{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.997184174103947,
  "global_step": 62800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 5e-05,
      "loss": 3.2345,
      "step": 500
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.959909555958242e-05,
      "loss": 1.994,
      "step": 1000
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.919819111916484e-05,
      "loss": 1.6753,
      "step": 1500
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.8797286678747256e-05,
      "loss": 1.5453,
      "step": 2000
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.8396382238329674e-05,
      "loss": 1.4013,
      "step": 2500
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.799547779791209e-05,
      "loss": 1.373,
      "step": 3000
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.759457335749451e-05,
      "loss": 1.2924,
      "step": 3500
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.719366891707693e-05,
      "loss": 1.2266,
      "step": 4000
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.6792764476659346e-05,
      "loss": 1.1925,
      "step": 4500
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.6391860036241764e-05,
      "loss": 1.22,
      "step": 5000
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.599095559582418e-05,
      "loss": 1.0955,
      "step": 5500
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.55900511554066e-05,
      "loss": 1.1189,
      "step": 6000
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.518914671498902e-05,
      "loss": 1.0632,
      "step": 6500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.4788242274571436e-05,
      "loss": 1.0091,
      "step": 7000
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.4387337834153854e-05,
      "loss": 1.0058,
      "step": 7500
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.398643339373627e-05,
      "loss": 1.0111,
      "step": 8000
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.358552895331869e-05,
      "loss": 0.8988,
      "step": 8500
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.318462451290111e-05,
      "loss": 0.9305,
      "step": 9000
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.2783720072483526e-05,
      "loss": 0.9372,
      "step": 9500
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.2382815632065944e-05,
      "loss": 0.9176,
      "step": 10000
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.198191119164836e-05,
      "loss": 0.8878,
      "step": 10500
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.158100675123078e-05,
      "loss": 0.872,
      "step": 11000
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.11801023108132e-05,
      "loss": 0.8525,
      "step": 11500
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.0779197870395616e-05,
      "loss": 0.8511,
      "step": 12000
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.0378293429978034e-05,
      "loss": 0.828,
      "step": 12500
    },
    {
      "epoch": 0.62,
      "learning_rate": 3.997738898956045e-05,
      "loss": 0.8198,
      "step": 13000
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.957648454914287e-05,
      "loss": 0.8225,
      "step": 13500
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.917558010872529e-05,
      "loss": 0.7818,
      "step": 14000
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.8774675668307706e-05,
      "loss": 0.7939,
      "step": 14500
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.8373771227890124e-05,
      "loss": 0.7799,
      "step": 15000
    },
    {
      "epoch": 0.74,
      "learning_rate": 3.7972866787472535e-05,
      "loss": 0.7588,
      "step": 15500
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.757196234705495e-05,
      "loss": 0.8004,
      "step": 16000
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.717105790663737e-05,
      "loss": 0.7468,
      "step": 16500
    },
    {
      "epoch": 0.81,
      "learning_rate": 3.677015346621979e-05,
      "loss": 0.7574,
      "step": 17000
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.636924902580221e-05,
      "loss": 0.7193,
      "step": 17500
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.5968344585384625e-05,
      "loss": 0.7122,
      "step": 18000
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.556744014496704e-05,
      "loss": 0.7215,
      "step": 18500
    },
    {
      "epoch": 0.91,
      "learning_rate": 3.516653570454946e-05,
      "loss": 0.6629,
      "step": 19000
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.476563126413188e-05,
      "loss": 0.7112,
      "step": 19500
    },
    {
      "epoch": 0.95,
      "learning_rate": 3.43647268237143e-05,
      "loss": 0.7116,
      "step": 20000
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.3963822383296715e-05,
      "loss": 0.6947,
      "step": 20500
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.356291794287913e-05,
      "loss": 0.6492,
      "step": 21000
    },
    {
      "epoch": 1.03,
      "learning_rate": 3.316201350246155e-05,
      "loss": 0.5083,
      "step": 21500
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.276110906204397e-05,
      "loss": 0.536,
      "step": 22000
    },
    {
      "epoch": 1.07,
      "learning_rate": 3.236020462162639e-05,
      "loss": 0.5196,
      "step": 22500
    },
    {
      "epoch": 1.1,
      "learning_rate": 3.1959300181208805e-05,
      "loss": 0.5637,
      "step": 23000
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.155839574079122e-05,
      "loss": 0.5338,
      "step": 23500
    },
    {
      "epoch": 1.15,
      "learning_rate": 3.115749130037364e-05,
      "loss": 0.5052,
      "step": 24000
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.075658685995606e-05,
      "loss": 0.5186,
      "step": 24500
    },
    {
      "epoch": 1.19,
      "learning_rate": 3.0355682419538477e-05,
      "loss": 0.5168,
      "step": 25000
    },
    {
      "epoch": 1.22,
      "learning_rate": 2.9954777979120895e-05,
      "loss": 0.4931,
      "step": 25500
    },
    {
      "epoch": 1.24,
      "learning_rate": 2.9553873538703313e-05,
      "loss": 0.4993,
      "step": 26000
    },
    {
      "epoch": 1.26,
      "learning_rate": 2.915296909828573e-05,
      "loss": 0.5286,
      "step": 26500
    },
    {
      "epoch": 1.29,
      "learning_rate": 2.875206465786815e-05,
      "loss": 0.4865,
      "step": 27000
    },
    {
      "epoch": 1.31,
      "learning_rate": 2.8351160217450567e-05,
      "loss": 0.503,
      "step": 27500
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.7950255777032985e-05,
      "loss": 0.5093,
      "step": 28000
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.7549351336615403e-05,
      "loss": 0.4978,
      "step": 28500
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.714844689619782e-05,
      "loss": 0.5042,
      "step": 29000
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.674754245578024e-05,
      "loss": 0.4909,
      "step": 29500
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.6346638015362657e-05,
      "loss": 0.4861,
      "step": 30000
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.5945733574945075e-05,
      "loss": 0.472,
      "step": 30500
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.5544829134527493e-05,
      "loss": 0.4673,
      "step": 31000
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.514392469410991e-05,
      "loss": 0.4747,
      "step": 31500
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.4743020253692332e-05,
      "loss": 0.4396,
      "step": 32000
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.434211581327475e-05,
      "loss": 0.4675,
      "step": 32500
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.3941211372857168e-05,
      "loss": 0.4421,
      "step": 33000
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.3540306932439586e-05,
      "loss": 0.4189,
      "step": 33500
    },
    {
      "epoch": 1.62,
      "learning_rate": 2.3139402492022004e-05,
      "loss": 0.4305,
      "step": 34000
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.273849805160442e-05,
      "loss": 0.4332,
      "step": 34500
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.2337593611186837e-05,
      "loss": 0.4314,
      "step": 35000
    },
    {
      "epoch": 1.69,
      "learning_rate": 2.1936689170769255e-05,
      "loss": 0.4138,
      "step": 35500
    },
    {
      "epoch": 1.72,
      "learning_rate": 2.1535784730351673e-05,
      "loss": 0.4179,
      "step": 36000
    },
    {
      "epoch": 1.74,
      "learning_rate": 2.113488028993409e-05,
      "loss": 0.4104,
      "step": 36500
    },
    {
      "epoch": 1.77,
      "learning_rate": 2.073397584951651e-05,
      "loss": 0.4078,
      "step": 37000
    },
    {
      "epoch": 1.79,
      "learning_rate": 2.0333071409098927e-05,
      "loss": 0.4113,
      "step": 37500
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.9932166968681345e-05,
      "loss": 0.3828,
      "step": 38000
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.9531262528263763e-05,
      "loss": 0.4095,
      "step": 38500
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.913035808784618e-05,
      "loss": 0.3782,
      "step": 39000
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.87294536474286e-05,
      "loss": 0.3494,
      "step": 39500
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.8328549207011017e-05,
      "loss": 0.3758,
      "step": 40000
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.7927644766593435e-05,
      "loss": 0.3628,
      "step": 40500
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.7526740326175853e-05,
      "loss": 0.3751,
      "step": 41000
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.712583588575827e-05,
      "loss": 0.3474,
      "step": 41500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.672493144534069e-05,
      "loss": 0.3201,
      "step": 42000
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.6324027004923107e-05,
      "loss": 0.237,
      "step": 42500
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.5923122564505525e-05,
      "loss": 0.2274,
      "step": 43000
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.5522218124087943e-05,
      "loss": 0.2453,
      "step": 43500
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.512131368367036e-05,
      "loss": 0.2502,
      "step": 44000
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.4720409243252779e-05,
      "loss": 0.2575,
      "step": 44500
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.4319504802835197e-05,
      "loss": 0.2533,
      "step": 45000
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3918600362417613e-05,
      "loss": 0.2496,
      "step": 45500
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.3517695922000031e-05,
      "loss": 0.2461,
      "step": 46000
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.3116791481582449e-05,
      "loss": 0.2509,
      "step": 46500
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.2715887041164867e-05,
      "loss": 0.2343,
      "step": 47000
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.2314982600747287e-05,
      "loss": 0.2549,
      "step": 47500
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.1914078160329705e-05,
      "loss": 0.2382,
      "step": 48000
    },
    {
      "epoch": 2.31,
      "learning_rate": 1.1513173719912123e-05,
      "loss": 0.2364,
      "step": 48500
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.111226927949454e-05,
      "loss": 0.2334,
      "step": 49000
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.0711364839076959e-05,
      "loss": 0.2291,
      "step": 49500
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.0310460398659377e-05,
      "loss": 0.2187,
      "step": 50000
    },
    {
      "epoch": 2.41,
      "learning_rate": 9.909555958241795e-06,
      "loss": 0.1919,
      "step": 50500
    },
    {
      "epoch": 2.43,
      "learning_rate": 9.50865151782421e-06,
      "loss": 0.2364,
      "step": 51000
    },
    {
      "epoch": 2.46,
      "learning_rate": 9.107747077406629e-06,
      "loss": 0.2045,
      "step": 51500
    },
    {
      "epoch": 2.48,
      "learning_rate": 8.706842636989047e-06,
      "loss": 0.2117,
      "step": 52000
    },
    {
      "epoch": 2.51,
      "learning_rate": 8.305938196571465e-06,
      "loss": 0.2266,
      "step": 52500
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.905033756153883e-06,
      "loss": 0.2023,
      "step": 53000
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.504129315736301e-06,
      "loss": 0.2032,
      "step": 53500
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.103224875318719e-06,
      "loss": 0.1957,
      "step": 54000
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.702320434901137e-06,
      "loss": 0.1869,
      "step": 54500
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.301415994483555e-06,
      "loss": 0.2016,
      "step": 55000
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.9005115540659736e-06,
      "loss": 0.1831,
      "step": 55500
    },
    {
      "epoch": 2.67,
      "learning_rate": 5.499607113648391e-06,
      "loss": 0.1707,
      "step": 56000
    },
    {
      "epoch": 2.7,
      "learning_rate": 5.098702673230809e-06,
      "loss": 0.1902,
      "step": 56500
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.697798232813227e-06,
      "loss": 0.1776,
      "step": 57000
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.296893792395645e-06,
      "loss": 0.2065,
      "step": 57500
    },
    {
      "epoch": 2.77,
      "learning_rate": 3.895989351978063e-06,
      "loss": 0.1955,
      "step": 58000
    },
    {
      "epoch": 2.79,
      "learning_rate": 3.4950849115604802e-06,
      "loss": 0.1747,
      "step": 58500
    },
    {
      "epoch": 2.82,
      "learning_rate": 3.0941804711428986e-06,
      "loss": 0.198,
      "step": 59000
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.6932760307253166e-06,
      "loss": 0.1885,
      "step": 59500
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.292371590307734e-06,
      "loss": 0.1537,
      "step": 60000
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.8914671498901522e-06,
      "loss": 0.1657,
      "step": 60500
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.4905627094725702e-06,
      "loss": 0.1631,
      "step": 61000
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.0896582690549882e-06,
      "loss": 0.1417,
      "step": 61500
    },
    {
      "epoch": 2.96,
      "learning_rate": 6.88753828637406e-07,
      "loss": 0.157,
      "step": 62000
    },
    {
      "epoch": 2.98,
      "learning_rate": 2.878493882198239e-07,
      "loss": 0.1536,
      "step": 62500
    }
  ],
  "max_steps": 62859,
  "num_train_epochs": 3,
  "total_flos": 4.536353409695568e+16,
  "trial_name": null,
  "trial_params": null
}