{
  "best_metric": 0.097813181579113,
  "best_model_checkpoint": "idbwbase/checkpoint-42381",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 47090,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11,
      "grad_norm": 3.36433744430542,
      "learning_rate": 0.0001978764068804417,
      "loss": 0.4791,
      "step": 500
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.1407039165496826,
      "learning_rate": 0.0001957528137608834,
      "loss": 0.454,
      "step": 1000
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.483966112136841,
      "learning_rate": 0.00019362922064132513,
      "loss": 0.4364,
      "step": 1500
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.059077501296997,
      "learning_rate": 0.00019150987470800598,
      "loss": 0.4149,
      "step": 2000
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.1167092323303223,
      "learning_rate": 0.00018938628158844767,
      "loss": 0.41,
      "step": 2500
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.9980089664459229,
      "learning_rate": 0.00018726268846888937,
      "loss": 0.3953,
      "step": 3000
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.5476139783859253,
      "learning_rate": 0.00018513909534933107,
      "loss": 0.3892,
      "step": 3500
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.8769443035125732,
      "learning_rate": 0.0001830197494160119,
      "loss": 0.3826,
      "step": 4000
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.0475404262542725,
      "learning_rate": 0.0001808961562964536,
      "loss": 0.373,
      "step": 4500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.870536469791843,
      "eval_loss": 0.30019834637641907,
      "eval_runtime": 363.8544,
      "eval_samples_per_second": 146.16,
      "eval_steps_per_second": 18.271,
      "step": 4709
    },
    {
      "epoch": 1.06,
      "grad_norm": 3.492917776107788,
      "learning_rate": 0.0001787725631768953,
      "loss": 0.3639,
      "step": 5000
    },
    {
      "epoch": 1.17,
      "grad_norm": 3.2921090126037598,
      "learning_rate": 0.00017664897005733704,
      "loss": 0.3553,
      "step": 5500
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.0384316444396973,
      "learning_rate": 0.00017452537693777873,
      "loss": 0.3499,
      "step": 6000
    },
    {
      "epoch": 1.38,
      "grad_norm": 2.055048942565918,
      "learning_rate": 0.00017240178381822043,
      "loss": 0.3506,
      "step": 6500
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.965047001838684,
      "learning_rate": 0.00017027819069866215,
      "loss": 0.3423,
      "step": 7000
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.3037126064300537,
      "learning_rate": 0.00016815459757910385,
      "loss": 0.3402,
      "step": 7500
    },
    {
      "epoch": 1.7,
      "grad_norm": 2.647536516189575,
      "learning_rate": 0.00016603100445954555,
      "loss": 0.3323,
      "step": 8000
    },
    {
      "epoch": 1.81,
      "grad_norm": 2.767207384109497,
      "learning_rate": 0.0001639116585262264,
      "loss": 0.3294,
      "step": 8500
    },
    {
      "epoch": 1.91,
      "grad_norm": 2.31170654296875,
      "learning_rate": 0.00016179231259290722,
      "loss": 0.3244,
      "step": 9000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9044395554803407,
      "eval_loss": 0.22620604932308197,
      "eval_runtime": 365.6085,
      "eval_samples_per_second": 145.459,
      "eval_steps_per_second": 18.183,
      "step": 9418
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.986091136932373,
      "learning_rate": 0.00015966871947334892,
      "loss": 0.3175,
      "step": 9500
    },
    {
      "epoch": 2.12,
      "grad_norm": 2.7250773906707764,
      "learning_rate": 0.0001575451263537906,
      "loss": 0.3038,
      "step": 10000
    },
    {
      "epoch": 2.23,
      "grad_norm": 2.0195140838623047,
      "learning_rate": 0.0001554215332342323,
      "loss": 0.3037,
      "step": 10500
    },
    {
      "epoch": 2.34,
      "grad_norm": 2.9471521377563477,
      "learning_rate": 0.00015329794011467404,
      "loss": 0.2911,
      "step": 11000
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.8268228769302368,
      "learning_rate": 0.00015117434699511573,
      "loss": 0.2931,
      "step": 11500
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.0443902015686035,
      "learning_rate": 0.00014905075387555746,
      "loss": 0.2903,
      "step": 12000
    },
    {
      "epoch": 2.65,
      "grad_norm": 2.8994696140289307,
      "learning_rate": 0.00014692716075599915,
      "loss": 0.2864,
      "step": 12500
    },
    {
      "epoch": 2.76,
      "grad_norm": 2.1983399391174316,
      "learning_rate": 0.00014480781482267998,
      "loss": 0.2833,
      "step": 13000
    },
    {
      "epoch": 2.87,
      "grad_norm": 2.4135706424713135,
      "learning_rate": 0.0001426884688893608,
      "loss": 0.2788,
      "step": 13500
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.9159047603607178,
      "learning_rate": 0.0001405648757698025,
      "loss": 0.2801,
      "step": 14000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9195953442018766,
      "eval_loss": 0.1986866444349289,
      "eval_runtime": 365.1106,
      "eval_samples_per_second": 145.657,
      "eval_steps_per_second": 18.208,
      "step": 14127
    },
    {
      "epoch": 3.08,
      "grad_norm": 2.0817108154296875,
      "learning_rate": 0.00013844128265024422,
      "loss": 0.2584,
      "step": 14500
    },
    {
      "epoch": 3.19,
      "grad_norm": 2.8534085750579834,
      "learning_rate": 0.00013631768953068594,
      "loss": 0.2559,
      "step": 15000
    },
    {
      "epoch": 3.29,
      "grad_norm": 3.98779559135437,
      "learning_rate": 0.00013419409641112764,
      "loss": 0.2511,
      "step": 15500
    },
    {
      "epoch": 3.4,
      "grad_norm": 3.104475975036621,
      "learning_rate": 0.00013207475047780846,
      "loss": 0.2517,
      "step": 16000
    },
    {
      "epoch": 3.5,
      "grad_norm": 2.7032172679901123,
      "learning_rate": 0.00012995115735825016,
      "loss": 0.2504,
      "step": 16500
    },
    {
      "epoch": 3.61,
      "grad_norm": 3.3657124042510986,
      "learning_rate": 0.00012782756423869186,
      "loss": 0.2461,
      "step": 17000
    },
    {
      "epoch": 3.72,
      "grad_norm": 2.80549693107605,
      "learning_rate": 0.00012570397111913358,
      "loss": 0.2451,
      "step": 17500
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.6164408922195435,
      "learning_rate": 0.0001235803779995753,
      "loss": 0.2421,
      "step": 18000
    },
    {
      "epoch": 3.93,
      "grad_norm": 2.589157819747925,
      "learning_rate": 0.00012145678488001699,
      "loss": 0.2366,
      "step": 18500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9344502735939527,
      "eval_loss": 0.17884854972362518,
      "eval_runtime": 360.8393,
      "eval_samples_per_second": 147.381,
      "eval_steps_per_second": 18.424,
      "step": 18836
    },
    {
      "epoch": 4.03,
      "grad_norm": 2.659468412399292,
      "learning_rate": 0.0001193331917604587,
      "loss": 0.2287,
      "step": 19000
    },
    {
      "epoch": 4.14,
      "grad_norm": 2.452038526535034,
      "learning_rate": 0.00011720959864090042,
      "loss": 0.2205,
      "step": 19500
    },
    {
      "epoch": 4.25,
      "grad_norm": 3.7109389305114746,
      "learning_rate": 0.00011509025270758123,
      "loss": 0.2076,
      "step": 20000
    },
    {
      "epoch": 4.35,
      "grad_norm": 2.5269851684570312,
      "learning_rate": 0.00011296665958802296,
      "loss": 0.2134,
      "step": 20500
    },
    {
      "epoch": 4.46,
      "grad_norm": 2.7143073081970215,
      "learning_rate": 0.00011084306646846465,
      "loss": 0.2144,
      "step": 21000
    },
    {
      "epoch": 4.57,
      "grad_norm": 3.1681575775146484,
      "learning_rate": 0.00010871947334890635,
      "loss": 0.2091,
      "step": 21500
    },
    {
      "epoch": 4.67,
      "grad_norm": 1.2098946571350098,
      "learning_rate": 0.00010659588022934806,
      "loss": 0.2065,
      "step": 22000
    },
    {
      "epoch": 4.78,
      "grad_norm": 1.6173245906829834,
      "learning_rate": 0.00010447228710978979,
      "loss": 0.2107,
      "step": 22500
    },
    {
      "epoch": 4.88,
      "grad_norm": 4.180131435394287,
      "learning_rate": 0.00010235294117647058,
      "loss": 0.2053,
      "step": 23000
    },
    {
      "epoch": 4.99,
      "grad_norm": 1.9634019136428833,
      "learning_rate": 0.0001002293480569123,
      "loss": 0.2051,
      "step": 23500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9483650175814671,
      "eval_loss": 0.1462855190038681,
      "eval_runtime": 362.8319,
      "eval_samples_per_second": 146.572,
      "eval_steps_per_second": 18.323,
      "step": 23545
    },
    {
      "epoch": 5.1,
      "grad_norm": 2.8337090015411377,
      "learning_rate": 9.810575493735401e-05,
      "loss": 0.1896,
      "step": 24000
    },
    {
      "epoch": 5.2,
      "grad_norm": 1.7976356744766235,
      "learning_rate": 9.598216181779571e-05,
      "loss": 0.1872,
      "step": 24500
    },
    {
      "epoch": 5.31,
      "grad_norm": 2.647800922393799,
      "learning_rate": 9.385856869823742e-05,
      "loss": 0.1819,
      "step": 25000
    },
    {
      "epoch": 5.42,
      "grad_norm": 2.1115944385528564,
      "learning_rate": 9.173922276491824e-05,
      "loss": 0.1837,
      "step": 25500
    },
    {
      "epoch": 5.52,
      "grad_norm": 3.922607183456421,
      "learning_rate": 8.961562964535995e-05,
      "loss": 0.1868,
      "step": 26000
    },
    {
      "epoch": 5.63,
      "grad_norm": 1.7281802892684937,
      "learning_rate": 8.749203652580165e-05,
      "loss": 0.1832,
      "step": 26500
    },
    {
      "epoch": 5.73,
      "grad_norm": 2.4881973266601562,
      "learning_rate": 8.536844340624338e-05,
      "loss": 0.1791,
      "step": 27000
    },
    {
      "epoch": 5.84,
      "grad_norm": 1.2277300357818604,
      "learning_rate": 8.324909747292418e-05,
      "loss": 0.1805,
      "step": 27500
    },
    {
      "epoch": 5.95,
      "grad_norm": 1.151065468788147,
      "learning_rate": 8.112975153960502e-05,
      "loss": 0.1764,
      "step": 28000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9593087756905662,
      "eval_loss": 0.1202462688088417,
      "eval_runtime": 359.7308,
      "eval_samples_per_second": 147.836,
      "eval_steps_per_second": 18.48,
      "step": 28254
    },
    {
      "epoch": 6.05,
      "grad_norm": 2.2972559928894043,
      "learning_rate": 7.900615842004673e-05,
      "loss": 0.1694,
      "step": 28500
    },
    {
      "epoch": 6.16,
      "grad_norm": 1.6081655025482178,
      "learning_rate": 7.688256530048843e-05,
      "loss": 0.1656,
      "step": 29000
    },
    {
      "epoch": 6.26,
      "grad_norm": 4.132796764373779,
      "learning_rate": 7.475897218093014e-05,
      "loss": 0.1589,
      "step": 29500
    },
    {
      "epoch": 6.37,
      "grad_norm": 2.355618476867676,
      "learning_rate": 7.263537906137185e-05,
      "loss": 0.1618,
      "step": 30000
    },
    {
      "epoch": 6.48,
      "grad_norm": 2.159538745880127,
      "learning_rate": 7.051178594181356e-05,
      "loss": 0.1572,
      "step": 30500
    },
    {
      "epoch": 6.58,
      "grad_norm": 3.115480661392212,
      "learning_rate": 6.838819282225526e-05,
      "loss": 0.163,
      "step": 31000
    },
    {
      "epoch": 6.69,
      "grad_norm": 2.2360424995422363,
      "learning_rate": 6.626459970269697e-05,
      "loss": 0.1607,
      "step": 31500
    },
    {
      "epoch": 6.8,
      "grad_norm": 2.970757246017456,
      "learning_rate": 6.414525376937779e-05,
      "loss": 0.1599,
      "step": 32000
    },
    {
      "epoch": 6.9,
      "grad_norm": 6.401217460632324,
      "learning_rate": 6.20216606498195e-05,
      "loss": 0.1595,
      "step": 32500
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9654763919444914,
      "eval_loss": 0.12429220974445343,
      "eval_runtime": 360.8363,
      "eval_samples_per_second": 147.383,
      "eval_steps_per_second": 18.424,
      "step": 32963
    },
    {
      "epoch": 7.01,
      "grad_norm": 3.2262797355651855,
      "learning_rate": 5.9898067530261204e-05,
      "loss": 0.1528,
      "step": 33000
    },
    {
      "epoch": 7.11,
      "grad_norm": 3.578483819961548,
      "learning_rate": 5.7774474410702914e-05,
      "loss": 0.1473,
      "step": 33500
    },
    {
      "epoch": 7.22,
      "grad_norm": 3.9829201698303223,
      "learning_rate": 5.565088129114462e-05,
      "loss": 0.1441,
      "step": 34000
    },
    {
      "epoch": 7.33,
      "grad_norm": 3.7269842624664307,
      "learning_rate": 5.3531535357825446e-05,
      "loss": 0.1385,
      "step": 34500
    },
    {
      "epoch": 7.43,
      "grad_norm": 2.6239781379699707,
      "learning_rate": 5.140794223826715e-05,
      "loss": 0.1406,
      "step": 35000
    },
    {
      "epoch": 7.54,
      "grad_norm": 1.6224650144577026,
      "learning_rate": 4.9284349118708855e-05,
      "loss": 0.1429,
      "step": 35500
    },
    {
      "epoch": 7.64,
      "grad_norm": 5.391174793243408,
      "learning_rate": 4.7160755999150565e-05,
      "loss": 0.1457,
      "step": 36000
    },
    {
      "epoch": 7.75,
      "grad_norm": 5.336463928222656,
      "learning_rate": 4.504141006583139e-05,
      "loss": 0.1359,
      "step": 36500
    },
    {
      "epoch": 7.86,
      "grad_norm": 3.005934238433838,
      "learning_rate": 4.29178169462731e-05,
      "loss": 0.1417,
      "step": 37000
    },
    {
      "epoch": 7.96,
      "grad_norm": 3.7323451042175293,
      "learning_rate": 4.079422382671481e-05,
      "loss": 0.1359,
      "step": 37500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9658712698144074,
      "eval_loss": 0.11875798553228378,
      "eval_runtime": 368.3038,
      "eval_samples_per_second": 144.394,
      "eval_steps_per_second": 18.05,
      "step": 37672
    },
    {
      "epoch": 8.07,
      "grad_norm": 2.7839584350585938,
      "learning_rate": 3.867063070715651e-05,
      "loss": 0.1335,
      "step": 38000
    },
    {
      "epoch": 8.18,
      "grad_norm": 4.733059406280518,
      "learning_rate": 3.6551284773837333e-05,
      "loss": 0.1313,
      "step": 38500
    },
    {
      "epoch": 8.28,
      "grad_norm": 2.712228536605835,
      "learning_rate": 3.4427691654279044e-05,
      "loss": 0.1325,
      "step": 39000
    },
    {
      "epoch": 8.39,
      "grad_norm": 4.026779651641846,
      "learning_rate": 3.230409853472075e-05,
      "loss": 0.1272,
      "step": 39500
    },
    {
      "epoch": 8.49,
      "grad_norm": 3.8425519466400146,
      "learning_rate": 3.018050541516246e-05,
      "loss": 0.1307,
      "step": 40000
    },
    {
      "epoch": 8.6,
      "grad_norm": 3.5189483165740967,
      "learning_rate": 2.806115948184328e-05,
      "loss": 0.129,
      "step": 40500
    },
    {
      "epoch": 8.71,
      "grad_norm": 4.5053486824035645,
      "learning_rate": 2.5937566362284988e-05,
      "loss": 0.1257,
      "step": 41000
    },
    {
      "epoch": 8.81,
      "grad_norm": 4.372600078582764,
      "learning_rate": 2.3818220428965812e-05,
      "loss": 0.1238,
      "step": 41500
    },
    {
      "epoch": 8.92,
      "grad_norm": 1.8681597709655762,
      "learning_rate": 2.169462730940752e-05,
      "loss": 0.1231,
      "step": 42000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9741072939583686,
      "eval_loss": 0.097813181579113,
      "eval_runtime": 359.2297,
      "eval_samples_per_second": 148.042,
      "eval_steps_per_second": 18.506,
      "step": 42381
    },
    {
      "epoch": 9.03,
      "grad_norm": 3.127023220062256,
      "learning_rate": 1.9571034189849227e-05,
      "loss": 0.1249,
      "step": 42500
    },
    {
      "epoch": 9.13,
      "grad_norm": 1.650771141052246,
      "learning_rate": 1.744744107029093e-05,
      "loss": 0.118,
      "step": 43000
    },
    {
      "epoch": 9.24,
      "grad_norm": 3.3975789546966553,
      "learning_rate": 1.532384795073264e-05,
      "loss": 0.1198,
      "step": 43500
    },
    {
      "epoch": 9.34,
      "grad_norm": 1.8186984062194824,
      "learning_rate": 1.3200254831174346e-05,
      "loss": 0.1167,
      "step": 44000
    },
    {
      "epoch": 9.45,
      "grad_norm": 7.110922336578369,
      "learning_rate": 1.1076661711616054e-05,
      "loss": 0.1198,
      "step": 44500
    },
    {
      "epoch": 9.56,
      "grad_norm": 6.389682292938232,
      "learning_rate": 8.953068592057761e-06,
      "loss": 0.1187,
      "step": 45000
    },
    {
      "epoch": 9.66,
      "grad_norm": 2.857072353363037,
      "learning_rate": 6.833722658738587e-06,
      "loss": 0.1129,
      "step": 45500
    },
    {
      "epoch": 9.77,
      "grad_norm": 0.7192772626876831,
      "learning_rate": 4.710129539180293e-06,
      "loss": 0.1133,
      "step": 46000
    },
    {
      "epoch": 9.87,
      "grad_norm": 3.4914393424987793,
      "learning_rate": 2.5865364196220007e-06,
      "loss": 0.1151,
      "step": 46500
    },
    {
      "epoch": 9.98,
      "grad_norm": 1.8187050819396973,
      "learning_rate": 4.671904863028244e-07,
      "loss": 0.1162,
      "step": 47000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9761004870160396,
      "eval_loss": 0.10011033713817596,
      "eval_runtime": 359.0659,
      "eval_samples_per_second": 148.109,
      "eval_steps_per_second": 18.515,
      "step": 47090
    },
    {
      "epoch": 10.0,
      "step": 47090,
      "total_flos": 2.383420595495315e+20,
      "train_loss": 0.22288281881624433,
      "train_runtime": 60913.5069,
      "train_samples_per_second": 49.473,
      "train_steps_per_second": 0.773
    }
  ],
  "logging_steps": 500,
  "max_steps": 47090,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 2.383420595495315e+20,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}