{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.724563644103874,
  "eval_steps": 500,
  "global_step": 800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017028522775649212,
      "grad_norm": 3.096078395843506,
      "learning_rate": 0.0001988623435722412,
      "loss": 2.9252,
      "mean_token_accuracy": 0.5491172857582569,
      "num_tokens": 4766.0,
      "step": 5
    },
    {
      "epoch": 0.034057045551298425,
      "grad_norm": 0.695277214050293,
      "learning_rate": 0.00019772468714448238,
      "loss": 0.924,
      "mean_token_accuracy": 0.8361630752682686,
      "num_tokens": 9556.0,
      "step": 10
    },
    {
      "epoch": 0.05108556832694764,
      "grad_norm": 0.541026771068573,
      "learning_rate": 0.00019658703071672356,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8945453137159347,
      "num_tokens": 14116.0,
      "step": 15
    },
    {
      "epoch": 0.06811409110259685,
      "grad_norm": 0.5480035543441772,
      "learning_rate": 0.00019544937428896475,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8836526513099671,
      "num_tokens": 18879.0,
      "step": 20
    },
    {
      "epoch": 0.08514261387824607,
      "grad_norm": 0.5408394932746887,
      "learning_rate": 0.00019431171786120593,
      "loss": 0.6289,
      "mean_token_accuracy": 0.8782178282737731,
      "num_tokens": 23699.0,
      "step": 25
    },
    {
      "epoch": 0.10217113665389528,
      "grad_norm": 0.583755612373352,
      "learning_rate": 0.00019317406143344712,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8905435591936112,
      "num_tokens": 28425.0,
      "step": 30
    },
    {
      "epoch": 0.11919965942954448,
      "grad_norm": 0.5654754638671875,
      "learning_rate": 0.0001920364050056883,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8963450387120246,
      "num_tokens": 33060.0,
      "step": 35
    },
    {
      "epoch": 0.1362281822051937,
      "grad_norm": 0.5462912321090698,
      "learning_rate": 0.00019089874857792946,
      "loss": 0.6819,
      "mean_token_accuracy": 0.8708900690078736,
      "num_tokens": 38017.0,
      "step": 40
    },
    {
      "epoch": 0.1532567049808429,
      "grad_norm": 0.45884689688682556,
      "learning_rate": 0.00018976109215017067,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8896441653370857,
      "num_tokens": 42760.0,
      "step": 45
    },
    {
      "epoch": 0.17028522775649213,
      "grad_norm": 0.5886663198471069,
      "learning_rate": 0.00018862343572241183,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8954117730259895,
      "num_tokens": 47455.0,
      "step": 50
    },
    {
      "epoch": 0.18731375053214133,
      "grad_norm": 0.5249215364456177,
      "learning_rate": 0.00018748577929465302,
      "loss": 0.4948,
      "mean_token_accuracy": 0.9001417383551598,
      "num_tokens": 52186.0,
      "step": 55
    },
    {
      "epoch": 0.20434227330779056,
      "grad_norm": 0.40858832001686096,
      "learning_rate": 0.0001863481228668942,
      "loss": 0.4616,
      "mean_token_accuracy": 0.9116851478815079,
      "num_tokens": 56765.0,
      "step": 60
    },
    {
      "epoch": 0.22137079608343976,
      "grad_norm": 0.4287504553794861,
      "learning_rate": 0.0001852104664391354,
      "loss": 0.5232,
      "mean_token_accuracy": 0.895601749420166,
      "num_tokens": 61436.0,
      "step": 65
    },
    {
      "epoch": 0.23839931885908897,
      "grad_norm": 0.5480343103408813,
      "learning_rate": 0.00018407281001137657,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8878508538007737,
      "num_tokens": 66190.0,
      "step": 70
    },
    {
      "epoch": 0.2554278416347382,
      "grad_norm": 0.4539305567741394,
      "learning_rate": 0.00018293515358361776,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8946248903870583,
      "num_tokens": 70888.0,
      "step": 75
    },
    {
      "epoch": 0.2724563644103874,
      "grad_norm": 0.5497062802314758,
      "learning_rate": 0.00018179749715585894,
      "loss": 0.4996,
      "mean_token_accuracy": 0.9008801370859146,
      "num_tokens": 75588.0,
      "step": 80
    },
    {
      "epoch": 0.2894848871860366,
      "grad_norm": 0.5979815721511841,
      "learning_rate": 0.00018065984072810013,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8904410973191261,
      "num_tokens": 80394.0,
      "step": 85
    },
    {
      "epoch": 0.3065134099616858,
      "grad_norm": 0.44527795910835266,
      "learning_rate": 0.0001795221843003413,
      "loss": 0.4675,
      "mean_token_accuracy": 0.9040657833218575,
      "num_tokens": 85088.0,
      "step": 90
    },
    {
      "epoch": 0.32354193273733506,
      "grad_norm": 0.5105463862419128,
      "learning_rate": 0.0001783845278725825,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8978576123714447,
      "num_tokens": 89799.0,
      "step": 95
    },
    {
      "epoch": 0.34057045551298426,
      "grad_norm": 0.5285545587539673,
      "learning_rate": 0.00017724687144482368,
      "loss": 0.5426,
      "mean_token_accuracy": 0.890034094452858,
      "num_tokens": 94573.0,
      "step": 100
    },
    {
      "epoch": 0.35759897828863346,
      "grad_norm": 0.5576184988021851,
      "learning_rate": 0.00017610921501706487,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8902270376682282,
      "num_tokens": 99357.0,
      "step": 105
    },
    {
      "epoch": 0.37462750106428266,
      "grad_norm": 0.5248317718505859,
      "learning_rate": 0.00017497155858930602,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8933822825551033,
      "num_tokens": 104092.0,
      "step": 110
    },
    {
      "epoch": 0.39165602383993187,
      "grad_norm": 0.45350518822669983,
      "learning_rate": 0.00017383390216154724,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8876873329281807,
      "num_tokens": 108902.0,
      "step": 115
    },
    {
      "epoch": 0.4086845466155811,
      "grad_norm": 0.5155291557312012,
      "learning_rate": 0.0001726962457337884,
      "loss": 0.5019,
      "mean_token_accuracy": 0.9029773235321045,
      "num_tokens": 113571.0,
      "step": 120
    },
    {
      "epoch": 0.4257130693912303,
      "grad_norm": 0.5053763389587402,
      "learning_rate": 0.0001715585893060296,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8906429842114448,
      "num_tokens": 118402.0,
      "step": 125
    },
    {
      "epoch": 0.4427415921668795,
      "grad_norm": 0.5493877530097961,
      "learning_rate": 0.00017042093287827076,
      "loss": 0.4821,
      "mean_token_accuracy": 0.9028852418065071,
      "num_tokens": 123147.0,
      "step": 130
    },
    {
      "epoch": 0.45977011494252873,
      "grad_norm": 0.426513135433197,
      "learning_rate": 0.00016928327645051198,
      "loss": 0.5311,
      "mean_token_accuracy": 0.8920483842492104,
      "num_tokens": 127878.0,
      "step": 135
    },
    {
      "epoch": 0.47679863771817793,
      "grad_norm": 0.5025385022163391,
      "learning_rate": 0.00016814562002275313,
      "loss": 0.4699,
      "mean_token_accuracy": 0.9039205580949783,
      "num_tokens": 132591.0,
      "step": 140
    },
    {
      "epoch": 0.49382716049382713,
      "grad_norm": 0.4677104651927948,
      "learning_rate": 0.00016700796359499432,
      "loss": 0.4876,
      "mean_token_accuracy": 0.9038951709866524,
      "num_tokens": 137221.0,
      "step": 145
    },
    {
      "epoch": 0.5108556832694764,
      "grad_norm": 0.5534231066703796,
      "learning_rate": 0.0001658703071672355,
      "loss": 0.4969,
      "mean_token_accuracy": 0.897729343175888,
      "num_tokens": 141981.0,
      "step": 150
    },
    {
      "epoch": 0.5278842060451255,
      "grad_norm": 0.4998553395271301,
      "learning_rate": 0.0001647326507394767,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8921540513634681,
      "num_tokens": 146807.0,
      "step": 155
    },
    {
      "epoch": 0.5449127288207748,
      "grad_norm": 0.545978307723999,
      "learning_rate": 0.00016359499431171787,
      "loss": 0.4543,
      "mean_token_accuracy": 0.9018902003765106,
      "num_tokens": 151544.0,
      "step": 160
    },
    {
      "epoch": 0.561941251596424,
      "grad_norm": 0.5749042630195618,
      "learning_rate": 0.00016245733788395906,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8976296812295914,
      "num_tokens": 156280.0,
      "step": 165
    },
    {
      "epoch": 0.5789697743720732,
      "grad_norm": 0.5217350125312805,
      "learning_rate": 0.00016131968145620024,
      "loss": 0.4741,
      "mean_token_accuracy": 0.9045222416520119,
      "num_tokens": 161011.0,
      "step": 170
    },
    {
      "epoch": 0.5959982971477225,
      "grad_norm": 0.5088939666748047,
      "learning_rate": 0.00016018202502844143,
      "loss": 0.4655,
      "mean_token_accuracy": 0.9078934848308563,
      "num_tokens": 165670.0,
      "step": 175
    },
    {
      "epoch": 0.6130268199233716,
      "grad_norm": 0.5888277292251587,
      "learning_rate": 0.0001590443686006826,
      "loss": 0.4692,
      "mean_token_accuracy": 0.9044368535280227,
      "num_tokens": 170377.0,
      "step": 180
    },
    {
      "epoch": 0.6300553426990209,
      "grad_norm": 0.6187074780464172,
      "learning_rate": 0.0001579067121729238,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8983473464846611,
      "num_tokens": 175090.0,
      "step": 185
    },
    {
      "epoch": 0.6470838654746701,
      "grad_norm": 0.5764688849449158,
      "learning_rate": 0.00015676905574516496,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8986694201827049,
      "num_tokens": 179933.0,
      "step": 190
    },
    {
      "epoch": 0.6641123882503193,
      "grad_norm": 0.5491649508476257,
      "learning_rate": 0.00015563139931740617,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8957466080784797,
      "num_tokens": 184767.0,
      "step": 195
    },
    {
      "epoch": 0.6811409110259685,
      "grad_norm": 0.4836632311344147,
      "learning_rate": 0.00015449374288964733,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8894904315471649,
      "num_tokens": 189624.0,
      "step": 200
    },
    {
      "epoch": 0.6981694338016177,
      "grad_norm": 0.4630073308944702,
      "learning_rate": 0.00015335608646188854,
      "loss": 0.4529,
      "mean_token_accuracy": 0.9051140874624253,
      "num_tokens": 194357.0,
      "step": 205
    },
    {
      "epoch": 0.7151979565772669,
      "grad_norm": 0.49463990330696106,
      "learning_rate": 0.0001522184300341297,
      "loss": 0.4165,
      "mean_token_accuracy": 0.9074924409389495,
      "num_tokens": 199038.0,
      "step": 210
    },
    {
      "epoch": 0.7322264793529162,
      "grad_norm": 0.5512949824333191,
      "learning_rate": 0.00015108077360637088,
      "loss": 0.4798,
      "mean_token_accuracy": 0.9015959203243256,
      "num_tokens": 203805.0,
      "step": 215
    },
    {
      "epoch": 0.7492550021285653,
      "grad_norm": 0.58586585521698,
      "learning_rate": 0.00014994311717861207,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8944625854492188,
      "num_tokens": 208579.0,
      "step": 220
    },
    {
      "epoch": 0.7662835249042146,
      "grad_norm": 0.4895232319831848,
      "learning_rate": 0.00014880546075085325,
      "loss": 0.4694,
      "mean_token_accuracy": 0.9098370909690857,
      "num_tokens": 213237.0,
      "step": 225
    },
    {
      "epoch": 0.7833120476798637,
      "grad_norm": 0.5064729452133179,
      "learning_rate": 0.00014766780432309444,
      "loss": 0.5047,
      "mean_token_accuracy": 0.8984142780303955,
      "num_tokens": 217964.0,
      "step": 230
    },
    {
      "epoch": 0.800340570455513,
      "grad_norm": 0.5260823965072632,
      "learning_rate": 0.00014653014789533562,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8943046078085899,
      "num_tokens": 222821.0,
      "step": 235
    },
    {
      "epoch": 0.8173690932311622,
      "grad_norm": 0.5452587604522705,
      "learning_rate": 0.0001453924914675768,
      "loss": 0.4438,
      "mean_token_accuracy": 0.905141781270504,
      "num_tokens": 227453.0,
      "step": 240
    },
    {
      "epoch": 0.8343976160068114,
      "grad_norm": 0.5265963077545166,
      "learning_rate": 0.000144254835039818,
      "loss": 0.472,
      "mean_token_accuracy": 0.9038746282458305,
      "num_tokens": 232153.0,
      "step": 245
    },
    {
      "epoch": 0.8514261387824607,
      "grad_norm": 0.655539333820343,
      "learning_rate": 0.00014311717861205915,
      "loss": 0.542,
      "mean_token_accuracy": 0.8909124419093132,
      "num_tokens": 236925.0,
      "step": 250
    },
    {
      "epoch": 0.8684546615581098,
      "grad_norm": 0.4830772578716278,
      "learning_rate": 0.00014197952218430036,
      "loss": 0.4432,
      "mean_token_accuracy": 0.9110842600464821,
      "num_tokens": 241546.0,
      "step": 255
    },
    {
      "epoch": 0.885483184333759,
      "grad_norm": 0.5120635032653809,
      "learning_rate": 0.00014084186575654152,
      "loss": 0.4758,
      "mean_token_accuracy": 0.904394218325615,
      "num_tokens": 246278.0,
      "step": 260
    },
    {
      "epoch": 0.9025117071094083,
      "grad_norm": 0.49142977595329285,
      "learning_rate": 0.00013970420932878273,
      "loss": 0.481,
      "mean_token_accuracy": 0.8970154285430908,
      "num_tokens": 251040.0,
      "step": 265
    },
    {
      "epoch": 0.9195402298850575,
      "grad_norm": 0.5381618738174438,
      "learning_rate": 0.0001385665529010239,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8992804601788521,
      "num_tokens": 255781.0,
      "step": 270
    },
    {
      "epoch": 0.9365687526607067,
      "grad_norm": 0.5239886045455933,
      "learning_rate": 0.0001374288964732651,
      "loss": 0.4957,
      "mean_token_accuracy": 0.9006548926234246,
      "num_tokens": 260512.0,
      "step": 275
    },
    {
      "epoch": 0.9535972754363559,
      "grad_norm": 0.5325700640678406,
      "learning_rate": 0.00013629124004550626,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8976434960961341,
      "num_tokens": 265234.0,
      "step": 280
    },
    {
      "epoch": 0.9706257982120051,
      "grad_norm": 0.52351975440979,
      "learning_rate": 0.00013515358361774744,
      "loss": 0.4408,
      "mean_token_accuracy": 0.9042558416724205,
      "num_tokens": 269939.0,
      "step": 285
    },
    {
      "epoch": 0.9876543209876543,
      "grad_norm": 0.5632317066192627,
      "learning_rate": 0.00013401592718998863,
      "loss": 0.4267,
      "mean_token_accuracy": 0.9127328768372536,
      "num_tokens": 274580.0,
      "step": 290
    },
    {
      "epoch": 1.0046828437633035,
      "grad_norm": 0.43729496002197266,
      "learning_rate": 0.00013287827076222981,
      "loss": 0.5103,
      "mean_token_accuracy": 0.9010089665651322,
      "num_tokens": 279321.0,
      "step": 295
    },
    {
      "epoch": 1.0217113665389528,
      "grad_norm": 0.45043033361434937,
      "learning_rate": 0.000131740614334471,
      "loss": 0.376,
      "mean_token_accuracy": 0.9193378701806069,
      "num_tokens": 284033.0,
      "step": 300
    },
    {
      "epoch": 1.038739889314602,
      "grad_norm": 0.6295264363288879,
      "learning_rate": 0.00013060295790671218,
      "loss": 0.3561,
      "mean_token_accuracy": 0.9209762245416642,
      "num_tokens": 288815.0,
      "step": 305
    },
    {
      "epoch": 1.055768412090251,
      "grad_norm": 0.5746341347694397,
      "learning_rate": 0.00012946530147895337,
      "loss": 0.361,
      "mean_token_accuracy": 0.9210935369133949,
      "num_tokens": 293454.0,
      "step": 310
    },
    {
      "epoch": 1.0727969348659003,
      "grad_norm": 0.5386441349983215,
      "learning_rate": 0.00012832764505119455,
      "loss": 0.3631,
      "mean_token_accuracy": 0.923116135597229,
      "num_tokens": 298136.0,
      "step": 315
    },
    {
      "epoch": 1.0898254576415496,
      "grad_norm": 0.637798547744751,
      "learning_rate": 0.0001271899886234357,
      "loss": 0.375,
      "mean_token_accuracy": 0.9151263251900673,
      "num_tokens": 302962.0,
      "step": 320
    },
    {
      "epoch": 1.1068539804171988,
      "grad_norm": 0.6014487147331238,
      "learning_rate": 0.00012605233219567692,
      "loss": 0.3584,
      "mean_token_accuracy": 0.9245758667588234,
      "num_tokens": 307661.0,
      "step": 325
    },
    {
      "epoch": 1.123882503192848,
      "grad_norm": 0.6004022359848022,
      "learning_rate": 0.00012491467576791808,
      "loss": 0.3569,
      "mean_token_accuracy": 0.921052211523056,
      "num_tokens": 312330.0,
      "step": 330
    },
    {
      "epoch": 1.1409110259684971,
      "grad_norm": 0.8666739463806152,
      "learning_rate": 0.0001237770193401593,
      "loss": 0.3778,
      "mean_token_accuracy": 0.9176680579781532,
      "num_tokens": 317107.0,
      "step": 335
    },
    {
      "epoch": 1.1579395487441464,
      "grad_norm": 0.5575281381607056,
      "learning_rate": 0.00012263936291240045,
      "loss": 0.3765,
      "mean_token_accuracy": 0.9149453848600387,
      "num_tokens": 321895.0,
      "step": 340
    },
    {
      "epoch": 1.1749680715197957,
      "grad_norm": 0.5432224273681641,
      "learning_rate": 0.00012150170648464165,
      "loss": 0.3546,
      "mean_token_accuracy": 0.9211765006184578,
      "num_tokens": 326592.0,
      "step": 345
    },
    {
      "epoch": 1.191996594295445,
      "grad_norm": 0.5993474721908569,
      "learning_rate": 0.00012036405005688282,
      "loss": 0.3721,
      "mean_token_accuracy": 0.9195932745933533,
      "num_tokens": 331391.0,
      "step": 350
    },
    {
      "epoch": 1.2090251170710942,
      "grad_norm": 0.6452960968017578,
      "learning_rate": 0.000119226393629124,
      "loss": 0.4292,
      "mean_token_accuracy": 0.9088605523109436,
      "num_tokens": 336238.0,
      "step": 355
    },
    {
      "epoch": 1.2260536398467432,
      "grad_norm": 0.5383236408233643,
      "learning_rate": 0.00011808873720136519,
      "loss": 0.3192,
      "mean_token_accuracy": 0.9233780831098557,
      "num_tokens": 340817.0,
      "step": 360
    },
    {
      "epoch": 1.2430821626223925,
      "grad_norm": 0.6566250324249268,
      "learning_rate": 0.00011695108077360638,
      "loss": 0.3617,
      "mean_token_accuracy": 0.916201414167881,
      "num_tokens": 345509.0,
      "step": 365
    },
    {
      "epoch": 1.2601106853980417,
      "grad_norm": 0.6461586952209473,
      "learning_rate": 0.00011581342434584756,
      "loss": 0.3971,
      "mean_token_accuracy": 0.9110044553875923,
      "num_tokens": 350303.0,
      "step": 370
    },
    {
      "epoch": 1.277139208173691,
      "grad_norm": 0.700552225112915,
      "learning_rate": 0.00011467576791808873,
      "loss": 0.3439,
      "mean_token_accuracy": 0.9284276351332664,
      "num_tokens": 354873.0,
      "step": 375
    },
    {
      "epoch": 1.29416773094934,
      "grad_norm": 0.6307278871536255,
      "learning_rate": 0.00011353811149032993,
      "loss": 0.3619,
      "mean_token_accuracy": 0.9204769432544708,
      "num_tokens": 359564.0,
      "step": 380
    },
    {
      "epoch": 1.3111962537249893,
      "grad_norm": 0.5559372901916504,
      "learning_rate": 0.0001124004550625711,
      "loss": 0.337,
      "mean_token_accuracy": 0.9221480071544648,
      "num_tokens": 364285.0,
      "step": 385
    },
    {
      "epoch": 1.3282247765006385,
      "grad_norm": 0.7321580648422241,
      "learning_rate": 0.00011126279863481229,
      "loss": 0.3864,
      "mean_token_accuracy": 0.917863991856575,
      "num_tokens": 369021.0,
      "step": 390
    },
    {
      "epoch": 1.3452532992762878,
      "grad_norm": 0.6482062339782715,
      "learning_rate": 0.00011012514220705347,
      "loss": 0.342,
      "mean_token_accuracy": 0.9234487116336823,
      "num_tokens": 373711.0,
      "step": 395
    },
    {
      "epoch": 1.362281822051937,
      "grad_norm": 0.7216457724571228,
      "learning_rate": 0.00010898748577929466,
      "loss": 0.3494,
      "mean_token_accuracy": 0.9197908401489258,
      "num_tokens": 378403.0,
      "step": 400
    },
    {
      "epoch": 1.3793103448275863,
      "grad_norm": 0.7166584730148315,
      "learning_rate": 0.00010784982935153584,
      "loss": 0.3768,
      "mean_token_accuracy": 0.9174183040857316,
      "num_tokens": 383100.0,
      "step": 405
    },
    {
      "epoch": 1.3963388676032356,
      "grad_norm": 0.8020215630531311,
      "learning_rate": 0.00010671217292377703,
      "loss": 0.3772,
      "mean_token_accuracy": 0.9173366680741311,
      "num_tokens": 387930.0,
      "step": 410
    },
    {
      "epoch": 1.4133673903788846,
      "grad_norm": 0.6404600143432617,
      "learning_rate": 0.00010557451649601821,
      "loss": 0.3345,
      "mean_token_accuracy": 0.9267360553145408,
      "num_tokens": 392624.0,
      "step": 415
    },
    {
      "epoch": 1.4303959131545338,
      "grad_norm": 0.7745601534843445,
      "learning_rate": 0.00010443686006825938,
      "loss": 0.3762,
      "mean_token_accuracy": 0.9124831840395927,
      "num_tokens": 397405.0,
      "step": 420
    },
    {
      "epoch": 1.447424435930183,
      "grad_norm": 0.6004043817520142,
      "learning_rate": 0.00010329920364050057,
      "loss": 0.402,
      "mean_token_accuracy": 0.914434814453125,
      "num_tokens": 402178.0,
      "step": 425
    },
    {
      "epoch": 1.4644529587058321,
      "grad_norm": 0.6424717903137207,
      "learning_rate": 0.00010216154721274175,
      "loss": 0.3269,
      "mean_token_accuracy": 0.9274717316031456,
      "num_tokens": 406801.0,
      "step": 430
    },
    {
      "epoch": 1.4814814814814814,
      "grad_norm": 0.7379339337348938,
      "learning_rate": 0.00010102389078498294,
      "loss": 0.3853,
      "mean_token_accuracy": 0.9116453334689141,
      "num_tokens": 411635.0,
      "step": 435
    },
    {
      "epoch": 1.4985100042571307,
      "grad_norm": 0.6798533201217651,
      "learning_rate": 9.988623435722412e-05,
      "loss": 0.3442,
      "mean_token_accuracy": 0.9197518125176429,
      "num_tokens": 416372.0,
      "step": 440
    },
    {
      "epoch": 1.51553852703278,
      "grad_norm": 0.5506631731987,
      "learning_rate": 9.874857792946531e-05,
      "loss": 0.3557,
      "mean_token_accuracy": 0.9198912262916565,
      "num_tokens": 421129.0,
      "step": 445
    },
    {
      "epoch": 1.5325670498084292,
      "grad_norm": 32.163753509521484,
      "learning_rate": 9.761092150170649e-05,
      "loss": 0.2988,
      "mean_token_accuracy": 0.9316724047064782,
      "num_tokens": 425749.0,
      "step": 450
    },
    {
      "epoch": 1.5495955725840784,
      "grad_norm": 0.6345335841178894,
      "learning_rate": 9.647326507394768e-05,
      "loss": 0.328,
      "mean_token_accuracy": 0.925884073972702,
      "num_tokens": 430356.0,
      "step": 455
    },
    {
      "epoch": 1.5666240953597277,
      "grad_norm": 0.4996885657310486,
      "learning_rate": 9.533560864618886e-05,
      "loss": 0.3357,
      "mean_token_accuracy": 0.9236923009157181,
      "num_tokens": 435131.0,
      "step": 460
    },
    {
      "epoch": 1.5836526181353767,
      "grad_norm": 0.7736502289772034,
      "learning_rate": 9.419795221843003e-05,
      "loss": 0.3363,
      "mean_token_accuracy": 0.919234861433506,
      "num_tokens": 439799.0,
      "step": 465
    },
    {
      "epoch": 1.600681140911026,
      "grad_norm": 0.5958617925643921,
      "learning_rate": 9.306029579067122e-05,
      "loss": 0.3636,
      "mean_token_accuracy": 0.92172020226717,
      "num_tokens": 444514.0,
      "step": 470
    },
    {
      "epoch": 1.617709663686675,
      "grad_norm": 0.7543319463729858,
      "learning_rate": 9.19226393629124e-05,
      "loss": 0.3993,
      "mean_token_accuracy": 0.9095039024949074,
      "num_tokens": 449453.0,
      "step": 475
    },
    {
      "epoch": 1.6347381864623243,
      "grad_norm": 0.6482803225517273,
      "learning_rate": 9.078498293515359e-05,
      "loss": 0.3359,
      "mean_token_accuracy": 0.924050772190094,
      "num_tokens": 454144.0,
      "step": 480
    },
    {
      "epoch": 1.6517667092379735,
      "grad_norm": 0.5878757834434509,
      "learning_rate": 8.964732650739477e-05,
      "loss": 0.3728,
      "mean_token_accuracy": 0.9194103494286537,
      "num_tokens": 458839.0,
      "step": 485
    },
    {
      "epoch": 1.6687952320136228,
      "grad_norm": 0.8008145689964294,
      "learning_rate": 8.850967007963596e-05,
      "loss": 0.3973,
      "mean_token_accuracy": 0.9146214991807937,
      "num_tokens": 463633.0,
      "step": 490
    },
    {
      "epoch": 1.685823754789272,
      "grad_norm": 0.6370537281036377,
      "learning_rate": 8.737201365187714e-05,
      "loss": 0.3528,
      "mean_token_accuracy": 0.9191272169351578,
      "num_tokens": 468319.0,
      "step": 495
    },
    {
      "epoch": 1.7028522775649213,
      "grad_norm": 0.6956344842910767,
      "learning_rate": 8.623435722411833e-05,
      "loss": 0.3584,
      "mean_token_accuracy": 0.9227298349142075,
      "num_tokens": 473018.0,
      "step": 500
    },
    {
      "epoch": 1.7198808003405706,
      "grad_norm": 0.5346872806549072,
      "learning_rate": 8.509670079635951e-05,
      "loss": 0.3094,
      "mean_token_accuracy": 0.9289079144597053,
      "num_tokens": 477564.0,
      "step": 505
    },
    {
      "epoch": 1.7369093231162198,
      "grad_norm": 0.6789258122444153,
      "learning_rate": 8.395904436860069e-05,
      "loss": 0.3999,
      "mean_token_accuracy": 0.9171444311738014,
      "num_tokens": 482329.0,
      "step": 510
    },
    {
      "epoch": 1.7539378458918689,
      "grad_norm": 0.7325442433357239,
      "learning_rate": 8.282138794084187e-05,
      "loss": 0.3576,
      "mean_token_accuracy": 0.919205529987812,
      "num_tokens": 487073.0,
      "step": 515
    },
    {
      "epoch": 1.770966368667518,
      "grad_norm": 0.6846973299980164,
      "learning_rate": 8.168373151308306e-05,
      "loss": 0.3547,
      "mean_token_accuracy": 0.9206994399428368,
      "num_tokens": 491796.0,
      "step": 520
    },
    {
      "epoch": 1.7879948914431671,
      "grad_norm": 0.5631270408630371,
      "learning_rate": 8.054607508532424e-05,
      "loss": 0.3473,
      "mean_token_accuracy": 0.9214297071099281,
      "num_tokens": 496498.0,
      "step": 525
    },
    {
      "epoch": 1.8050234142188164,
      "grad_norm": 0.6066986322402954,
      "learning_rate": 7.940841865756543e-05,
      "loss": 0.3774,
      "mean_token_accuracy": 0.9192397773265839,
      "num_tokens": 501235.0,
      "step": 530
    },
    {
      "epoch": 1.8220519369944657,
      "grad_norm": 0.6359360218048096,
      "learning_rate": 7.827076222980661e-05,
      "loss": 0.324,
      "mean_token_accuracy": 0.9266907840967178,
      "num_tokens": 505887.0,
      "step": 535
    },
    {
      "epoch": 1.839080459770115,
      "grad_norm": 0.836995005607605,
      "learning_rate": 7.71331058020478e-05,
      "loss": 0.3708,
      "mean_token_accuracy": 0.915412899851799,
      "num_tokens": 510670.0,
      "step": 540
    },
    {
      "epoch": 1.8561089825457642,
      "grad_norm": 0.6547362208366394,
      "learning_rate": 7.599544937428897e-05,
      "loss": 0.394,
      "mean_token_accuracy": 0.9127551719546318,
      "num_tokens": 515432.0,
      "step": 545
    },
    {
      "epoch": 1.8731375053214134,
      "grad_norm": 0.6609506607055664,
      "learning_rate": 7.485779294653015e-05,
      "loss": 0.4105,
      "mean_token_accuracy": 0.910239189863205,
      "num_tokens": 520330.0,
      "step": 550
    },
    {
      "epoch": 1.8901660280970627,
      "grad_norm": 0.6680045127868652,
      "learning_rate": 7.372013651877134e-05,
      "loss": 0.3478,
      "mean_token_accuracy": 0.9242256492376327,
      "num_tokens": 524980.0,
      "step": 555
    },
    {
      "epoch": 1.907194550872712,
      "grad_norm": 0.579949676990509,
      "learning_rate": 7.258248009101252e-05,
      "loss": 0.389,
      "mean_token_accuracy": 0.9161396250128746,
      "num_tokens": 529754.0,
      "step": 560
    },
    {
      "epoch": 1.924223073648361,
      "grad_norm": 0.6184946298599243,
      "learning_rate": 7.14448236632537e-05,
      "loss": 0.3253,
      "mean_token_accuracy": 0.924703124165535,
      "num_tokens": 534525.0,
      "step": 565
    },
    {
      "epoch": 1.9412515964240102,
      "grad_norm": 0.657850980758667,
      "learning_rate": 7.030716723549489e-05,
      "loss": 0.3926,
      "mean_token_accuracy": 0.9116024389863014,
      "num_tokens": 539435.0,
      "step": 570
    },
    {
      "epoch": 1.9582801191996593,
      "grad_norm": 0.7310178875923157,
      "learning_rate": 6.916951080773608e-05,
      "loss": 0.4016,
      "mean_token_accuracy": 0.9142703726887703,
      "num_tokens": 544326.0,
      "step": 575
    },
    {
      "epoch": 1.9753086419753085,
      "grad_norm": 0.7617822289466858,
      "learning_rate": 6.803185437997726e-05,
      "loss": 0.3878,
      "mean_token_accuracy": 0.9160768195986748,
      "num_tokens": 549062.0,
      "step": 580
    },
    {
      "epoch": 1.9923371647509578,
      "grad_norm": 0.5839700102806091,
      "learning_rate": 6.689419795221843e-05,
      "loss": 0.3557,
      "mean_token_accuracy": 0.9169671639800072,
      "num_tokens": 553832.0,
      "step": 585
    },
    {
      "epoch": 2.009365687526607,
      "grad_norm": 0.5534153580665588,
      "learning_rate": 6.575654152445962e-05,
      "loss": 0.3887,
      "mean_token_accuracy": 0.9231562212109565,
      "num_tokens": 558589.0,
      "step": 590
    },
    {
      "epoch": 2.0263942103022563,
      "grad_norm": 0.5666193962097168,
      "learning_rate": 6.46188850967008e-05,
      "loss": 0.2234,
      "mean_token_accuracy": 0.9479417443275452,
      "num_tokens": 563291.0,
      "step": 595
    },
    {
      "epoch": 2.0434227330779056,
      "grad_norm": 0.6161417961120605,
      "learning_rate": 6.348122866894199e-05,
      "loss": 0.2351,
      "mean_token_accuracy": 0.9460513457655907,
      "num_tokens": 568054.0,
      "step": 600
    },
    {
      "epoch": 2.060451255853555,
      "grad_norm": 0.8185880184173584,
      "learning_rate": 6.234357224118317e-05,
      "loss": 0.2028,
      "mean_token_accuracy": 0.9509358927607536,
      "num_tokens": 572684.0,
      "step": 605
    },
    {
      "epoch": 2.077479778629204,
      "grad_norm": 0.6658746600151062,
      "learning_rate": 6.120591581342436e-05,
      "loss": 0.2188,
      "mean_token_accuracy": 0.949580205976963,
      "num_tokens": 577466.0,
      "step": 610
    },
    {
      "epoch": 2.0945083014048533,
      "grad_norm": 0.8151549696922302,
      "learning_rate": 6.0068259385665536e-05,
      "loss": 0.2381,
      "mean_token_accuracy": 0.9443762481212616,
      "num_tokens": 582301.0,
      "step": 615
    },
    {
      "epoch": 2.111536824180502,
      "grad_norm": 0.5441663861274719,
      "learning_rate": 5.8930602957906714e-05,
      "loss": 0.2159,
      "mean_token_accuracy": 0.9448505908250808,
      "num_tokens": 587032.0,
      "step": 620
    },
    {
      "epoch": 2.1285653469561514,
      "grad_norm": 0.6140492558479309,
      "learning_rate": 5.77929465301479e-05,
      "loss": 0.2298,
      "mean_token_accuracy": 0.9426965191960335,
      "num_tokens": 591780.0,
      "step": 625
    },
    {
      "epoch": 2.1455938697318007,
      "grad_norm": 0.7466714978218079,
      "learning_rate": 5.665529010238908e-05,
      "loss": 0.2421,
      "mean_token_accuracy": 0.9439867153763771,
      "num_tokens": 596536.0,
      "step": 630
    },
    {
      "epoch": 2.16262239250745,
      "grad_norm": 0.6266794204711914,
      "learning_rate": 5.551763367463026e-05,
      "loss": 0.2192,
      "mean_token_accuracy": 0.948202782869339,
      "num_tokens": 601092.0,
      "step": 635
    },
    {
      "epoch": 2.179650915283099,
      "grad_norm": 1.0131019353866577,
      "learning_rate": 5.437997724687145e-05,
      "loss": 0.217,
      "mean_token_accuracy": 0.9489307940006256,
      "num_tokens": 605806.0,
      "step": 640
    },
    {
      "epoch": 2.1966794380587484,
      "grad_norm": 0.7090656161308289,
      "learning_rate": 5.324232081911263e-05,
      "loss": 0.2162,
      "mean_token_accuracy": 0.9507017344236374,
      "num_tokens": 610609.0,
      "step": 645
    },
    {
      "epoch": 2.2137079608343977,
      "grad_norm": 0.8066163659095764,
      "learning_rate": 5.210466439135382e-05,
      "loss": 0.2557,
      "mean_token_accuracy": 0.9380737692117691,
      "num_tokens": 615357.0,
      "step": 650
    },
    {
      "epoch": 2.230736483610047,
      "grad_norm": 0.7409054040908813,
      "learning_rate": 5.0967007963594995e-05,
      "loss": 0.2409,
      "mean_token_accuracy": 0.9453119024634361,
      "num_tokens": 620139.0,
      "step": 655
    },
    {
      "epoch": 2.247765006385696,
      "grad_norm": 0.782002866268158,
      "learning_rate": 4.982935153583618e-05,
      "loss": 0.2433,
      "mean_token_accuracy": 0.9432441100478173,
      "num_tokens": 624815.0,
      "step": 660
    },
    {
      "epoch": 2.264793529161345,
      "grad_norm": 0.6300613284111023,
      "learning_rate": 4.8691695108077365e-05,
      "loss": 0.2643,
      "mean_token_accuracy": 0.9443272948265076,
      "num_tokens": 629544.0,
      "step": 665
    },
    {
      "epoch": 2.2818220519369943,
      "grad_norm": 0.6463762521743774,
      "learning_rate": 4.755403868031855e-05,
      "loss": 0.2027,
      "mean_token_accuracy": 0.9486809656023979,
      "num_tokens": 634156.0,
      "step": 670
    },
    {
      "epoch": 2.2988505747126435,
      "grad_norm": 0.6527837514877319,
      "learning_rate": 4.641638225255973e-05,
      "loss": 0.2131,
      "mean_token_accuracy": 0.9498034939169884,
      "num_tokens": 638784.0,
      "step": 675
    },
    {
      "epoch": 2.315879097488293,
      "grad_norm": 0.7105056047439575,
      "learning_rate": 4.527872582480091e-05,
      "loss": 0.2292,
      "mean_token_accuracy": 0.9485589876770973,
      "num_tokens": 643477.0,
      "step": 680
    },
    {
      "epoch": 2.332907620263942,
      "grad_norm": 0.7441533803939819,
      "learning_rate": 4.41410693970421e-05,
      "loss": 0.2386,
      "mean_token_accuracy": 0.9443776711821557,
      "num_tokens": 648285.0,
      "step": 685
    },
    {
      "epoch": 2.3499361430395913,
      "grad_norm": 0.9468763470649719,
      "learning_rate": 4.300341296928328e-05,
      "loss": 0.2555,
      "mean_token_accuracy": 0.9390515595674515,
      "num_tokens": 653168.0,
      "step": 690
    },
    {
      "epoch": 2.3669646658152406,
      "grad_norm": 0.6610742211341858,
      "learning_rate": 4.186575654152446e-05,
      "loss": 0.2506,
      "mean_token_accuracy": 0.9435861468315124,
      "num_tokens": 657863.0,
      "step": 695
    },
    {
      "epoch": 2.38399318859089,
      "grad_norm": 0.9869385957717896,
      "learning_rate": 4.0728100113765646e-05,
      "loss": 0.2387,
      "mean_token_accuracy": 0.9400592625141144,
      "num_tokens": 662727.0,
      "step": 700
    },
    {
      "epoch": 2.401021711366539,
      "grad_norm": 0.7145782113075256,
      "learning_rate": 3.959044368600683e-05,
      "loss": 0.222,
      "mean_token_accuracy": 0.9447844803333283,
      "num_tokens": 667411.0,
      "step": 705
    },
    {
      "epoch": 2.4180502341421883,
      "grad_norm": 0.793662965297699,
      "learning_rate": 3.8452787258248016e-05,
      "loss": 0.2749,
      "mean_token_accuracy": 0.9354437962174416,
      "num_tokens": 672370.0,
      "step": 710
    },
    {
      "epoch": 2.4350787569178376,
      "grad_norm": 0.6723162531852722,
      "learning_rate": 3.7315130830489194e-05,
      "loss": 0.2199,
      "mean_token_accuracy": 0.949055691063404,
      "num_tokens": 676945.0,
      "step": 715
    },
    {
      "epoch": 2.4521072796934864,
      "grad_norm": 0.7849971652030945,
      "learning_rate": 3.617747440273038e-05,
      "loss": 0.2349,
      "mean_token_accuracy": 0.9447068989276886,
      "num_tokens": 681628.0,
      "step": 720
    },
    {
      "epoch": 2.4691358024691357,
      "grad_norm": 0.7382046580314636,
      "learning_rate": 3.5039817974971564e-05,
      "loss": 0.2298,
      "mean_token_accuracy": 0.9459434017539025,
      "num_tokens": 686342.0,
      "step": 725
    },
    {
      "epoch": 2.486164325244785,
      "grad_norm": 0.8477870225906372,
      "learning_rate": 3.390216154721274e-05,
      "loss": 0.2226,
      "mean_token_accuracy": 0.9453668460249901,
      "num_tokens": 691011.0,
      "step": 730
    },
    {
      "epoch": 2.503192848020434,
      "grad_norm": 0.7759321928024292,
      "learning_rate": 3.276450511945393e-05,
      "loss": 0.2051,
      "mean_token_accuracy": 0.9518747553229332,
      "num_tokens": 695661.0,
      "step": 735
    },
    {
      "epoch": 2.5202213707960834,
      "grad_norm": 0.7729942798614502,
      "learning_rate": 3.162684869169511e-05,
      "loss": 0.2196,
      "mean_token_accuracy": 0.9440933063626289,
      "num_tokens": 700360.0,
      "step": 740
    },
    {
      "epoch": 2.5372498935717327,
      "grad_norm": 0.8978192806243896,
      "learning_rate": 3.0489192263936294e-05,
      "loss": 0.2259,
      "mean_token_accuracy": 0.9469935670495033,
      "num_tokens": 705254.0,
      "step": 745
    },
    {
      "epoch": 2.554278416347382,
      "grad_norm": 0.8663476705551147,
      "learning_rate": 2.9351535836177476e-05,
      "loss": 0.239,
      "mean_token_accuracy": 0.9402577832341195,
      "num_tokens": 709984.0,
      "step": 750
    },
    {
      "epoch": 2.571306939123031,
      "grad_norm": 0.7382346391677856,
      "learning_rate": 2.8213879408418657e-05,
      "loss": 0.2145,
      "mean_token_accuracy": 0.9489600315690041,
      "num_tokens": 714606.0,
      "step": 755
    },
    {
      "epoch": 2.58833546189868,
      "grad_norm": 0.7550341486930847,
      "learning_rate": 2.7076222980659842e-05,
      "loss": 0.2303,
      "mean_token_accuracy": 0.9455819338560104,
      "num_tokens": 719367.0,
      "step": 760
    },
    {
      "epoch": 2.6053639846743293,
      "grad_norm": 0.6596015095710754,
      "learning_rate": 2.5938566552901024e-05,
      "loss": 0.2133,
      "mean_token_accuracy": 0.9476204797625541,
      "num_tokens": 724055.0,
      "step": 765
    },
    {
      "epoch": 2.6223925074499785,
      "grad_norm": 0.9189562201499939,
      "learning_rate": 2.480091012514221e-05,
      "loss": 0.2355,
      "mean_token_accuracy": 0.9445128008723259,
      "num_tokens": 728885.0,
      "step": 770
    },
    {
      "epoch": 2.639421030225628,
      "grad_norm": 0.8011082410812378,
      "learning_rate": 2.366325369738339e-05,
      "loss": 0.2193,
      "mean_token_accuracy": 0.9496066018939018,
      "num_tokens": 733655.0,
      "step": 775
    },
    {
      "epoch": 2.656449553001277,
      "grad_norm": 0.8749405741691589,
      "learning_rate": 2.2525597269624575e-05,
      "loss": 0.2571,
      "mean_token_accuracy": 0.9410566672682762,
      "num_tokens": 738508.0,
      "step": 780
    },
    {
      "epoch": 2.6734780757769263,
      "grad_norm": 0.7753709554672241,
      "learning_rate": 2.138794084186576e-05,
      "loss": 0.2281,
      "mean_token_accuracy": 0.9464891225099563,
      "num_tokens": 743232.0,
      "step": 785
    },
    {
      "epoch": 2.6905065985525756,
      "grad_norm": 0.7801012992858887,
      "learning_rate": 2.025028441410694e-05,
      "loss": 0.2284,
      "mean_token_accuracy": 0.9456154644489289,
      "num_tokens": 747910.0,
      "step": 790
    },
    {
      "epoch": 2.707535121328225,
      "grad_norm": 1.0080945491790771,
      "learning_rate": 1.9112627986348127e-05,
      "loss": 0.2242,
      "mean_token_accuracy": 0.9471994698047638,
      "num_tokens": 752624.0,
      "step": 795
    },
    {
      "epoch": 2.724563644103874,
      "grad_norm": 0.6461349129676819,
      "learning_rate": 1.7974971558589308e-05,
      "loss": 0.2158,
      "mean_token_accuracy": 0.9480220451951027,
      "num_tokens": 757321.0,
      "step": 800
    }
  ],
  "logging_steps": 5,
  "max_steps": 879,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.231180922051891e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}