{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 55479,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 4.954937904432308e-05,
      "loss": 1.8047,
      "step": 500
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.9098758088646155e-05,
      "loss": 1.0161,
      "step": 1000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.864813713296923e-05,
      "loss": 0.9184,
      "step": 1500
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.8197516177292315e-05,
      "loss": 0.8381,
      "step": 2000
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.774689522161539e-05,
      "loss": 0.8042,
      "step": 2500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.729627426593847e-05,
      "loss": 0.7784,
      "step": 3000
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.6845653310261544e-05,
      "loss": 0.7531,
      "step": 3500
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.639503235458462e-05,
      "loss": 0.7263,
      "step": 4000
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.59444113989077e-05,
      "loss": 0.7171,
      "step": 4500
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.549379044323077e-05,
      "loss": 0.6689,
      "step": 5000
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.504316948755385e-05,
      "loss": 0.6679,
      "step": 5500
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.4592548531876926e-05,
      "loss": 0.6308,
      "step": 6000
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.41419275762e-05,
      "loss": 0.6113,
      "step": 6500
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.369130662052308e-05,
      "loss": 0.6029,
      "step": 7000
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.324068566484616e-05,
      "loss": 0.6022,
      "step": 7500
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.279006470916924e-05,
      "loss": 0.5795,
      "step": 8000
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.2339443753492315e-05,
      "loss": 0.5509,
      "step": 8500
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.188882279781539e-05,
      "loss": 0.5502,
      "step": 9000
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.143820184213847e-05,
      "loss": 0.5475,
      "step": 9500
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.0987580886461544e-05,
      "loss": 0.5176,
      "step": 10000
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.053695993078463e-05,
      "loss": 0.524,
      "step": 10500
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.0086338975107703e-05,
      "loss": 0.4914,
      "step": 11000
    },
    {
      "epoch": 0.62,
      "learning_rate": 3.963571801943078e-05,
      "loss": 0.4989,
      "step": 11500
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.9185097063753856e-05,
      "loss": 0.4852,
      "step": 12000
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.873447610807693e-05,
      "loss": 0.4751,
      "step": 12500
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.828385515240001e-05,
      "loss": 0.4766,
      "step": 13000
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.7833234196723085e-05,
      "loss": 0.4558,
      "step": 13500
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.738261324104617e-05,
      "loss": 0.46,
      "step": 14000
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.693199228536924e-05,
      "loss": 0.4379,
      "step": 14500
    },
    {
      "epoch": 0.81,
      "learning_rate": 3.6481371329692315e-05,
      "loss": 0.4219,
      "step": 15000
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.603075037401539e-05,
      "loss": 0.4261,
      "step": 15500
    },
    {
      "epoch": 0.87,
      "learning_rate": 3.558012941833847e-05,
      "loss": 0.4075,
      "step": 16000
    },
    {
      "epoch": 0.89,
      "learning_rate": 3.512950846266155e-05,
      "loss": 0.4027,
      "step": 16500
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.467888750698463e-05,
      "loss": 0.3952,
      "step": 17000
    },
    {
      "epoch": 0.95,
      "learning_rate": 3.4228266551307703e-05,
      "loss": 0.3859,
      "step": 17500
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.377764559563078e-05,
      "loss": 0.3845,
      "step": 18000
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.3327024639953856e-05,
      "loss": 0.3848,
      "step": 18500
    },
    {
      "epoch": 1.03,
      "learning_rate": 3.287640368427693e-05,
      "loss": 0.3174,
      "step": 19000
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.2425782728600016e-05,
      "loss": 0.3251,
      "step": 19500
    },
    {
      "epoch": 1.08,
      "learning_rate": 3.197516177292309e-05,
      "loss": 0.3244,
      "step": 20000
    },
    {
      "epoch": 1.11,
      "learning_rate": 3.152454081724617e-05,
      "loss": 0.3186,
      "step": 20500
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.1073919861569245e-05,
      "loss": 0.3151,
      "step": 21000
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.062329890589232e-05,
      "loss": 0.2993,
      "step": 21500
    },
    {
      "epoch": 1.19,
      "learning_rate": 3.01726779502154e-05,
      "loss": 0.2997,
      "step": 22000
    },
    {
      "epoch": 1.22,
      "learning_rate": 2.9722056994538478e-05,
      "loss": 0.3078,
      "step": 22500
    },
    {
      "epoch": 1.24,
      "learning_rate": 2.9271436038861554e-05,
      "loss": 0.2988,
      "step": 23000
    },
    {
      "epoch": 1.27,
      "learning_rate": 2.8820815083184634e-05,
      "loss": 0.2912,
      "step": 23500
    },
    {
      "epoch": 1.3,
      "learning_rate": 2.8370194127507703e-05,
      "loss": 0.288,
      "step": 24000
    },
    {
      "epoch": 1.32,
      "learning_rate": 2.7919573171830783e-05,
      "loss": 0.277,
      "step": 24500
    },
    {
      "epoch": 1.35,
      "learning_rate": 2.746895221615386e-05,
      "loss": 0.2886,
      "step": 25000
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.7018331260476936e-05,
      "loss": 0.2659,
      "step": 25500
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.6567710304800016e-05,
      "loss": 0.2713,
      "step": 26000
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.6117089349123092e-05,
      "loss": 0.2755,
      "step": 26500
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.566646839344617e-05,
      "loss": 0.2723,
      "step": 27000
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.5215847437769248e-05,
      "loss": 0.261,
      "step": 27500
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.4765226482092325e-05,
      "loss": 0.2628,
      "step": 28000
    },
    {
      "epoch": 1.54,
      "learning_rate": 2.43146055264154e-05,
      "loss": 0.2538,
      "step": 28500
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.3863984570738477e-05,
      "loss": 0.2526,
      "step": 29000
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.3413363615061557e-05,
      "loss": 0.2595,
      "step": 29500
    },
    {
      "epoch": 1.62,
      "learning_rate": 2.2962742659384634e-05,
      "loss": 0.2523,
      "step": 30000
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.251212170370771e-05,
      "loss": 0.2431,
      "step": 30500
    },
    {
      "epoch": 1.68,
      "learning_rate": 2.2061500748030786e-05,
      "loss": 0.2351,
      "step": 31000
    },
    {
      "epoch": 1.7,
      "learning_rate": 2.1610879792353863e-05,
      "loss": 0.2409,
      "step": 31500
    },
    {
      "epoch": 1.73,
      "learning_rate": 2.1160258836676943e-05,
      "loss": 0.23,
      "step": 32000
    },
    {
      "epoch": 1.76,
      "learning_rate": 2.070963788100002e-05,
      "loss": 0.223,
      "step": 32500
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.0259016925323095e-05,
      "loss": 0.2247,
      "step": 33000
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.9808395969646172e-05,
      "loss": 0.221,
      "step": 33500
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.935777501396925e-05,
      "loss": 0.222,
      "step": 34000
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.8907154058292328e-05,
      "loss": 0.2262,
      "step": 34500
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.8456533102615404e-05,
      "loss": 0.2137,
      "step": 35000
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.8005912146938484e-05,
      "loss": 0.2143,
      "step": 35500
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.7555291191261557e-05,
      "loss": 0.21,
      "step": 36000
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.7104670235584637e-05,
      "loss": 0.2025,
      "step": 36500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.6654049279907713e-05,
      "loss": 0.2038,
      "step": 37000
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.620342832423079e-05,
      "loss": 0.1669,
      "step": 37500
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.575280736855387e-05,
      "loss": 0.1703,
      "step": 38000
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.5302186412876946e-05,
      "loss": 0.1641,
      "step": 38500
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.4851565457200022e-05,
      "loss": 0.1578,
      "step": 39000
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.44009445015231e-05,
      "loss": 0.1523,
      "step": 39500
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.3950323545846179e-05,
      "loss": 0.1566,
      "step": 40000
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.3499702590169253e-05,
      "loss": 0.1602,
      "step": 40500
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.304908163449233e-05,
      "loss": 0.1523,
      "step": 41000
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.2598460678815408e-05,
      "loss": 0.1551,
      "step": 41500
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.2147839723138484e-05,
      "loss": 0.1533,
      "step": 42000
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.1697218767461562e-05,
      "loss": 0.1517,
      "step": 42500
    },
    {
      "epoch": 2.33,
      "learning_rate": 1.124659781178464e-05,
      "loss": 0.1394,
      "step": 43000
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.0795976856107717e-05,
      "loss": 0.1471,
      "step": 43500
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.0345355900430793e-05,
      "loss": 0.1484,
      "step": 44000
    },
    {
      "epoch": 2.41,
      "learning_rate": 9.894734944753871e-06,
      "loss": 0.1454,
      "step": 44500
    },
    {
      "epoch": 2.43,
      "learning_rate": 9.444113989076948e-06,
      "loss": 0.1368,
      "step": 45000
    },
    {
      "epoch": 2.46,
      "learning_rate": 8.993493033400026e-06,
      "loss": 0.1404,
      "step": 45500
    },
    {
      "epoch": 2.49,
      "learning_rate": 8.542872077723104e-06,
      "loss": 0.1362,
      "step": 46000
    },
    {
      "epoch": 2.51,
      "learning_rate": 8.09225112204618e-06,
      "loss": 0.135,
      "step": 46500
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.641630166369257e-06,
      "loss": 0.1329,
      "step": 47000
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.191009210692335e-06,
      "loss": 0.1313,
      "step": 47500
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.740388255015412e-06,
      "loss": 0.1327,
      "step": 48000
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.289767299338489e-06,
      "loss": 0.1242,
      "step": 48500
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.839146343661566e-06,
      "loss": 0.1232,
      "step": 49000
    },
    {
      "epoch": 2.68,
      "learning_rate": 5.388525387984643e-06,
      "loss": 0.1294,
      "step": 49500
    },
    {
      "epoch": 2.7,
      "learning_rate": 4.93790443230772e-06,
      "loss": 0.1275,
      "step": 50000
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.487283476630798e-06,
      "loss": 0.1261,
      "step": 50500
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.0366625209538745e-06,
      "loss": 0.1282,
      "step": 51000
    },
    {
      "epoch": 2.78,
      "learning_rate": 3.586041565276952e-06,
      "loss": 0.1245,
      "step": 51500
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.135420609600029e-06,
      "loss": 0.1247,
      "step": 52000
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.6847996539231063e-06,
      "loss": 0.121,
      "step": 52500
    },
    {
      "epoch": 2.87,
      "learning_rate": 2.2341786982461835e-06,
      "loss": 0.1104,
      "step": 53000
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.7835577425692605e-06,
      "loss": 0.1234,
      "step": 53500
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.3329367868923376e-06,
      "loss": 0.1146,
      "step": 54000
    },
    {
      "epoch": 2.95,
      "learning_rate": 8.823158312154149e-07,
      "loss": 0.1197,
      "step": 54500
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.3169487553849206e-07,
      "loss": 0.1116,
      "step": 55000
    },
    {
      "epoch": 3.0,
      "step": 55479,
      "total_flos": 2.3702775297552e+17,
      "train_loss": 0.3350037185057538,
      "train_runtime": 233521.0358,
      "train_samples_per_second": 7.602,
      "train_steps_per_second": 0.238
    }
  ],
  "logging_steps": 500,
  "max_steps": 55479,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 2.3702775297552e+17,
  "trial_name": null,
  "trial_params": null
}