004,
      "learning_rate": 8.386131371909948e-06,
      "loss": 0.5198,
      "step": 825
    },
    {
      "epoch": 5.0920245398773005,
      "grad_norm": 0.5226123332977295,
      "learning_rate": 8.35977818470044e-06,
      "loss": 0.5235,
      "step": 830
    },
    {
      "epoch": 5.122699386503068,
      "grad_norm": 0.5327089428901672,
      "learning_rate": 8.33325371887168e-06,
      "loss": 0.5267,
      "step": 835
    },
    {
      "epoch": 5.153374233128835,
      "grad_norm": 0.5217044949531555,
      "learning_rate": 8.30655932661826e-06,
      "loss": 0.5215,
      "step": 840
    },
    {
      "epoch": 5.184049079754601,
      "grad_norm": 0.5113396048545837,
      "learning_rate": 8.279696368797471e-06,
      "loss": 0.5227,
      "step": 845
    },
    {
      "epoch": 5.214723926380368,
      "grad_norm": 0.5224789977073669,
      "learning_rate": 8.252666214859936e-06,
      "loss": 0.5211,
      "step": 850
    },
    {
      "epoch": 5.245398773006135,
      "grad_norm": 0.47648775577545166,
      "learning_rate": 8.225470242779791e-06,
      "loss": 0.5262,
      "step": 855
    },
    {
      "epoch": 5.276073619631902,
      "grad_norm": 0.5127813816070557,
      "learning_rate": 8.19810983898444e-06,
      "loss": 0.5256,
      "step": 860
    },
    {
      "epoch": 5.306748466257669,
      "grad_norm": 0.5499957799911499,
      "learning_rate": 8.170586398283878e-06,
      "loss": 0.5267,
      "step": 865
    },
    {
      "epoch": 5.337423312883436,
      "grad_norm": 0.5156161189079285,
      "learning_rate": 8.142901323799578e-06,
      "loss": 0.5252,
      "step": 870
    },
    {
      "epoch": 5.368098159509202,
      "grad_norm": 0.4825647175312042,
      "learning_rate": 8.115056026892965e-06,
      "loss": 0.5266,
      "step": 875
    },
    {
      "epoch": 5.398773006134969,
      "grad_norm": 0.5043969750404358,
      "learning_rate": 8.08705192709347e-06,
      "loss": 0.5241,
      "step": 880
    },
    {
      "epoch": 5.429447852760736,
      "grad_norm": 0.47043198347091675,
      "learning_rate": 8.058890452026155e-06,
      "loss": 0.5272,
      "step": 885
    },
    {
      "epoch": 5.460122699386503,
      "grad_norm": 0.471513956785202,
      "learning_rate": 8.030573037338942e-06,
      "loss": 0.5321,
      "step": 890
    },
    {
      "epoch": 5.49079754601227,
      "grad_norm": 0.48654860258102417,
      "learning_rate": 8.002101126629422e-06,
      "loss": 0.5267,
      "step": 895
    },
    {
      "epoch": 5.521472392638037,
      "grad_norm": 0.4844699501991272,
      "learning_rate": 7.973476171371255e-06,
      "loss": 0.5301,
      "step": 900
    },
    {
      "epoch": 5.552147239263804,
      "grad_norm": 0.4706033170223236,
      "learning_rate": 7.94469963084019e-06,
      "loss": 0.5276,
      "step": 905
    },
    {
      "epoch": 5.58282208588957,
      "grad_norm": 0.47882503271102905,
      "learning_rate": 7.91577297203966e-06,
      "loss": 0.5267,
      "step": 910
    },
    {
      "epoch": 5.613496932515337,
      "grad_norm": 0.5027369856834412,
      "learning_rate": 7.886697669625995e-06,
      "loss": 0.5313,
      "step": 915
    },
    {
      "epoch": 5.644171779141105,
      "grad_norm": 0.4971833825111389,
      "learning_rate": 7.857475205833255e-06,
      "loss": 0.5219,
      "step": 920
    },
    {
      "epoch": 5.674846625766871,
      "grad_norm": 0.488737016916275,
      "learning_rate": 7.828107070397657e-06,
      "loss": 0.5233,
      "step": 925
    },
    {
      "epoch": 5.705521472392638,
      "grad_norm": 0.49035337567329407,
      "learning_rate": 7.798594760481639e-06,
      "loss": 0.5254,
      "step": 930
    },
    {
      "epoch": 5.736196319018405,
      "grad_norm": 0.5278029441833496,
      "learning_rate": 7.768939780597523e-06,
      "loss": 0.5246,
      "step": 935
    },
    {
      "epoch": 5.766871165644172,
      "grad_norm": 0.4724002480506897,
      "learning_rate": 7.739143642530833e-06,
      "loss": 0.5218,
      "step": 940
    },
    {
      "epoch": 5.7975460122699385,
      "grad_norm": 0.5203721523284912,
      "learning_rate": 7.70920786526321e-06,
      "loss": 0.5256,
      "step": 945
    },
    {
      "epoch": 5.828220858895706,
      "grad_norm": 0.5066553950309753,
      "learning_rate": 7.679133974894984e-06,
      "loss": 0.5272,
      "step": 950
    },
    {
      "epoch": 5.858895705521473,
      "grad_norm": 0.5213795304298401,
      "learning_rate": 7.648923504567374e-06,
      "loss": 0.5293,
      "step": 955
    },
    {
      "epoch": 5.889570552147239,
      "grad_norm": 0.5189357399940491,
      "learning_rate": 7.618577994384324e-06,
      "loss": 0.525,
      "step": 960
    },
    {
      "epoch": 5.920245398773006,
      "grad_norm": 0.45614001154899597,
      "learning_rate": 7.588098991334001e-06,
      "loss": 0.5254,
      "step": 965
    },
    {
      "epoch": 5.950920245398773,
      "grad_norm": 0.5392826199531555,
      "learning_rate": 7.557488049209921e-06,
      "loss": 0.5215,
      "step": 970
    },
    {
      "epoch": 5.9815950920245395,
      "grad_norm": 0.5028013586997986,
      "learning_rate": 7.52674672853174e-06,
      "loss": 0.5249,
      "step": 975
    },
    {
      "epoch": 6.012269938650307,
      "grad_norm": 0.5520371198654175,
      "learning_rate": 7.495876596465703e-06,
      "loss": 0.5087,
      "step": 980
    },
    {
      "epoch": 6.042944785276074,
      "grad_norm": 0.6230819225311279,
      "learning_rate": 7.464879226744748e-06,
      "loss": 0.4881,
      "step": 985
    },
    {
      "epoch": 6.07361963190184,
      "grad_norm": 0.581471860408783,
      "learning_rate": 7.433756199588282e-06,
      "loss": 0.4891,
      "step": 990
    },
    {
      "epoch": 6.104294478527607,
      "grad_norm": 0.6113958358764648,
      "learning_rate": 7.402509101621618e-06,
      "loss": 0.4848,
      "step": 995
    },
    {
      "epoch": 6.134969325153374,
      "grad_norm": 0.5332604050636292,
      "learning_rate": 7.371139525795094e-06,
      "loss": 0.488,
      "step": 1000
    },
    {
      "epoch": 6.1656441717791415,
      "grad_norm": 0.538058340549469,
      "learning_rate": 7.3396490713028674e-06,
      "loss": 0.4904,
      "step": 1005
    },
    {
      "epoch": 6.196319018404908,
      "grad_norm": 0.5761954188346863,
      "learning_rate": 7.308039343501381e-06,
      "loss": 0.4907,
      "step": 1010
    },
    {
      "epoch": 6.226993865030675,
      "grad_norm": 0.5636390447616577,
      "learning_rate": 7.276311953827533e-06,
      "loss": 0.4872,
      "step": 1015
    },
    {
      "epoch": 6.257668711656442,
      "grad_norm": 0.5386704802513123,
      "learning_rate": 7.244468519716521e-06,
      "loss": 0.4946,
      "step": 1020
    },
    {
      "epoch": 6.288343558282208,
      "grad_norm": 0.585161566734314,
      "learning_rate": 7.212510664519391e-06,
      "loss": 0.4934,
      "step": 1025
    },
    {
      "epoch": 6.319018404907975,
      "grad_norm": 0.5781230330467224,
      "learning_rate": 7.180440017420277e-06,
      "loss": 0.494,
      "step": 1030
    },
    {
      "epoch": 6.3496932515337425,
      "grad_norm": 0.5272154211997986,
      "learning_rate": 7.148258213353347e-06,
      "loss": 0.4944,
      "step": 1035
    },
    {
      "epoch": 6.38036809815951,
      "grad_norm": 0.5416070222854614,
      "learning_rate": 7.115966892919459e-06,
      "loss": 0.4936,
      "step": 1040
    },
    {
      "epoch": 6.411042944785276,
      "grad_norm": 0.5524059534072876,
      "learning_rate": 7.083567702302517e-06,
      "loss": 0.4901,
      "step": 1045
    },
    {
      "epoch": 6.441717791411043,
      "grad_norm": 0.5226433277130127,
      "learning_rate": 7.05106229318556e-06,
      "loss": 0.4931,
      "step": 1050
    },
    {
      "epoch": 6.47239263803681,
      "grad_norm": 0.5324561595916748,
      "learning_rate": 7.018452322666549e-06,
      "loss": 0.4906,
      "step": 1055
    },
    {
      "epoch": 6.5030674846625764,
      "grad_norm": 0.5099997520446777,
      "learning_rate": 6.985739453173903e-06,
      "loss": 0.4909,
      "step": 1060
    },
    {
      "epoch": 6.533742331288344,
      "grad_norm": 0.5172037482261658,
      "learning_rate": 6.9529253523817396e-06,
      "loss": 0.4936,
      "step": 1065
    },
    {
      "epoch": 6.564417177914111,
      "grad_norm": 0.5281527638435364,
      "learning_rate": 6.9200116931248575e-06,
      "loss": 0.4949,
      "step": 1070
    },
    {
      "epoch": 6.595092024539877,
      "grad_norm": 0.5232874751091003,
      "learning_rate": 6.887000153313468e-06,
      "loss": 0.493,
      "step": 1075
    },
    {
      "epoch": 6.625766871165644,
      "grad_norm": 0.5422880053520203,
      "learning_rate": 6.853892415847645e-06,
      "loss": 0.4944,
      "step": 1080
    },
    {
      "epoch": 6.656441717791411,
      "grad_norm": 0.5587918758392334,
      "learning_rate": 6.8206901685315366e-06,
      "loss": 0.4974,
      "step": 1085
    },
    {
      "epoch": 6.6871165644171775,
      "grad_norm": 0.5716516971588135,
      "learning_rate": 6.787395103987323e-06,
      "loss": 0.4959,
      "step": 1090
    },
    {
      "epoch": 6.717791411042945,
      "grad_norm": 0.5290241837501526,
      "learning_rate": 6.754008919568927e-06,
      "loss": 0.49,
      "step": 1095
    },
    {
      "epoch": 6.748466257668712,
      "grad_norm": 0.545229971408844,
      "learning_rate": 6.72053331727549e-06,
      "loss": 0.4913,
      "step": 1100
    },
    {
      "epoch": 6.779141104294479,
      "grad_norm": 0.5339862108230591,
      "learning_rate": 6.686970003664588e-06,
      "loss": 0.4933,
      "step": 1105
    },
    {
      "epoch": 6.809815950920245,
      "grad_norm": 0.5517235398292542,
      "learning_rate": 6.653320689765257e-06,
      "loss": 0.4957,
      "step": 1110
    },
    {
      "epoch": 6.840490797546012,
      "grad_norm": 0.5221419334411621,
      "learning_rate": 6.619587090990748e-06,
      "loss": 0.4938,
      "step": 1115
    },
    {
      "epoch": 6.871165644171779,
      "grad_norm": 0.5578407049179077,
      "learning_rate": 6.585770927051085e-06,
      "loss": 0.4992,
      "step": 1120
    },
    {
      "epoch": 6.901840490797546,
      "grad_norm": 0.5061231255531311,
      "learning_rate": 6.551873921865393e-06,
      "loss": 0.4968,
      "step": 1125
    },
    {
      "epoch": 6.932515337423313,
      "grad_norm": 0.5198781490325928,
      "learning_rate": 6.517897803474011e-06,
      "loss": 0.4982,
      "step": 1130
    },
    {
      "epoch": 6.96319018404908,
      "grad_norm": 0.5066606998443604,
      "learning_rate": 6.483844303950411e-06,
      "loss": 0.4975,
      "step": 1135
    },
    {
      "epoch": 6.993865030674847,
      "grad_norm": 0.5105178952217102,
      "learning_rate": 6.4497151593128795e-06,
      "loss": 0.5015,
      "step": 1140
    },
    {
      "epoch": 7.024539877300613,
      "grad_norm": 0.6838037967681885,
      "learning_rate": 6.415512109436031e-06,
      "loss": 0.4634,
      "step": 1145
    },
    {
      "epoch": 7.0552147239263805,
      "grad_norm": 0.6098262667655945,
      "learning_rate": 6.381236897962102e-06,
      "loss": 0.4616,
      "step": 1150
    },
    {
      "epoch": 7.085889570552148,
      "grad_norm": 0.5857096314430237,
      "learning_rate": 6.3468912722120715e-06,
      "loss": 0.4557,
      "step": 1155
    },
    {
      "epoch": 7.116564417177914,
      "grad_norm": 0.5579211115837097,
      "learning_rate": 6.312476983096573e-06,
      "loss": 0.454,
      "step": 1160
    },
    {
      "epoch": 7.147239263803681,
      "grad_norm": 0.5579200387001038,
      "learning_rate": 6.277995785026642e-06,
      "loss": 0.4606,
      "step": 1165
    },
    {
      "epoch": 7.177914110429448,
      "grad_norm": 0.5602801442146301,
      "learning_rate": 6.243449435824276e-06,
      "loss": 0.4557,
      "step": 1170
    },
    {
      "epoch": 7.208588957055214,
      "grad_norm": 0.5627221465110779,
      "learning_rate": 6.2088396966328155e-06,
      "loss": 0.4564,
      "step": 1175
    },
    {
      "epoch": 7.2392638036809815,
      "grad_norm": 0.5709859728813171,
      "learning_rate": 6.174168331827179e-06,
      "loss": 0.4596,
      "step": 1180
    },
    {
      "epoch": 7.269938650306749,
      "grad_norm": 0.554961621761322,
      "learning_rate": 6.139437108923898e-06,
      "loss": 0.4583,
      "step": 1185
    },
    {
      "epoch": 7.300613496932515,
      "grad_norm": 0.5725740790367126,
      "learning_rate": 6.1046477984910215e-06,
      "loss": 0.458,
      "step": 1190
    },
    {
      "epoch": 7.331288343558282,
      "grad_norm": 0.5590230226516724,
      "learning_rate": 6.069802174057849e-06,
      "loss": 0.4604,
      "step": 1195
    },
    {
      "epoch": 7.361963190184049,
      "grad_norm": 0.5823590159416199,
      "learning_rate": 6.034902012024521e-06,
      "loss": 0.4575,
      "step": 1200
    }
  ],
  "logging_steps": 5,
  "max_steps": 2445,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 400,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6846057011425575e+19,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}