{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 1036,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06791171477079797,
      "grad_norm": 39.06894209264497,
      "learning_rate": 9.000000000000001e-07,
      "loss": 2.5672,
      "step": 10
    },
    {
      "epoch": 0.13582342954159593,
      "grad_norm": 9.408641592679151,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 1.6271,
      "step": 20
    },
    {
      "epoch": 0.2037351443123939,
      "grad_norm": 3.5736160912411656,
      "learning_rate": 2.9e-06,
      "loss": 1.0035,
      "step": 30
    },
    {
      "epoch": 0.27164685908319186,
      "grad_norm": 3.159979908410021,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.7978,
      "step": 40
    },
    {
      "epoch": 0.3395585738539898,
      "grad_norm": 3.23976253664356,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.7342,
      "step": 50
    },
    {
      "epoch": 0.4074702886247878,
      "grad_norm": 2.5239516324886613,
      "learning_rate": 5.9e-06,
      "loss": 0.6387,
      "step": 60
    },
    {
      "epoch": 0.47538200339558573,
      "grad_norm": 2.864395183943674,
      "learning_rate": 6.9e-06,
      "loss": 0.6566,
      "step": 70
    },
    {
      "epoch": 0.5432937181663837,
      "grad_norm": 2.221989312084371,
      "learning_rate": 7.9e-06,
      "loss": 0.6369,
      "step": 80
    },
    {
      "epoch": 0.6112054329371817,
      "grad_norm": 2.622798646579124,
      "learning_rate": 8.900000000000001e-06,
      "loss": 0.5945,
      "step": 90
    },
    {
      "epoch": 0.6791171477079796,
      "grad_norm": 2.727376011862893,
      "learning_rate": 9.9e-06,
      "loss": 0.598,
      "step": 100
    },
    {
      "epoch": 0.7470288624787776,
      "grad_norm": 2.588442039618945,
      "learning_rate": 9.997718922447669e-06,
      "loss": 0.569,
      "step": 110
    },
    {
      "epoch": 0.8149405772495756,
      "grad_norm": 2.3467194997795335,
      "learning_rate": 9.989836388165012e-06,
      "loss": 0.5606,
      "step": 120
    },
    {
      "epoch": 0.8828522920203735,
      "grad_norm": 2.3847506796962556,
      "learning_rate": 9.976333112771488e-06,
      "loss": 0.5565,
      "step": 130
    },
    {
      "epoch": 0.9507640067911715,
      "grad_norm": 2.1204168574702957,
      "learning_rate": 9.957224306869053e-06,
      "loss": 0.5414,
      "step": 140
    },
    {
      "epoch": 1.0135823429541595,
      "grad_norm": 2.0962709577545704,
      "learning_rate": 9.932531495342005e-06,
      "loss": 0.5142,
      "step": 150
    },
    {
      "epoch": 1.0814940577249577,
      "grad_norm": 2.420185175502751,
      "learning_rate": 9.90228249311053e-06,
      "loss": 0.4346,
      "step": 160
    },
    {
      "epoch": 1.1494057724957556,
      "grad_norm": 2.175482934612183,
      "learning_rate": 9.86651137379893e-06,
      "loss": 0.4156,
      "step": 170
    },
    {
      "epoch": 1.2173174872665535,
      "grad_norm": 2.015548954416845,
      "learning_rate": 9.82525843135379e-06,
      "loss": 0.4199,
      "step": 180
    },
    {
      "epoch": 1.2852292020373515,
      "grad_norm": 2.2212942924087415,
      "learning_rate": 9.778570134655351e-06,
      "loss": 0.4088,
      "step": 190
    },
    {
      "epoch": 1.3531409168081494,
      "grad_norm": 2.2774181484385014,
      "learning_rate": 9.726499075173201e-06,
      "loss": 0.4198,
      "step": 200
    },
    {
      "epoch": 1.4210526315789473,
      "grad_norm": 1.951548193520802,
      "learning_rate": 9.669103907725243e-06,
      "loss": 0.4203,
      "step": 210
    },
    {
      "epoch": 1.4889643463497453,
      "grad_norm": 2.0166680690881154,
      "learning_rate": 9.606449284406685e-06,
      "loss": 0.4288,
      "step": 220
    },
    {
      "epoch": 1.5568760611205432,
      "grad_norm": 2.1215872162254445,
      "learning_rate": 9.538605781763464e-06,
      "loss": 0.403,
      "step": 230
    },
    {
      "epoch": 1.6247877758913414,
      "grad_norm": 2.21540975350655,
      "learning_rate": 9.465649821292148e-06,
      "loss": 0.4014,
      "step": 240
    },
    {
      "epoch": 1.692699490662139,
      "grad_norm": 2.2219332688595976,
      "learning_rate": 9.387663583355862e-06,
      "loss": 0.4104,
      "step": 250
    },
    {
      "epoch": 1.7606112054329373,
      "grad_norm": 2.139070251547491,
      "learning_rate": 9.3047349146132e-06,
      "loss": 0.4106,
      "step": 260
    },
    {
      "epoch": 1.8285229202037352,
      "grad_norm": 2.114857623710178,
      "learning_rate": 9.21695722906443e-06,
      "loss": 0.4055,
      "step": 270
    },
    {
      "epoch": 1.8964346349745331,
      "grad_norm": 2.0517966830796586,
      "learning_rate": 9.124429402826398e-06,
      "loss": 0.4238,
      "step": 280
    },
    {
      "epoch": 1.964346349745331,
      "grad_norm": 1.8301929541835302,
      "learning_rate": 9.02725566275473e-06,
      "loss": 0.3975,
      "step": 290
    },
    {
      "epoch": 2.027164685908319,
      "grad_norm": 1.61769307419241,
      "learning_rate": 8.92554546903875e-06,
      "loss": 0.347,
      "step": 300
    },
    {
      "epoch": 2.095076400679117,
      "grad_norm": 1.7835414384395631,
      "learning_rate": 8.819413391901353e-06,
      "loss": 0.2279,
      "step": 310
    },
    {
      "epoch": 2.1629881154499153,
      "grad_norm": 1.7449573933209457,
      "learning_rate": 8.708978982542766e-06,
      "loss": 0.2456,
      "step": 320
    },
    {
      "epoch": 2.230899830220713,
      "grad_norm": 1.8762842446890258,
      "learning_rate": 8.59436663847354e-06,
      "loss": 0.2394,
      "step": 330
    },
    {
      "epoch": 2.298811544991511,
      "grad_norm": 1.7977301694269836,
      "learning_rate": 8.475705463388467e-06,
      "loss": 0.2537,
      "step": 340
    },
    {
      "epoch": 2.366723259762309,
      "grad_norm": 1.693865524446407,
      "learning_rate": 8.35312912173928e-06,
      "loss": 0.2555,
      "step": 350
    },
    {
      "epoch": 2.434634974533107,
      "grad_norm": 1.9696999644652728,
      "learning_rate": 8.226775688169943e-06,
      "loss": 0.2424,
      "step": 360
    },
    {
      "epoch": 2.502546689303905,
      "grad_norm": 1.7928317986379525,
      "learning_rate": 8.096787491984128e-06,
      "loss": 0.2458,
      "step": 370
    },
    {
      "epoch": 2.570458404074703,
      "grad_norm": 1.6035422026428618,
      "learning_rate": 7.963310956820085e-06,
      "loss": 0.2545,
      "step": 380
    },
    {
      "epoch": 2.6383701188455007,
      "grad_norm": 1.7086388697991841,
      "learning_rate": 7.826496435713497e-06,
      "loss": 0.2544,
      "step": 390
    },
    {
      "epoch": 2.706281833616299,
      "grad_norm": 1.7784954050779234,
      "learning_rate": 7.686498041734121e-06,
      "loss": 0.2587,
      "step": 400
    },
    {
      "epoch": 2.774193548387097,
      "grad_norm": 1.8489681135255784,
      "learning_rate": 7.543473474386962e-06,
      "loss": 0.2537,
      "step": 410
    },
    {
      "epoch": 2.8421052631578947,
      "grad_norm": 1.6154424089494572,
      "learning_rate": 7.397583841973584e-06,
      "loss": 0.2549,
      "step": 420
    },
    {
      "epoch": 2.910016977928693,
      "grad_norm": 1.6945683455240643,
      "learning_rate": 7.248993480113585e-06,
      "loss": 0.2576,
      "step": 430
    },
    {
      "epoch": 2.9779286926994906,
      "grad_norm": 1.7774529548842273,
      "learning_rate": 7.09786976663073e-06,
      "loss": 0.2597,
      "step": 440
    },
    {
      "epoch": 3.0407470288624787,
      "grad_norm": 1.3291164143277696,
      "learning_rate": 6.944382933012207e-06,
      "loss": 0.1914,
      "step": 450
    },
    {
      "epoch": 3.108658743633277,
      "grad_norm": 1.4399162979563198,
      "learning_rate": 6.788705872653429e-06,
      "loss": 0.1304,
      "step": 460
    },
    {
      "epoch": 3.1765704584040746,
      "grad_norm": 1.4090235140542846,
      "learning_rate": 6.631013946104348e-06,
      "loss": 0.1348,
      "step": 470
    },
    {
      "epoch": 3.2444821731748728,
      "grad_norm": 1.5024124740765719,
      "learning_rate": 6.4714847835366625e-06,
      "loss": 0.1306,
      "step": 480
    },
    {
      "epoch": 3.3123938879456705,
      "grad_norm": 1.4037148184869794,
      "learning_rate": 6.310298084654467e-06,
      "loss": 0.1341,
      "step": 490
    },
    {
      "epoch": 3.3803056027164686,
      "grad_norm": 1.5382989963970903,
      "learning_rate": 6.147635416273679e-06,
      "loss": 0.138,
      "step": 500
    },
    {
      "epoch": 3.4482173174872663,
      "grad_norm": 1.4075458945664348,
      "learning_rate": 5.983680007798277e-06,
      "loss": 0.1367,
      "step": 510
    },
    {
      "epoch": 3.5161290322580645,
      "grad_norm": 1.379295875695038,
      "learning_rate": 5.818616544823755e-06,
      "loss": 0.1341,
      "step": 520
    },
    {
      "epoch": 3.5840407470288627,
      "grad_norm": 1.6516703780226065,
      "learning_rate": 5.65263096110026e-06,
      "loss": 0.1417,
      "step": 530
    },
    {
      "epoch": 3.6519524617996604,
      "grad_norm": 1.4236024491217225,
      "learning_rate": 5.485910229089739e-06,
      "loss": 0.1432,
      "step": 540
    },
    {
      "epoch": 3.7198641765704585,
      "grad_norm": 1.5628385416315445,
      "learning_rate": 5.3186421493530845e-06,
      "loss": 0.1378,
      "step": 550
    },
    {
      "epoch": 3.7877758913412563,
      "grad_norm": 1.5399714943796112,
      "learning_rate": 5.151015139004445e-06,
      "loss": 0.1396,
      "step": 560
    },
    {
      "epoch": 3.8556876061120544,
      "grad_norm": 1.5044051222954915,
      "learning_rate": 4.983218019471027e-06,
      "loss": 0.1443,
      "step": 570
    },
    {
      "epoch": 3.923599320882852,
      "grad_norm": 1.2731582701152808,
      "learning_rate": 4.815439803797483e-06,
      "loss": 0.1328,
      "step": 580
    },
    {
      "epoch": 3.9915110356536503,
      "grad_norm": 1.2890321476898117,
      "learning_rate": 4.647869483734441e-06,
      "loss": 0.1334,
      "step": 590
    },
    {
      "epoch": 4.054329371816638,
      "grad_norm": 1.3973597889025486,
      "learning_rate": 4.4806958168510235e-06,
      "loss": 0.0788,
      "step": 600
    },
    {
      "epoch": 4.122241086587437,
      "grad_norm": 1.1936189947404265,
      "learning_rate": 4.314107113911151e-06,
      "loss": 0.062,
      "step": 610
    },
    {
      "epoch": 4.190152801358234,
      "grad_norm": 1.091238397517897,
      "learning_rate": 4.148291026753159e-06,
      "loss": 0.0649,
      "step": 620
    },
    {
      "epoch": 4.258064516129032,
      "grad_norm": 1.1958745470057501,
      "learning_rate": 3.983434336911636e-06,
      "loss": 0.0662,
      "step": 630
    },
    {
      "epoch": 4.325976230899831,
      "grad_norm": 1.1212711189708877,
      "learning_rate": 3.819722745219617e-06,
      "loss": 0.0633,
      "step": 640
    },
    {
      "epoch": 4.393887945670628,
      "grad_norm": 1.3264229165705554,
      "learning_rate": 3.657340662628116e-06,
      "loss": 0.0662,
      "step": 650
    },
    {
      "epoch": 4.461799660441426,
      "grad_norm": 1.1770811940347883,
      "learning_rate": 3.4964710024786354e-06,
      "loss": 0.0625,
      "step": 660
    },
    {
      "epoch": 4.529711375212224,
      "grad_norm": 1.2943958821442108,
      "learning_rate": 3.337294974462623e-06,
      "loss": 0.0628,
      "step": 670
    },
    {
      "epoch": 4.597623089983022,
      "grad_norm": 1.345364224404323,
      "learning_rate": 3.1799918805000097e-06,
      "loss": 0.0655,
      "step": 680
    },
    {
      "epoch": 4.66553480475382,
      "grad_norm": 1.0977379456012588,
      "learning_rate": 3.024738912766717e-06,
      "loss": 0.0616,
      "step": 690
    },
    {
      "epoch": 4.733446519524618,
      "grad_norm": 1.1868517773950296,
      "learning_rate": 2.87171095409866e-06,
      "loss": 0.0614,
      "step": 700
    },
    {
      "epoch": 4.8013582342954155,
      "grad_norm": 1.110887449804357,
      "learning_rate": 2.7210803809970853e-06,
      "loss": 0.0635,
      "step": 710
    },
    {
      "epoch": 4.869269949066214,
      "grad_norm": 1.2822194078234743,
      "learning_rate": 2.5730168694571457e-06,
      "loss": 0.0614,
      "step": 720
    },
    {
      "epoch": 4.937181663837012,
      "grad_norm": 1.0805554157065451,
      "learning_rate": 2.4276872038384037e-06,
      "loss": 0.0598,
      "step": 730
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.320950039949113,
      "learning_rate": 2.2852550889926067e-06,
      "loss": 0.0595,
      "step": 740
    },
    {
      "epoch": 5.067911714770798,
      "grad_norm": 0.7541720353663811,
      "learning_rate": 2.1458809658603288e-06,
      "loss": 0.024,
      "step": 750
    },
    {
      "epoch": 5.135823429541596,
      "grad_norm": 0.9116539389923982,
      "learning_rate": 2.009721830744201e-06,
      "loss": 0.022,
      "step": 760
    },
    {
      "epoch": 5.203735144312394,
      "grad_norm": 0.8703094574463964,
      "learning_rate": 1.87693105846232e-06,
      "loss": 0.0214,
      "step": 770
    },
    {
      "epoch": 5.271646859083192,
      "grad_norm": 0.8382504818211445,
      "learning_rate": 1.7476582295810314e-06,
      "loss": 0.0224,
      "step": 780
    },
    {
      "epoch": 5.3395585738539895,
      "grad_norm": 0.8489442445106106,
      "learning_rate": 1.6220489619216988e-06,
      "loss": 0.0212,
      "step": 790
    },
    {
      "epoch": 5.407470288624788,
      "grad_norm": 0.8903040566842944,
      "learning_rate": 1.5002447465312675e-06,
      "loss": 0.0213,
      "step": 800
    },
    {
      "epoch": 5.475382003395586,
      "grad_norm": 0.9430023584705125,
      "learning_rate": 1.3823827883013723e-06,
      "loss": 0.0201,
      "step": 810
    },
    {
      "epoch": 5.5432937181663835,
      "grad_norm": 0.877235765811111,
      "learning_rate": 1.2685958514155456e-06,
      "loss": 0.0214,
      "step": 820
    },
    {
      "epoch": 5.611205432937181,
      "grad_norm": 1.0592040952393595,
      "learning_rate": 1.159012109798598e-06,
      "loss": 0.0213,
      "step": 830
    },
    {
      "epoch": 5.67911714770798,
      "grad_norm": 0.9416315998080429,
      "learning_rate": 1.0537550027366444e-06,
      "loss": 0.0191,
      "step": 840
    },
    {
      "epoch": 5.7470288624787775,
      "grad_norm": 0.8017668057640382,
      "learning_rate": 9.529430958304092e-07,
      "loss": 0.0206,
      "step": 850
    },
    {
      "epoch": 5.814940577249575,
      "grad_norm": 0.8439085846162264,
      "learning_rate": 8.566899474384299e-07,
      "loss": 0.0192,
      "step": 860
    },
    {
      "epoch": 5.882852292020374,
      "grad_norm": 0.9517820142849852,
      "learning_rate": 7.651039807606137e-07,
      "loss": 0.0213,
      "step": 870
    },
    {
      "epoch": 5.950764006791172,
      "grad_norm": 0.9056469249009282,
      "learning_rate": 6.782883617062286e-07,
      "loss": 0.0189,
      "step": 880
    },
    {
      "epoch": 6.01358234295416,
      "grad_norm": 0.4576722465983699,
      "learning_rate": 5.963408826839079e-07,
      "loss": 0.0157,
      "step": 890
    },
    {
      "epoch": 6.081494057724957,
      "grad_norm": 0.28634320188291257,
      "learning_rate": 5.19353852444569e-07,
      "loss": 0.0064,
      "step": 900
    },
    {
      "epoch": 6.149405772495755,
      "grad_norm": 0.4684376300431715,
      "learning_rate": 4.4741399210133154e-07,
      "loss": 0.0063,
      "step": 910
    },
    {
      "epoch": 6.217317487266554,
      "grad_norm": 0.30013129798891736,
      "learning_rate": 3.8060233744356634e-07,
      "loss": 0.0054,
      "step": 920
    },
    {
      "epoch": 6.2852292020373515,
      "grad_norm": 0.5743674767243917,
      "learning_rate": 3.189941476551034e-07,
      "loss": 0.0059,
      "step": 930
    },
    {
      "epoch": 6.353140916808149,
      "grad_norm": 0.42277606258594225,
      "learning_rate": 2.6265882053943204e-07,
      "loss": 0.0055,
      "step": 940
    },
    {
      "epoch": 6.421052631578947,
      "grad_norm": 0.45134401222170456,
      "learning_rate": 2.1165981434738025e-07,
      "loss": 0.0055,
      "step": 950
    },
    {
      "epoch": 6.4889643463497455,
      "grad_norm": 0.37311468378603657,
      "learning_rate": 1.6605457629533417e-07,
      "loss": 0.0049,
      "step": 960
    },
    {
      "epoch": 6.556876061120543,
      "grad_norm": 0.4991066248356068,
      "learning_rate": 1.2589447785450903e-07,
      "loss": 0.0056,
      "step": 970
    },
    {
      "epoch": 6.624787775891341,
      "grad_norm": 0.42890580429476455,
      "learning_rate": 9.122475688417954e-08,
      "loss": 0.0053,
      "step": 980
    },
    {
      "epoch": 6.6926994906621395,
      "grad_norm": 0.44354296838652474,
      "learning_rate": 6.208446667403167e-08,
      "loss": 0.0054,
      "step": 990
    },
    {
      "epoch": 6.760611205432937,
      "grad_norm": 0.32840772053670925,
      "learning_rate": 3.8506431953060365e-08,
      "loss": 0.0051,
      "step": 1000
    },
    {
      "epoch": 6.828522920203735,
      "grad_norm": 0.6268189928914281,
      "learning_rate": 2.0517211914545254e-08,
      "loss": 0.0052,
      "step": 1010
    },
    {
      "epoch": 6.896434634974533,
      "grad_norm": 0.2742284364295698,
      "learning_rate": 8.137070298769312e-09,
      "loss": 0.0052,
      "step": 1020
    },
    {
      "epoch": 6.964346349745331,
      "grad_norm": 0.4536117363233092,
      "learning_rate": 1.3799525671720492e-09,
      "loss": 0.0053,
      "step": 1030
    },
    {
      "epoch": 7.0,
      "step": 1036,
      "total_flos": 273471074074624.0,
      "train_loss": 0.24805861423295494,
      "train_runtime": 7529.4023,
      "train_samples_per_second": 17.517,
      "train_steps_per_second": 0.138
    }
  ],
  "logging_steps": 10,
  "max_steps": 1036,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 273471074074624.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}