{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.629948364888124,
  "eval_steps": 5000,
  "global_step": 8500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1721170395869191,
      "grad_norm": 134.16664123535156,
      "learning_rate": 2.224770642201835e-06,
      "loss": 10.8697,
      "step": 100
    },
    {
      "epoch": 0.3442340791738382,
      "grad_norm": 69.6041259765625,
      "learning_rate": 4.5183486238532115e-06,
      "loss": 9.1125,
      "step": 200
    },
    {
      "epoch": 0.5163511187607573,
      "grad_norm": 103.4394760131836,
      "learning_rate": 6.8119266055045875e-06,
      "loss": 6.8873,
      "step": 300
    },
    {
      "epoch": 0.6884681583476764,
      "grad_norm": 32.237342834472656,
      "learning_rate": 9.08256880733945e-06,
      "loss": 3.1124,
      "step": 400
    },
    {
      "epoch": 0.8605851979345955,
      "grad_norm": 13.523033142089844,
      "learning_rate": 1.1376146788990828e-05,
      "loss": 1.0882,
      "step": 500
    },
    {
      "epoch": 1.0327022375215147,
      "grad_norm": 56.13002395629883,
      "learning_rate": 1.3669724770642203e-05,
      "loss": 0.869,
      "step": 600
    },
    {
      "epoch": 1.2048192771084336,
      "grad_norm": 4.41276741027832,
      "learning_rate": 1.5963302752293578e-05,
      "loss": 0.6952,
      "step": 700
    },
    {
      "epoch": 1.3769363166953528,
      "grad_norm": 1.8771318197250366,
      "learning_rate": 1.8256880733944955e-05,
      "loss": 0.5522,
      "step": 800
    },
    {
      "epoch": 1.549053356282272,
      "grad_norm": 5.131401062011719,
      "learning_rate": 1.9938798928981258e-05,
      "loss": 0.5184,
      "step": 900
    },
    {
      "epoch": 1.721170395869191,
      "grad_norm": 3.2520999908447266,
      "learning_rate": 1.9683794466403164e-05,
      "loss": 0.3996,
      "step": 1000
    },
    {
      "epoch": 1.8932874354561102,
      "grad_norm": 1.1916402578353882,
      "learning_rate": 1.9428790003825067e-05,
      "loss": 0.6316,
      "step": 1100
    },
    {
      "epoch": 2.0654044750430294,
      "grad_norm": 3.3565962314605713,
      "learning_rate": 1.9173785541246974e-05,
      "loss": 0.5352,
      "step": 1200
    },
    {
      "epoch": 2.2375215146299485,
      "grad_norm": 2.182133436203003,
      "learning_rate": 1.8918781078668877e-05,
      "loss": 0.3731,
      "step": 1300
    },
    {
      "epoch": 2.4096385542168672,
      "grad_norm": 3.537748098373413,
      "learning_rate": 1.8663776616090783e-05,
      "loss": 0.3376,
      "step": 1400
    },
    {
      "epoch": 2.581755593803787,
      "grad_norm": 2.066549777984619,
      "learning_rate": 1.840877215351269e-05,
      "loss": 0.597,
      "step": 1500
    },
    {
      "epoch": 2.7538726333907055,
      "grad_norm": 2.867453098297119,
      "learning_rate": 1.8153767690934592e-05,
      "loss": 0.5737,
      "step": 1600
    },
    {
      "epoch": 2.9259896729776247,
      "grad_norm": 0.8096536993980408,
      "learning_rate": 1.7898763228356495e-05,
      "loss": 0.7107,
      "step": 1700
    },
    {
      "epoch": 3.098106712564544,
      "grad_norm": 5.293230056762695,
      "learning_rate": 1.7643758765778402e-05,
      "loss": 0.4356,
      "step": 1800
    },
    {
      "epoch": 3.270223752151463,
      "grad_norm": 2.1939845085144043,
      "learning_rate": 1.7388754303200308e-05,
      "loss": 0.5581,
      "step": 1900
    },
    {
      "epoch": 3.442340791738382,
      "grad_norm": 2.1973116397857666,
      "learning_rate": 1.713374984062221e-05,
      "loss": 0.2012,
      "step": 2000
    },
    {
      "epoch": 3.6144578313253013,
      "grad_norm": 1.3364547491073608,
      "learning_rate": 1.6878745378044118e-05,
      "loss": 0.3906,
      "step": 2100
    },
    {
      "epoch": 3.7865748709122204,
      "grad_norm": 3.2359094619750977,
      "learning_rate": 1.662374091546602e-05,
      "loss": 0.5386,
      "step": 2200
    },
    {
      "epoch": 3.958691910499139,
      "grad_norm": 31.699663162231445,
      "learning_rate": 1.6368736452887927e-05,
      "loss": 0.2624,
      "step": 2300
    },
    {
      "epoch": 4.130808950086059,
      "grad_norm": 92.98713684082031,
      "learning_rate": 1.611373199030983e-05,
      "loss": 0.3573,
      "step": 2400
    },
    {
      "epoch": 4.3029259896729775,
      "grad_norm": 2.056157350540161,
      "learning_rate": 1.5858727527731736e-05,
      "loss": 0.4798,
      "step": 2500
    },
    {
      "epoch": 4.475043029259897,
      "grad_norm": 7.822810649871826,
      "learning_rate": 1.5606273109779423e-05,
      "loss": 0.2465,
      "step": 2600
    },
    {
      "epoch": 4.647160068846816,
      "grad_norm": 1.6002038717269897,
      "learning_rate": 1.5351268647201326e-05,
      "loss": 0.3482,
      "step": 2700
    },
    {
      "epoch": 4.8192771084337345,
      "grad_norm": 2.061086416244507,
      "learning_rate": 1.5096264184623233e-05,
      "loss": 0.1915,
      "step": 2800
    },
    {
      "epoch": 4.991394148020654,
      "grad_norm": 1.1744683980941772,
      "learning_rate": 1.4841259722045136e-05,
      "loss": 0.4617,
      "step": 2900
    },
    {
      "epoch": 5.163511187607573,
      "grad_norm": 2.5757875442504883,
      "learning_rate": 1.4586255259467042e-05,
      "loss": 0.2874,
      "step": 3000
    },
    {
      "epoch": 5.335628227194492,
      "grad_norm": 8.106232643127441,
      "learning_rate": 1.4331250796888947e-05,
      "loss": 0.4636,
      "step": 3100
    },
    {
      "epoch": 5.507745266781411,
      "grad_norm": 2.139594316482544,
      "learning_rate": 1.4076246334310853e-05,
      "loss": 0.1344,
      "step": 3200
    },
    {
      "epoch": 5.679862306368331,
      "grad_norm": 8.198427200317383,
      "learning_rate": 1.3821241871732756e-05,
      "loss": 0.3615,
      "step": 3300
    },
    {
      "epoch": 5.851979345955249,
      "grad_norm": 0.706113338470459,
      "learning_rate": 1.3566237409154661e-05,
      "loss": 0.309,
      "step": 3400
    },
    {
      "epoch": 6.024096385542169,
      "grad_norm": 1.0154913663864136,
      "learning_rate": 1.3311232946576567e-05,
      "loss": 0.1883,
      "step": 3500
    },
    {
      "epoch": 6.196213425129088,
      "grad_norm": 27.715837478637695,
      "learning_rate": 1.3056228483998472e-05,
      "loss": 0.4029,
      "step": 3600
    },
    {
      "epoch": 6.368330464716007,
      "grad_norm": 3.0514609813690186,
      "learning_rate": 1.2801224021420375e-05,
      "loss": 0.2082,
      "step": 3700
    },
    {
      "epoch": 6.540447504302926,
      "grad_norm": 3.8193249702453613,
      "learning_rate": 1.2546219558842281e-05,
      "loss": 0.1333,
      "step": 3800
    },
    {
      "epoch": 6.712564543889846,
      "grad_norm": 1.4768047332763672,
      "learning_rate": 1.2291215096264186e-05,
      "loss": 0.1509,
      "step": 3900
    },
    {
      "epoch": 6.884681583476764,
      "grad_norm": 1.5106594562530518,
      "learning_rate": 1.2036210633686089e-05,
      "loss": 0.6264,
      "step": 4000
    },
    {
      "epoch": 7.056798623063683,
      "grad_norm": 1.1024622917175293,
      "learning_rate": 1.1781206171107995e-05,
      "loss": 0.2177,
      "step": 4100
    },
    {
      "epoch": 7.228915662650603,
      "grad_norm": 0.900026798248291,
      "learning_rate": 1.15262017085299e-05,
      "loss": 0.1957,
      "step": 4200
    },
    {
      "epoch": 7.401032702237521,
      "grad_norm": 144.5244140625,
      "learning_rate": 1.1271197245951807e-05,
      "loss": 0.2887,
      "step": 4300
    },
    {
      "epoch": 7.573149741824441,
      "grad_norm": 4.466265678405762,
      "learning_rate": 1.101619278337371e-05,
      "loss": 0.2271,
      "step": 4400
    },
    {
      "epoch": 7.74526678141136,
      "grad_norm": 2.862029790878296,
      "learning_rate": 1.0761188320795614e-05,
      "loss": 0.3486,
      "step": 4500
    },
    {
      "epoch": 7.917383820998279,
      "grad_norm": 1.178603172302246,
      "learning_rate": 1.050618385821752e-05,
      "loss": 0.4429,
      "step": 4600
    },
    {
      "epoch": 8.089500860585199,
      "grad_norm": 6.430075645446777,
      "learning_rate": 1.0251179395639424e-05,
      "loss": 0.4398,
      "step": 4700
    },
    {
      "epoch": 8.261617900172118,
      "grad_norm": 6.42482852935791,
      "learning_rate": 9.996174933061328e-06,
      "loss": 0.31,
      "step": 4800
    },
    {
      "epoch": 8.433734939759036,
      "grad_norm": 6.2779622077941895,
      "learning_rate": 9.743720515109015e-06,
      "loss": 0.2045,
      "step": 4900
    },
    {
      "epoch": 8.605851979345955,
      "grad_norm": 4.175030708312988,
      "learning_rate": 9.48871605253092e-06,
      "loss": 0.2583,
      "step": 5000
    },
    {
      "epoch": 8.605851979345955,
      "eval_loss": 0.23712533712387085,
      "eval_runtime": 13.9021,
      "eval_samples_per_second": 1335.195,
      "eval_steps_per_second": 10.502,
      "step": 5000
    },
    {
      "epoch": 8.777969018932874,
      "grad_norm": 1.0522035360336304,
      "learning_rate": 9.233711589952825e-06,
      "loss": 0.2774,
      "step": 5100
    },
    {
      "epoch": 8.950086058519794,
      "grad_norm": 0.9467515349388123,
      "learning_rate": 8.98125717200051e-06,
      "loss": 0.1902,
      "step": 5200
    },
    {
      "epoch": 9.122203098106713,
      "grad_norm": 9.148195266723633,
      "learning_rate": 8.726252709422416e-06,
      "loss": 0.3058,
      "step": 5300
    },
    {
      "epoch": 9.294320137693632,
      "grad_norm": 9.301542282104492,
      "learning_rate": 8.471248246844321e-06,
      "loss": 0.3742,
      "step": 5400
    },
    {
      "epoch": 9.46643717728055,
      "grad_norm": 17.278079986572266,
      "learning_rate": 8.216243784266226e-06,
      "loss": 0.2972,
      "step": 5500
    },
    {
      "epoch": 9.638554216867469,
      "grad_norm": 4.043286323547363,
      "learning_rate": 7.96123932168813e-06,
      "loss": 0.3084,
      "step": 5600
    },
    {
      "epoch": 9.81067125645439,
      "grad_norm": 2.227259874343872,
      "learning_rate": 7.706234859110035e-06,
      "loss": 0.1215,
      "step": 5700
    },
    {
      "epoch": 9.982788296041308,
      "grad_norm": 0.9134290218353271,
      "learning_rate": 7.45123039653194e-06,
      "loss": 0.1876,
      "step": 5800
    },
    {
      "epoch": 10.154905335628227,
      "grad_norm": 1.4163002967834473,
      "learning_rate": 7.196225933953844e-06,
      "loss": 0.1702,
      "step": 5900
    },
    {
      "epoch": 10.327022375215146,
      "grad_norm": 1.564228892326355,
      "learning_rate": 6.94122147137575e-06,
      "loss": 0.2506,
      "step": 6000
    },
    {
      "epoch": 10.499139414802066,
      "grad_norm": 5.47558069229126,
      "learning_rate": 6.686217008797654e-06,
      "loss": 0.2852,
      "step": 6100
    },
    {
      "epoch": 10.671256454388985,
      "grad_norm": 10.801889419555664,
      "learning_rate": 6.431212546219559e-06,
      "loss": 0.2354,
      "step": 6200
    },
    {
      "epoch": 10.843373493975903,
      "grad_norm": 1.8754569292068481,
      "learning_rate": 6.176208083641464e-06,
      "loss": 0.214,
      "step": 6300
    },
    {
      "epoch": 11.015490533562822,
      "grad_norm": 2.237508773803711,
      "learning_rate": 5.9212036210633696e-06,
      "loss": 0.3815,
      "step": 6400
    },
    {
      "epoch": 11.187607573149743,
      "grad_norm": 13.412964820861816,
      "learning_rate": 5.666199158485273e-06,
      "loss": 0.0803,
      "step": 6500
    },
    {
      "epoch": 11.359724612736661,
      "grad_norm": 0.996343195438385,
      "learning_rate": 5.411194695907179e-06,
      "loss": 0.1941,
      "step": 6600
    },
    {
      "epoch": 11.53184165232358,
      "grad_norm": 64.30641174316406,
      "learning_rate": 5.156190233329084e-06,
      "loss": 0.1576,
      "step": 6700
    },
    {
      "epoch": 11.703958691910499,
      "grad_norm": 0.9045078158378601,
      "learning_rate": 4.901185770750988e-06,
      "loss": 0.2911,
      "step": 6800
    },
    {
      "epoch": 11.876075731497417,
      "grad_norm": 1.798627495765686,
      "learning_rate": 4.646181308172894e-06,
      "loss": 0.4913,
      "step": 6900
    },
    {
      "epoch": 12.048192771084338,
      "grad_norm": 6.165831565856934,
      "learning_rate": 4.3911768455947986e-06,
      "loss": 0.2759,
      "step": 7000
    },
    {
      "epoch": 12.220309810671257,
      "grad_norm": 3.460507392883301,
      "learning_rate": 4.136172383016703e-06,
      "loss": 0.2928,
      "step": 7100
    },
    {
      "epoch": 12.392426850258175,
      "grad_norm": 5.5960187911987305,
      "learning_rate": 3.881167920438608e-06,
      "loss": 0.2181,
      "step": 7200
    },
    {
      "epoch": 12.564543889845094,
      "grad_norm": 1.473883032798767,
      "learning_rate": 3.6261634578605126e-06,
      "loss": 0.1286,
      "step": 7300
    },
    {
      "epoch": 12.736660929432015,
      "grad_norm": 10.179828643798828,
      "learning_rate": 3.3711589952824173e-06,
      "loss": 0.3342,
      "step": 7400
    },
    {
      "epoch": 12.908777969018933,
      "grad_norm": 3.2593960762023926,
      "learning_rate": 3.1187045773301034e-06,
      "loss": 0.1577,
      "step": 7500
    },
    {
      "epoch": 13.080895008605852,
      "grad_norm": 1.169028878211975,
      "learning_rate": 2.863700114752008e-06,
      "loss": 0.2578,
      "step": 7600
    },
    {
      "epoch": 13.25301204819277,
      "grad_norm": 1.2204866409301758,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.2844,
      "step": 7700
    },
    {
      "epoch": 13.42512908777969,
      "grad_norm": 1.1637088060379028,
      "learning_rate": 2.353691189595818e-06,
      "loss": 0.0917,
      "step": 7800
    },
    {
      "epoch": 13.59724612736661,
      "grad_norm": 0.6639829277992249,
      "learning_rate": 2.098686727017723e-06,
      "loss": 0.2617,
      "step": 7900
    },
    {
      "epoch": 13.769363166953529,
      "grad_norm": 4.155405044555664,
      "learning_rate": 1.843682264439628e-06,
      "loss": 0.3021,
      "step": 8000
    },
    {
      "epoch": 13.941480206540447,
      "grad_norm": 1.9663244485855103,
      "learning_rate": 1.5886778018615326e-06,
      "loss": 0.1036,
      "step": 8100
    },
    {
      "epoch": 14.113597246127366,
      "grad_norm": 32.85494613647461,
      "learning_rate": 1.3336733392834375e-06,
      "loss": 0.5471,
      "step": 8200
    },
    {
      "epoch": 14.285714285714286,
      "grad_norm": 0.6067169904708862,
      "learning_rate": 1.0786688767053424e-06,
      "loss": 0.2395,
      "step": 8300
    },
    {
      "epoch": 14.457831325301205,
      "grad_norm": 0.9747382998466492,
      "learning_rate": 8.236644141272474e-07,
      "loss": 0.2664,
      "step": 8400
    },
    {
      "epoch": 14.629948364888124,
      "grad_norm": 21.624757766723633,
      "learning_rate": 5.686599515491522e-07,
      "loss": 0.2697,
      "step": 8500
    }
  ],
  "logging_steps": 100,
  "max_steps": 8715,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}