{
  "best_metric": 0.9143675417661098,
  "best_model_checkpoint": "./save_models/qnli/roberta-base_lr1e-05_run0/checkpoint-17676",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 58920,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "learning_rate": 1.4140271493212672e-06,
      "loss": 0.6933,
      "step": 500
    },
    {
      "epoch": 0.17,
      "learning_rate": 2.8280542986425343e-06,
      "loss": 0.542,
      "step": 1000
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.242081447963801e-06,
      "loss": 0.4328,
      "step": 1500
    },
    {
      "epoch": 0.34,
      "learning_rate": 5.656108597285069e-06,
      "loss": 0.3932,
      "step": 2000
    },
    {
      "epoch": 0.42,
      "learning_rate": 7.070135746606336e-06,
      "loss": 0.3613,
      "step": 2500
    },
    {
      "epoch": 0.51,
      "learning_rate": 8.484162895927603e-06,
      "loss": 0.3553,
      "step": 3000
    },
    {
      "epoch": 0.59,
      "learning_rate": 9.89819004524887e-06,
      "loss": 0.3362,
      "step": 3500
    },
    {
      "epoch": 0.68,
      "learning_rate": 9.916221291347683e-06,
      "loss": 0.3362,
      "step": 4000
    },
    {
      "epoch": 0.76,
      "learning_rate": 9.825942510472339e-06,
      "loss": 0.3298,
      "step": 4500
    },
    {
      "epoch": 0.85,
      "learning_rate": 9.735663729596995e-06,
      "loss": 0.3075,
      "step": 5000
    },
    {
      "epoch": 0.93,
      "learning_rate": 9.645384948721653e-06,
      "loss": 0.3071,
      "step": 5500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8982338902147972,
      "eval_loss": 0.273178368806839,
      "eval_runtime": 14.5434,
      "eval_samples_per_second": 720.258,
      "eval_steps_per_second": 45.038,
      "step": 5892
    },
    {
      "epoch": 1.02,
      "learning_rate": 9.55510616784631e-06,
      "loss": 0.2836,
      "step": 6000
    },
    {
      "epoch": 1.1,
      "learning_rate": 9.464827386970968e-06,
      "loss": 0.2656,
      "step": 6500
    },
    {
      "epoch": 1.19,
      "learning_rate": 9.374548606095624e-06,
      "loss": 0.2548,
      "step": 7000
    },
    {
      "epoch": 1.27,
      "learning_rate": 9.28426982522028e-06,
      "loss": 0.2535,
      "step": 7500
    },
    {
      "epoch": 1.36,
      "learning_rate": 9.193991044344937e-06,
      "loss": 0.258,
      "step": 8000
    },
    {
      "epoch": 1.44,
      "learning_rate": 9.103712263469595e-06,
      "loss": 0.2539,
      "step": 8500
    },
    {
      "epoch": 1.53,
      "learning_rate": 9.013433482594251e-06,
      "loss": 0.2465,
      "step": 9000
    },
    {
      "epoch": 1.61,
      "learning_rate": 8.92315470171891e-06,
      "loss": 0.2586,
      "step": 9500
    },
    {
      "epoch": 1.7,
      "learning_rate": 8.832875920843566e-06,
      "loss": 0.2407,
      "step": 10000
    },
    {
      "epoch": 1.78,
      "learning_rate": 8.742597139968224e-06,
      "loss": 0.2453,
      "step": 10500
    },
    {
      "epoch": 1.87,
      "learning_rate": 8.65231835909288e-06,
      "loss": 0.2506,
      "step": 11000
    },
    {
      "epoch": 1.95,
      "learning_rate": 8.562039578217536e-06,
      "loss": 0.2426,
      "step": 11500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.911217183770883,
      "eval_loss": 0.24326153099536896,
      "eval_runtime": 14.4104,
      "eval_samples_per_second": 726.905,
      "eval_steps_per_second": 45.453,
      "step": 11784
    },
    {
      "epoch": 2.04,
      "learning_rate": 8.471760797342193e-06,
      "loss": 0.2253,
      "step": 12000
    },
    {
      "epoch": 2.12,
      "learning_rate": 8.381482016466849e-06,
      "loss": 0.2004,
      "step": 12500
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.291203235591507e-06,
      "loss": 0.2032,
      "step": 13000
    },
    {
      "epoch": 2.29,
      "learning_rate": 8.200924454716163e-06,
      "loss": 0.2098,
      "step": 13500
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.110645673840822e-06,
      "loss": 0.1996,
      "step": 14000
    },
    {
      "epoch": 2.46,
      "learning_rate": 8.020366892965478e-06,
      "loss": 0.2065,
      "step": 14500
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.930088112090136e-06,
      "loss": 0.2011,
      "step": 15000
    },
    {
      "epoch": 2.63,
      "learning_rate": 7.839809331214792e-06,
      "loss": 0.2135,
      "step": 15500
    },
    {
      "epoch": 2.72,
      "learning_rate": 7.749530550339449e-06,
      "loss": 0.2116,
      "step": 16000
    },
    {
      "epoch": 2.8,
      "learning_rate": 7.659251769464105e-06,
      "loss": 0.221,
      "step": 16500
    },
    {
      "epoch": 2.89,
      "learning_rate": 7.568972988588763e-06,
      "loss": 0.2179,
      "step": 17000
    },
    {
      "epoch": 2.97,
      "learning_rate": 7.478694207713419e-06,
      "loss": 0.2141,
      "step": 17500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9143675417661098,
      "eval_loss": 0.27570974826812744,
      "eval_runtime": 14.419,
      "eval_samples_per_second": 726.472,
      "eval_steps_per_second": 45.426,
      "step": 17676
    },
    {
      "epoch": 3.05,
      "learning_rate": 7.388415426838077e-06,
      "loss": 0.1871,
      "step": 18000
    },
    {
      "epoch": 3.14,
      "learning_rate": 7.298136645962734e-06,
      "loss": 0.183,
      "step": 18500
    },
    {
      "epoch": 3.22,
      "learning_rate": 7.207857865087391e-06,
      "loss": 0.1757,
      "step": 19000
    },
    {
      "epoch": 3.31,
      "learning_rate": 7.117579084212047e-06,
      "loss": 0.1783,
      "step": 19500
    },
    {
      "epoch": 3.39,
      "learning_rate": 7.027300303336704e-06,
      "loss": 0.1787,
      "step": 20000
    },
    {
      "epoch": 3.48,
      "learning_rate": 6.937021522461362e-06,
      "loss": 0.1834,
      "step": 20500
    },
    {
      "epoch": 3.56,
      "learning_rate": 6.846742741586018e-06,
      "loss": 0.1722,
      "step": 21000
    },
    {
      "epoch": 3.65,
      "learning_rate": 6.756463960710675e-06,
      "loss": 0.1894,
      "step": 21500
    },
    {
      "epoch": 3.73,
      "learning_rate": 6.6661851798353315e-06,
      "loss": 0.1708,
      "step": 22000
    },
    {
      "epoch": 3.82,
      "learning_rate": 6.57590639895999e-06,
      "loss": 0.1962,
      "step": 22500
    },
    {
      "epoch": 3.9,
      "learning_rate": 6.485627618084646e-06,
      "loss": 0.1831,
      "step": 23000
    },
    {
      "epoch": 3.99,
      "learning_rate": 6.395348837209303e-06,
      "loss": 0.1893,
      "step": 23500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9129355608591886,
      "eval_loss": 0.39158299565315247,
      "eval_runtime": 14.3752,
      "eval_samples_per_second": 728.688,
      "eval_steps_per_second": 45.565,
      "step": 23568
    },
    {
      "epoch": 4.07,
      "learning_rate": 6.3050700563339595e-06,
      "loss": 0.1408,
      "step": 24000
    },
    {
      "epoch": 4.16,
      "learning_rate": 6.214791275458617e-06,
      "loss": 0.1505,
      "step": 24500
    },
    {
      "epoch": 4.24,
      "learning_rate": 6.124512494583274e-06,
      "loss": 0.1364,
      "step": 25000
    },
    {
      "epoch": 4.33,
      "learning_rate": 6.034233713707931e-06,
      "loss": 0.1519,
      "step": 25500
    },
    {
      "epoch": 4.41,
      "learning_rate": 5.943954932832587e-06,
      "loss": 0.1367,
      "step": 26000
    },
    {
      "epoch": 4.5,
      "learning_rate": 5.853676151957245e-06,
      "loss": 0.154,
      "step": 26500
    },
    {
      "epoch": 4.58,
      "learning_rate": 5.763397371081901e-06,
      "loss": 0.1437,
      "step": 27000
    },
    {
      "epoch": 4.67,
      "learning_rate": 5.673118590206559e-06,
      "loss": 0.1496,
      "step": 27500
    },
    {
      "epoch": 4.75,
      "learning_rate": 5.582839809331215e-06,
      "loss": 0.1668,
      "step": 28000
    },
    {
      "epoch": 4.84,
      "learning_rate": 5.492561028455872e-06,
      "loss": 0.1448,
      "step": 28500
    },
    {
      "epoch": 4.92,
      "learning_rate": 5.402282247580529e-06,
      "loss": 0.1638,
      "step": 29000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9121718377088306,
      "eval_loss": 0.4099659323692322,
      "eval_runtime": 14.3864,
      "eval_samples_per_second": 728.119,
      "eval_steps_per_second": 45.529,
      "step": 29460
    },
    {
      "epoch": 5.01,
      "learning_rate": 5.312003466705185e-06,
      "loss": 0.1375,
      "step": 29500
    },
    {
      "epoch": 5.09,
      "learning_rate": 5.221724685829843e-06,
      "loss": 0.1032,
      "step": 30000
    },
    {
      "epoch": 5.18,
      "learning_rate": 5.1314459049545e-06,
      "loss": 0.1106,
      "step": 30500
    },
    {
      "epoch": 5.26,
      "learning_rate": 5.041167124079157e-06,
      "loss": 0.1091,
      "step": 31000
    },
    {
      "epoch": 5.35,
      "learning_rate": 4.950888343203813e-06,
      "loss": 0.1094,
      "step": 31500
    },
    {
      "epoch": 5.43,
      "learning_rate": 4.86060956232847e-06,
      "loss": 0.1223,
      "step": 32000
    },
    {
      "epoch": 5.52,
      "learning_rate": 4.7703307814531275e-06,
      "loss": 0.1084,
      "step": 32500
    },
    {
      "epoch": 5.6,
      "learning_rate": 4.680052000577785e-06,
      "loss": 0.1097,
      "step": 33000
    },
    {
      "epoch": 5.69,
      "learning_rate": 4.589773219702441e-06,
      "loss": 0.1216,
      "step": 33500
    },
    {
      "epoch": 5.77,
      "learning_rate": 4.499494438827098e-06,
      "loss": 0.1191,
      "step": 34000
    },
    {
      "epoch": 5.86,
      "learning_rate": 4.4092156579517554e-06,
      "loss": 0.1215,
      "step": 34500
    },
    {
      "epoch": 5.94,
      "learning_rate": 4.318936877076413e-06,
      "loss": 0.1356,
      "step": 35000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9077804295942721,
      "eval_loss": 0.484077125787735,
      "eval_runtime": 14.4023,
      "eval_samples_per_second": 727.317,
      "eval_steps_per_second": 45.479,
      "step": 35352
    },
    {
      "epoch": 6.03,
      "learning_rate": 4.228658096201069e-06,
      "loss": 0.1124,
      "step": 35500
    },
    {
      "epoch": 6.11,
      "learning_rate": 4.138379315325726e-06,
      "loss": 0.0893,
      "step": 36000
    },
    {
      "epoch": 6.19,
      "learning_rate": 4.048100534450383e-06,
      "loss": 0.0763,
      "step": 36500
    },
    {
      "epoch": 6.28,
      "learning_rate": 3.9578217535750406e-06,
      "loss": 0.0889,
      "step": 37000
    },
    {
      "epoch": 6.36,
      "learning_rate": 3.867542972699697e-06,
      "loss": 0.0941,
      "step": 37500
    },
    {
      "epoch": 6.45,
      "learning_rate": 3.777264191824354e-06,
      "loss": 0.0931,
      "step": 38000
    },
    {
      "epoch": 6.53,
      "learning_rate": 3.686985410949011e-06,
      "loss": 0.0916,
      "step": 38500
    },
    {
      "epoch": 6.62,
      "learning_rate": 3.596706630073668e-06,
      "loss": 0.0916,
      "step": 39000
    },
    {
      "epoch": 6.7,
      "learning_rate": 3.5064278491983244e-06,
      "loss": 0.1038,
      "step": 39500
    },
    {
      "epoch": 6.79,
      "learning_rate": 3.4161490683229816e-06,
      "loss": 0.089,
      "step": 40000
    },
    {
      "epoch": 6.87,
      "learning_rate": 3.3258702874476384e-06,
      "loss": 0.1008,
      "step": 40500
    },
    {
      "epoch": 6.96,
      "learning_rate": 3.2355915065722956e-06,
      "loss": 0.0867,
      "step": 41000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9097852028639618,
      "eval_loss": 0.5044658184051514,
      "eval_runtime": 14.3742,
      "eval_samples_per_second": 728.736,
      "eval_steps_per_second": 45.568,
      "step": 41244
    },
    {
      "epoch": 7.04,
      "learning_rate": 3.1453127256969523e-06,
      "loss": 0.0746,
      "step": 41500
    },
    {
      "epoch": 7.13,
      "learning_rate": 3.055033944821609e-06,
      "loss": 0.0624,
      "step": 42000
    },
    {
      "epoch": 7.21,
      "learning_rate": 2.9647551639462663e-06,
      "loss": 0.0695,
      "step": 42500
    },
    {
      "epoch": 7.3,
      "learning_rate": 2.874476383070923e-06,
      "loss": 0.0804,
      "step": 43000
    },
    {
      "epoch": 7.38,
      "learning_rate": 2.7841976021955803e-06,
      "loss": 0.0695,
      "step": 43500
    },
    {
      "epoch": 7.47,
      "learning_rate": 2.693918821320237e-06,
      "loss": 0.0737,
      "step": 44000
    },
    {
      "epoch": 7.55,
      "learning_rate": 2.6036400404448942e-06,
      "loss": 0.0773,
      "step": 44500
    },
    {
      "epoch": 7.64,
      "learning_rate": 2.513361259569551e-06,
      "loss": 0.0756,
      "step": 45000
    },
    {
      "epoch": 7.72,
      "learning_rate": 2.4230824786942078e-06,
      "loss": 0.0697,
      "step": 45500
    },
    {
      "epoch": 7.81,
      "learning_rate": 2.332803697818865e-06,
      "loss": 0.0731,
      "step": 46000
    },
    {
      "epoch": 7.89,
      "learning_rate": 2.2425249169435217e-06,
      "loss": 0.0584,
      "step": 46500
    },
    {
      "epoch": 7.98,
      "learning_rate": 2.152246136068179e-06,
      "loss": 0.071,
      "step": 47000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9092124105011933,
      "eval_loss": 0.5765216946601868,
      "eval_runtime": 14.3597,
      "eval_samples_per_second": 729.473,
      "eval_steps_per_second": 45.614,
      "step": 47136
    },
    {
      "epoch": 8.06,
      "learning_rate": 2.0619673551928357e-06,
      "loss": 0.0614,
      "step": 47500
    },
    {
      "epoch": 8.15,
      "learning_rate": 1.9716885743174925e-06,
      "loss": 0.0588,
      "step": 48000
    },
    {
      "epoch": 8.23,
      "learning_rate": 1.8814097934421497e-06,
      "loss": 0.0575,
      "step": 48500
    },
    {
      "epoch": 8.32,
      "learning_rate": 1.7911310125668062e-06,
      "loss": 0.0454,
      "step": 49000
    },
    {
      "epoch": 8.4,
      "learning_rate": 1.7008522316914632e-06,
      "loss": 0.0554,
      "step": 49500
    },
    {
      "epoch": 8.49,
      "learning_rate": 1.6105734508161202e-06,
      "loss": 0.0559,
      "step": 50000
    },
    {
      "epoch": 8.57,
      "learning_rate": 1.5202946699407772e-06,
      "loss": 0.0561,
      "step": 50500
    },
    {
      "epoch": 8.66,
      "learning_rate": 1.4300158890654341e-06,
      "loss": 0.0505,
      "step": 51000
    },
    {
      "epoch": 8.74,
      "learning_rate": 1.3397371081900911e-06,
      "loss": 0.0552,
      "step": 51500
    },
    {
      "epoch": 8.83,
      "learning_rate": 1.249458327314748e-06,
      "loss": 0.0604,
      "step": 52000
    },
    {
      "epoch": 8.91,
      "learning_rate": 1.1591795464394049e-06,
      "loss": 0.0626,
      "step": 52500
    },
    {
      "epoch": 9.0,
      "learning_rate": 1.0689007655640619e-06,
      "loss": 0.0636,
      "step": 53000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9113126491646778,
      "eval_loss": 0.5765402913093567,
      "eval_runtime": 14.3825,
      "eval_samples_per_second": 728.315,
      "eval_steps_per_second": 45.541,
      "step": 53028
    },
    {
      "epoch": 9.08,
      "learning_rate": 9.786219846887188e-07,
      "loss": 0.0439,
      "step": 53500
    },
    {
      "epoch": 9.16,
      "learning_rate": 8.883432038133757e-07,
      "loss": 0.0447,
      "step": 54000
    },
    {
      "epoch": 9.25,
      "learning_rate": 7.980644229380327e-07,
      "loss": 0.0391,
      "step": 54500
    },
    {
      "epoch": 9.33,
      "learning_rate": 7.077856420626897e-07,
      "loss": 0.0404,
      "step": 55000
    },
    {
      "epoch": 9.42,
      "learning_rate": 6.175068611873466e-07,
      "loss": 0.0535,
      "step": 55500
    },
    {
      "epoch": 9.5,
      "learning_rate": 5.272280803120034e-07,
      "loss": 0.038,
      "step": 56000
    },
    {
      "epoch": 9.59,
      "learning_rate": 4.3694929943666047e-07,
      "loss": 0.047,
      "step": 56500
    },
    {
      "epoch": 9.67,
      "learning_rate": 3.4667051856131735e-07,
      "loss": 0.0414,
      "step": 57000
    },
    {
      "epoch": 9.76,
      "learning_rate": 2.5639173768597433e-07,
      "loss": 0.0447,
      "step": 57500
    },
    {
      "epoch": 9.84,
      "learning_rate": 1.6611295681063126e-07,
      "loss": 0.0364,
      "step": 58000
    },
    {
      "epoch": 9.93,
      "learning_rate": 7.583417593528817e-08,
      "loss": 0.0492,
      "step": 58500
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9117899761336515,
      "eval_loss": 0.6481573581695557,
      "eval_runtime": 14.4339,
      "eval_samples_per_second": 725.723,
      "eval_steps_per_second": 45.379,
      "step": 58920
    },
    {
      "epoch": 10.0,
      "step": 58920,
      "total_flos": 4.325458183470288e+16,
      "train_loss": 0.15599560614596883,
      "train_runtime": 4172.8624,
      "train_samples_per_second": 225.907,
      "train_steps_per_second": 14.12
    }
  ],
  "logging_steps": 500,
  "max_steps": 58920,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 4.325458183470288e+16,
  "trial_name": null,
  "trial_params": null
}