{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.919645785503444,
  "global_step": 60000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "learning_rate": 5e-06,
      "loss": 8.6674,
      "step": 500
    },
    {
      "epoch": 0.08,
      "learning_rate": 1e-05,
      "loss": 6.0699,
      "step": 1000
    },
    {
      "epoch": 0.12,
      "learning_rate": 1.5e-05,
      "loss": 5.3064,
      "step": 1500
    },
    {
      "epoch": 0.16,
      "learning_rate": 2e-05,
      "loss": 4.6019,
      "step": 2000
    },
    {
      "epoch": 0.2,
      "learning_rate": 2.5e-05,
      "loss": 3.9841,
      "step": 2500
    },
    {
      "epoch": 0.25,
      "learning_rate": 3e-05,
      "loss": 3.5033,
      "step": 3000
    },
    {
      "epoch": 0.29,
      "learning_rate": 3.5e-05,
      "loss": 3.1509,
      "step": 3500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4e-05,
      "loss": 2.8726,
      "step": 4000
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.5e-05,
      "loss": 2.6501,
      "step": 4500
    },
    {
      "epoch": 0.41,
      "learning_rate": 5e-05,
      "loss": 2.4705,
      "step": 5000
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.955341193283315e-05,
      "loss": 2.329,
      "step": 5500
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.9106823865666315e-05,
      "loss": 2.2016,
      "step": 6000
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.8660235798499465e-05,
      "loss": 2.108,
      "step": 6500
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.821364773133262e-05,
      "loss": 2.0367,
      "step": 7000
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.776705966416578e-05,
      "loss": 1.9619,
      "step": 7500
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.7320471596998934e-05,
      "loss": 1.905,
      "step": 8000
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.6873883529832084e-05,
      "loss": 1.8638,
      "step": 8500
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.642729546266524e-05,
      "loss": 1.8176,
      "step": 9000
    },
    {
      "epoch": 0.78,
      "learning_rate": 4.59807073954984e-05,
      "loss": 1.7665,
      "step": 9500
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.5534119328331546e-05,
      "loss": 1.7427,
      "step": 10000
    },
    {
      "epoch": 0.86,
      "learning_rate": 4.50875312611647e-05,
      "loss": 1.7148,
      "step": 10500
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.464094319399786e-05,
      "loss": 1.686,
      "step": 11000
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.419435512683101e-05,
      "loss": 1.6532,
      "step": 11500
    },
    {
      "epoch": 0.98,
      "learning_rate": 4.374776705966417e-05,
      "loss": 1.6338,
      "step": 12000
    },
    {
      "epoch": 1.02,
      "learning_rate": 4.330117899249732e-05,
      "loss": 1.6049,
      "step": 12500
    },
    {
      "epoch": 1.07,
      "learning_rate": 4.285459092533048e-05,
      "loss": 1.5844,
      "step": 13000
    },
    {
      "epoch": 1.11,
      "learning_rate": 4.2408002858163634e-05,
      "loss": 1.5617,
      "step": 13500
    },
    {
      "epoch": 1.15,
      "learning_rate": 4.196141479099679e-05,
      "loss": 1.5475,
      "step": 14000
    },
    {
      "epoch": 1.19,
      "learning_rate": 4.151482672382994e-05,
      "loss": 1.533,
      "step": 14500
    },
    {
      "epoch": 1.23,
      "learning_rate": 4.10682386566631e-05,
      "loss": 1.5207,
      "step": 15000
    },
    {
      "epoch": 1.27,
      "learning_rate": 4.062165058949625e-05,
      "loss": 1.5027,
      "step": 15500
    },
    {
      "epoch": 1.31,
      "learning_rate": 4.01750625223294e-05,
      "loss": 1.4911,
      "step": 16000
    },
    {
      "epoch": 1.35,
      "learning_rate": 3.972847445516256e-05,
      "loss": 1.4758,
      "step": 16500
    },
    {
      "epoch": 1.39,
      "learning_rate": 3.9281886387995716e-05,
      "loss": 1.4632,
      "step": 17000
    },
    {
      "epoch": 1.43,
      "learning_rate": 3.8835298320828865e-05,
      "loss": 1.4508,
      "step": 17500
    },
    {
      "epoch": 1.48,
      "learning_rate": 3.838871025366202e-05,
      "loss": 1.4448,
      "step": 18000
    },
    {
      "epoch": 1.52,
      "learning_rate": 3.794212218649518e-05,
      "loss": 1.4325,
      "step": 18500
    },
    {
      "epoch": 1.56,
      "learning_rate": 3.7495534119328335e-05,
      "loss": 1.4219,
      "step": 19000
    },
    {
      "epoch": 1.6,
      "learning_rate": 3.7048946052161484e-05,
      "loss": 1.4095,
      "step": 19500
    },
    {
      "epoch": 1.64,
      "learning_rate": 3.660235798499465e-05,
      "loss": 1.4042,
      "step": 20000
    },
    {
      "epoch": 1.68,
      "learning_rate": 3.61557699178278e-05,
      "loss": 1.3934,
      "step": 20500
    },
    {
      "epoch": 1.72,
      "learning_rate": 3.570918185066095e-05,
      "loss": 1.3905,
      "step": 21000
    },
    {
      "epoch": 1.76,
      "learning_rate": 3.526259378349411e-05,
      "loss": 1.3783,
      "step": 21500
    },
    {
      "epoch": 1.8,
      "learning_rate": 3.481600571632726e-05,
      "loss": 1.3627,
      "step": 22000
    },
    {
      "epoch": 1.84,
      "learning_rate": 3.4369417649160416e-05,
      "loss": 1.3597,
      "step": 22500
    },
    {
      "epoch": 1.89,
      "learning_rate": 3.392282958199357e-05,
      "loss": 1.3594,
      "step": 23000
    },
    {
      "epoch": 1.93,
      "learning_rate": 3.347624151482672e-05,
      "loss": 1.3407,
      "step": 23500
    },
    {
      "epoch": 1.97,
      "learning_rate": 3.302965344765988e-05,
      "loss": 1.3396,
      "step": 24000
    },
    {
      "epoch": 2.01,
      "learning_rate": 3.2583065380493035e-05,
      "loss": 1.3309,
      "step": 24500
    },
    {
      "epoch": 2.05,
      "learning_rate": 3.213647731332619e-05,
      "loss": 1.3214,
      "step": 25000
    },
    {
      "epoch": 2.09,
      "learning_rate": 3.168988924615934e-05,
      "loss": 1.3113,
      "step": 25500
    },
    {
      "epoch": 2.13,
      "learning_rate": 3.1243301178992504e-05,
      "loss": 1.3082,
      "step": 26000
    },
    {
      "epoch": 2.17,
      "learning_rate": 3.0796713111825654e-05,
      "loss": 1.3057,
      "step": 26500
    },
    {
      "epoch": 2.21,
      "learning_rate": 3.0350125044658807e-05,
      "loss": 1.2962,
      "step": 27000
    },
    {
      "epoch": 2.25,
      "learning_rate": 2.9903536977491963e-05,
      "loss": 1.2969,
      "step": 27500
    },
    {
      "epoch": 2.3,
      "learning_rate": 2.9456948910325116e-05,
      "loss": 1.2909,
      "step": 28000
    },
    {
      "epoch": 2.34,
      "learning_rate": 2.9010360843158273e-05,
      "loss": 1.2803,
      "step": 28500
    },
    {
      "epoch": 2.38,
      "learning_rate": 2.8563772775991426e-05,
      "loss": 1.2762,
      "step": 29000
    },
    {
      "epoch": 2.42,
      "learning_rate": 2.811718470882458e-05,
      "loss": 1.2691,
      "step": 29500
    },
    {
      "epoch": 2.46,
      "learning_rate": 2.7670596641657735e-05,
      "loss": 1.2698,
      "step": 30000
    },
    {
      "epoch": 2.5,
      "learning_rate": 2.722400857449089e-05,
      "loss": 1.2646,
      "step": 30500
    },
    {
      "epoch": 2.54,
      "learning_rate": 2.6777420507324048e-05,
      "loss": 1.2598,
      "step": 31000
    },
    {
      "epoch": 2.58,
      "learning_rate": 2.6330832440157198e-05,
      "loss": 1.2598,
      "step": 31500
    },
    {
      "epoch": 2.62,
      "learning_rate": 2.5884244372990358e-05,
      "loss": 1.2499,
      "step": 32000
    },
    {
      "epoch": 2.66,
      "learning_rate": 2.543765630582351e-05,
      "loss": 1.2492,
      "step": 32500
    },
    {
      "epoch": 2.71,
      "learning_rate": 2.4991068238656664e-05,
      "loss": 1.2417,
      "step": 33000
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.454448017148982e-05,
      "loss": 1.2458,
      "step": 33500
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.4097892104322976e-05,
      "loss": 1.2428,
      "step": 34000
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.3651304037156126e-05,
      "loss": 1.2319,
      "step": 34500
    },
    {
      "epoch": 2.87,
      "learning_rate": 2.3204715969989283e-05,
      "loss": 1.2334,
      "step": 35000
    },
    {
      "epoch": 2.91,
      "learning_rate": 2.275812790282244e-05,
      "loss": 1.2275,
      "step": 35500
    },
    {
      "epoch": 2.95,
      "learning_rate": 2.2311539835655592e-05,
      "loss": 1.2234,
      "step": 36000
    },
    {
      "epoch": 2.99,
      "learning_rate": 2.186495176848875e-05,
      "loss": 1.2297,
      "step": 36500
    },
    {
      "epoch": 3.03,
      "learning_rate": 2.14183637013219e-05,
      "loss": 1.2149,
      "step": 37000
    },
    {
      "epoch": 3.07,
      "learning_rate": 2.0971775634155054e-05,
      "loss": 1.2124,
      "step": 37500
    },
    {
      "epoch": 3.12,
      "learning_rate": 2.052518756698821e-05,
      "loss": 1.2099,
      "step": 38000
    },
    {
      "epoch": 3.16,
      "learning_rate": 2.0078599499821364e-05,
      "loss": 1.1997,
      "step": 38500
    },
    {
      "epoch": 3.2,
      "learning_rate": 1.963201143265452e-05,
      "loss": 1.2052,
      "step": 39000
    },
    {
      "epoch": 3.24,
      "learning_rate": 1.9185423365487677e-05,
      "loss": 1.2032,
      "step": 39500
    },
    {
      "epoch": 3.28,
      "learning_rate": 1.873883529832083e-05,
      "loss": 1.1958,
      "step": 40000
    },
    {
      "epoch": 3.32,
      "learning_rate": 1.8292247231153983e-05,
      "loss": 1.1938,
      "step": 40500
    },
    {
      "epoch": 3.36,
      "learning_rate": 1.784565916398714e-05,
      "loss": 1.1927,
      "step": 41000
    },
    {
      "epoch": 3.4,
      "learning_rate": 1.7399071096820292e-05,
      "loss": 1.1898,
      "step": 41500
    },
    {
      "epoch": 3.44,
      "learning_rate": 1.695248302965345e-05,
      "loss": 1.1871,
      "step": 42000
    },
    {
      "epoch": 3.48,
      "learning_rate": 1.6505894962486605e-05,
      "loss": 1.1847,
      "step": 42500
    },
    {
      "epoch": 3.53,
      "learning_rate": 1.6059306895319758e-05,
      "loss": 1.1805,
      "step": 43000
    },
    {
      "epoch": 3.57,
      "learning_rate": 1.561271882815291e-05,
      "loss": 1.182,
      "step": 43500
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.5166130760986066e-05,
      "loss": 1.1766,
      "step": 44000
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.4719542693819222e-05,
      "loss": 1.177,
      "step": 44500
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.4272954626652377e-05,
      "loss": 1.1756,
      "step": 45000
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.3826366559485532e-05,
      "loss": 1.1744,
      "step": 45500
    },
    {
      "epoch": 3.77,
      "learning_rate": 1.3379778492318687e-05,
      "loss": 1.172,
      "step": 46000
    },
    {
      "epoch": 3.81,
      "learning_rate": 1.293319042515184e-05,
      "loss": 1.1656,
      "step": 46500
    },
    {
      "epoch": 3.85,
      "learning_rate": 1.2486602357984996e-05,
      "loss": 1.1658,
      "step": 47000
    },
    {
      "epoch": 3.89,
      "learning_rate": 1.2040014290818149e-05,
      "loss": 1.1656,
      "step": 47500
    },
    {
      "epoch": 3.94,
      "learning_rate": 1.1593426223651305e-05,
      "loss": 1.1576,
      "step": 48000
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.114683815648446e-05,
      "loss": 1.1572,
      "step": 48500
    },
    {
      "epoch": 4.02,
      "learning_rate": 1.0700250089317613e-05,
      "loss": 1.1591,
      "step": 49000
    },
    {
      "epoch": 4.06,
      "learning_rate": 1.0253662022150768e-05,
      "loss": 1.1567,
      "step": 49500
    },
    {
      "epoch": 4.1,
      "learning_rate": 9.807073954983924e-06,
      "loss": 1.153,
      "step": 50000
    },
    {
      "epoch": 4.14,
      "learning_rate": 9.360485887817077e-06,
      "loss": 1.1554,
      "step": 50500
    },
    {
      "epoch": 4.18,
      "learning_rate": 8.913897820650232e-06,
      "loss": 1.1463,
      "step": 51000
    },
    {
      "epoch": 4.22,
      "learning_rate": 8.467309753483389e-06,
      "loss": 1.1527,
      "step": 51500
    },
    {
      "epoch": 4.26,
      "learning_rate": 8.020721686316542e-06,
      "loss": 1.15,
      "step": 52000
    },
    {
      "epoch": 4.3,
      "learning_rate": 7.574133619149697e-06,
      "loss": 1.1503,
      "step": 52500
    },
    {
      "epoch": 4.35,
      "learning_rate": 7.127545551982852e-06,
      "loss": 1.1455,
      "step": 53000
    },
    {
      "epoch": 4.39,
      "learning_rate": 6.680957484816006e-06,
      "loss": 1.1473,
      "step": 53500
    },
    {
      "epoch": 4.43,
      "learning_rate": 6.2343694176491604e-06,
      "loss": 1.1453,
      "step": 54000
    },
    {
      "epoch": 4.47,
      "learning_rate": 5.787781350482315e-06,
      "loss": 1.1441,
      "step": 54500
    },
    {
      "epoch": 4.51,
      "learning_rate": 5.34119328331547e-06,
      "loss": 1.1399,
      "step": 55000
    },
    {
      "epoch": 4.55,
      "learning_rate": 4.894605216148625e-06,
      "loss": 1.1354,
      "step": 55500
    },
    {
      "epoch": 4.59,
      "learning_rate": 4.448017148981779e-06,
      "loss": 1.1371,
      "step": 56000
    },
    {
      "epoch": 4.63,
      "learning_rate": 4.001429081814934e-06,
      "loss": 1.1399,
      "step": 56500
    },
    {
      "epoch": 4.67,
      "learning_rate": 3.554841014648089e-06,
      "loss": 1.1317,
      "step": 57000
    },
    {
      "epoch": 4.71,
      "learning_rate": 3.1082529474812435e-06,
      "loss": 1.1412,
      "step": 57500
    },
    {
      "epoch": 4.76,
      "learning_rate": 2.6616648803143982e-06,
      "loss": 1.1368,
      "step": 58000
    },
    {
      "epoch": 4.8,
      "learning_rate": 2.2150768131475525e-06,
      "loss": 1.1345,
      "step": 58500
    },
    {
      "epoch": 4.84,
      "learning_rate": 1.7684887459807075e-06,
      "loss": 1.1316,
      "step": 59000
    },
    {
      "epoch": 4.88,
      "learning_rate": 1.3219006788138622e-06,
      "loss": 1.1362,
      "step": 59500
    },
    {
      "epoch": 4.92,
      "learning_rate": 8.753126116470169e-07,
      "loss": 1.1358,
      "step": 60000
    }
  ],
  "max_steps": 60980,
  "num_train_epochs": 5,
  "total_flos": 5.053516613471232e+17,
  "trial_name": null,
  "trial_params": null
}