{
  "best_metric": 0.47372516989707947,
  "best_model_checkpoint": "/data/jcanete/all_results/xnli/albeto_xlarge/epochs_2_bs_16_lr_5e-6/checkpoint-22000",
  "epoch": 2.0,
  "global_step": 49088,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 4.9493766297262065e-06,
      "loss": 1.0212,
      "step": 500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.898549543676663e-06,
      "loss": 0.8644,
      "step": 1000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.847722457627119e-06,
      "loss": 0.765,
      "step": 1500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.7967935136897e-06,
      "loss": 0.7077,
      "step": 2000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.7052208835341366,
      "eval_loss": 0.6958776712417603,
      "eval_runtime": 15.7507,
      "eval_samples_per_second": 158.088,
      "eval_steps_per_second": 9.904,
      "step": 2000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.7458645697522824e-06,
      "loss": 0.6694,
      "step": 2500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.694935625814864e-06,
      "loss": 0.6511,
      "step": 3000
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.644006681877445e-06,
      "loss": 0.6403,
      "step": 3500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.593179595827901e-06,
      "loss": 0.6062,
      "step": 4000
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.7610441767068273,
      "eval_loss": 0.6156352162361145,
      "eval_runtime": 15.8305,
      "eval_samples_per_second": 157.291,
      "eval_steps_per_second": 9.854,
      "step": 4000
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.542250651890483e-06,
      "loss": 0.6191,
      "step": 4500
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.491321707953064e-06,
      "loss": 0.6144,
      "step": 5000
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.440392764015646e-06,
      "loss": 0.6089,
      "step": 5500
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.389463820078227e-06,
      "loss": 0.584,
      "step": 6000
    },
    {
      "epoch": 0.24,
      "eval_accuracy": 0.7698795180722892,
      "eval_loss": 0.6093908548355103,
      "eval_runtime": 15.9514,
      "eval_samples_per_second": 156.099,
      "eval_steps_per_second": 9.78,
      "step": 6000
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.338636734028683e-06,
      "loss": 0.583,
      "step": 6500
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.287707790091265e-06,
      "loss": 0.5849,
      "step": 7000
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.236778846153847e-06,
      "loss": 0.5819,
      "step": 7500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.185849902216428e-06,
      "loss": 0.5567,
      "step": 8000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.7718875502008032,
      "eval_loss": 0.5614191889762878,
      "eval_runtime": 15.8137,
      "eval_samples_per_second": 157.458,
      "eval_steps_per_second": 9.865,
      "step": 8000
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.134920958279009e-06,
      "loss": 0.5554,
      "step": 8500
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.083992014341591e-06,
      "loss": 0.5476,
      "step": 9000
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.033063070404172e-06,
      "loss": 0.5531,
      "step": 9500
    },
    {
      "epoch": 0.41,
      "learning_rate": 3.982134126466754e-06,
      "loss": 0.5464,
      "step": 10000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.7847389558232932,
      "eval_loss": 0.5493358373641968,
      "eval_runtime": 15.7384,
      "eval_samples_per_second": 158.211,
      "eval_steps_per_second": 9.912,
      "step": 10000
    },
    {
      "epoch": 0.43,
      "learning_rate": 3.9312051825293356e-06,
      "loss": 0.5482,
      "step": 10500
    },
    {
      "epoch": 0.45,
      "learning_rate": 3.880378096479792e-06,
      "loss": 0.5451,
      "step": 11000
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.8294491525423735e-06,
      "loss": 0.5402,
      "step": 11500
    },
    {
      "epoch": 0.49,
      "learning_rate": 3.7786220664928295e-06,
      "loss": 0.5393,
      "step": 12000
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.7911646586345381,
      "eval_loss": 0.5225389003753662,
      "eval_runtime": 15.9954,
      "eval_samples_per_second": 155.67,
      "eval_steps_per_second": 9.753,
      "step": 12000
    },
    {
      "epoch": 0.51,
      "learning_rate": 3.7276931225554107e-06,
      "loss": 0.5344,
      "step": 12500
    },
    {
      "epoch": 0.53,
      "learning_rate": 3.6767641786179927e-06,
      "loss": 0.5495,
      "step": 13000
    },
    {
      "epoch": 0.55,
      "learning_rate": 3.6258352346805742e-06,
      "loss": 0.5362,
      "step": 13500
    },
    {
      "epoch": 0.57,
      "learning_rate": 3.5750081486310302e-06,
      "loss": 0.5308,
      "step": 14000
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.7919678714859437,
      "eval_loss": 0.527259111404419,
      "eval_runtime": 16.0033,
      "eval_samples_per_second": 155.593,
      "eval_steps_per_second": 9.748,
      "step": 14000
    },
    {
      "epoch": 0.59,
      "learning_rate": 3.5240792046936118e-06,
      "loss": 0.5218,
      "step": 14500
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.4731502607561934e-06,
      "loss": 0.5421,
      "step": 15000
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.4222213168187745e-06,
      "loss": 0.5173,
      "step": 15500
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.371292372881356e-06,
      "loss": 0.5246,
      "step": 16000
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.7975903614457831,
      "eval_loss": 0.506671130657196,
      "eval_runtime": 15.9337,
      "eval_samples_per_second": 156.272,
      "eval_steps_per_second": 9.791,
      "step": 16000
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.3204652868318125e-06,
      "loss": 0.5237,
      "step": 16500
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.2695363428943936e-06,
      "loss": 0.5186,
      "step": 17000
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.2186073989569756e-06,
      "loss": 0.5267,
      "step": 17500
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.167678455019557e-06,
      "loss": 0.5075,
      "step": 18000
    },
    {
      "epoch": 0.73,
      "eval_accuracy": 0.7967871485943775,
      "eval_loss": 0.5081815719604492,
      "eval_runtime": 16.0036,
      "eval_samples_per_second": 155.59,
      "eval_steps_per_second": 9.748,
      "step": 18000
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.1167495110821384e-06,
      "loss": 0.5029,
      "step": 18500
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.06582056714472e-06,
      "loss": 0.5123,
      "step": 19000
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.0148916232073015e-06,
      "loss": 0.5031,
      "step": 19500
    },
    {
      "epoch": 0.81,
      "learning_rate": 2.9639626792698826e-06,
      "loss": 0.5008,
      "step": 20000
    },
    {
      "epoch": 0.81,
      "eval_accuracy": 0.804417670682731,
      "eval_loss": 0.49554213881492615,
      "eval_runtime": 16.1075,
      "eval_samples_per_second": 154.587,
      "eval_steps_per_second": 9.685,
      "step": 20000
    },
    {
      "epoch": 0.84,
      "learning_rate": 2.9130337353324646e-06,
      "loss": 0.5054,
      "step": 20500
    },
    {
      "epoch": 0.86,
      "learning_rate": 2.862104791395046e-06,
      "loss": 0.4996,
      "step": 21000
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.8111758474576274e-06,
      "loss": 0.4849,
      "step": 21500
    },
    {
      "epoch": 0.9,
      "learning_rate": 2.760246903520209e-06,
      "loss": 0.507,
      "step": 22000
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.8152610441767069,
      "eval_loss": 0.47372516989707947,
      "eval_runtime": 16.3654,
      "eval_samples_per_second": 152.15,
      "eval_steps_per_second": 9.532,
      "step": 22000
    },
    {
      "epoch": 0.92,
      "learning_rate": 2.70931795958279e-06,
      "loss": 0.4995,
      "step": 22500
    },
    {
      "epoch": 0.94,
      "learning_rate": 2.6584908735332465e-06,
      "loss": 0.5007,
      "step": 23000
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.607663787483703e-06,
      "loss": 0.4944,
      "step": 23500
    },
    {
      "epoch": 0.98,
      "learning_rate": 2.5567348435462845e-06,
      "loss": 0.5053,
      "step": 24000
    },
    {
      "epoch": 0.98,
      "eval_accuracy": 0.8072289156626506,
      "eval_loss": 0.49480196833610535,
      "eval_runtime": 16.2328,
      "eval_samples_per_second": 153.393,
      "eval_steps_per_second": 9.61,
      "step": 24000
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.5058058996088656e-06,
      "loss": 0.4946,
      "step": 24500
    },
    {
      "epoch": 1.02,
      "learning_rate": 2.4548769556714476e-06,
      "loss": 0.4133,
      "step": 25000
    },
    {
      "epoch": 1.04,
      "learning_rate": 2.4039480117340288e-06,
      "loss": 0.4103,
      "step": 25500
    },
    {
      "epoch": 1.06,
      "learning_rate": 2.3530190677966103e-06,
      "loss": 0.4063,
      "step": 26000
    },
    {
      "epoch": 1.06,
      "eval_accuracy": 0.8040160642570281,
      "eval_loss": 0.4989866316318512,
      "eval_runtime": 15.9626,
      "eval_samples_per_second": 155.99,
      "eval_steps_per_second": 9.773,
      "step": 26000
    },
    {
      "epoch": 1.08,
      "learning_rate": 2.3021919817470667e-06,
      "loss": 0.4123,
      "step": 26500
    },
    {
      "epoch": 1.1,
      "learning_rate": 2.2512630378096483e-06,
      "loss": 0.3875,
      "step": 27000
    },
    {
      "epoch": 1.12,
      "learning_rate": 2.2003340938722295e-06,
      "loss": 0.4103,
      "step": 27500
    },
    {
      "epoch": 1.14,
      "learning_rate": 2.149405149934811e-06,
      "loss": 0.4061,
      "step": 28000
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.797991967871486,
      "eval_loss": 0.5204265117645264,
      "eval_runtime": 16.59,
      "eval_samples_per_second": 150.09,
      "eval_steps_per_second": 9.403,
      "step": 28000
    },
    {
      "epoch": 1.16,
      "learning_rate": 2.0985780638852674e-06,
      "loss": 0.4178,
      "step": 28500
    },
    {
      "epoch": 1.18,
      "learning_rate": 2.047750977835724e-06,
      "loss": 0.4062,
      "step": 29000
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.9968220338983054e-06,
      "loss": 0.4142,
      "step": 29500
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.9458930899608866e-06,
      "loss": 0.4118,
      "step": 30000
    },
    {
      "epoch": 1.22,
      "eval_accuracy": 0.8120481927710843,
      "eval_loss": 0.5043097138404846,
      "eval_runtime": 15.9992,
      "eval_samples_per_second": 155.632,
      "eval_steps_per_second": 9.75,
      "step": 30000
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.8949641460234683e-06,
      "loss": 0.4032,
      "step": 30500
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.8440352020860497e-06,
      "loss": 0.4004,
      "step": 31000
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.793106258148631e-06,
      "loss": 0.3999,
      "step": 31500
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.7421773142112129e-06,
      "loss": 0.4057,
      "step": 32000
    },
    {
      "epoch": 1.3,
      "eval_accuracy": 0.8176706827309237,
      "eval_loss": 0.4870663285255432,
      "eval_runtime": 15.9746,
      "eval_samples_per_second": 155.872,
      "eval_steps_per_second": 9.765,
      "step": 32000
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.691350228161669e-06,
      "loss": 0.4008,
      "step": 32500
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.6404212842242504e-06,
      "loss": 0.4112,
      "step": 33000
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.589492340286832e-06,
      "loss": 0.4101,
      "step": 33500
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.5385633963494133e-06,
      "loss": 0.4108,
      "step": 34000
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.8052208835341366,
      "eval_loss": 0.5033333897590637,
      "eval_runtime": 16.0841,
      "eval_samples_per_second": 154.811,
      "eval_steps_per_second": 9.699,
      "step": 34000
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.487634452411995e-06,
      "loss": 0.3994,
      "step": 34500
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.4367055084745765e-06,
      "loss": 0.4046,
      "step": 35000
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.3857765645371578e-06,
      "loss": 0.3958,
      "step": 35500
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.334949478487614e-06,
      "loss": 0.3927,
      "step": 36000
    },
    {
      "epoch": 1.47,
      "eval_accuracy": 0.8068273092369478,
      "eval_loss": 0.4963078498840332,
      "eval_runtime": 15.7763,
      "eval_samples_per_second": 157.832,
      "eval_steps_per_second": 9.888,
      "step": 36000
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.2840205345501958e-06,
      "loss": 0.3971,
      "step": 36500
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.2330915906127772e-06,
      "loss": 0.3985,
      "step": 37000
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.1821626466753585e-06,
      "loss": 0.3922,
      "step": 37500
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.1312337027379401e-06,
      "loss": 0.4077,
      "step": 38000
    },
    {
      "epoch": 1.55,
      "eval_accuracy": 0.814859437751004,
      "eval_loss": 0.47995567321777344,
      "eval_runtime": 15.901,
      "eval_samples_per_second": 156.594,
      "eval_steps_per_second": 9.811,
      "step": 38000
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.0803047588005217e-06,
      "loss": 0.4048,
      "step": 38500
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.029375814863103e-06,
      "loss": 0.3926,
      "step": 39000
    },
    {
      "epoch": 1.61,
      "learning_rate": 9.784468709256846e-07,
      "loss": 0.4051,
      "step": 39500
    },
    {
      "epoch": 1.63,
      "learning_rate": 9.27517926988266e-07,
      "loss": 0.4029,
      "step": 40000
    },
    {
      "epoch": 1.63,
      "eval_accuracy": 0.810843373493976,
      "eval_loss": 0.4851303994655609,
      "eval_runtime": 15.8122,
      "eval_samples_per_second": 157.474,
      "eval_steps_per_second": 9.866,
      "step": 40000
    },
    {
      "epoch": 1.65,
      "learning_rate": 8.765889830508476e-07,
      "loss": 0.385,
      "step": 40500
    },
    {
      "epoch": 1.67,
      "learning_rate": 8.256600391134289e-07,
      "loss": 0.4012,
      "step": 41000
    },
    {
      "epoch": 1.69,
      "learning_rate": 7.747310951760105e-07,
      "loss": 0.3998,
      "step": 41500
    },
    {
      "epoch": 1.71,
      "learning_rate": 7.239040091264668e-07,
      "loss": 0.3926,
      "step": 42000
    },
    {
      "epoch": 1.71,
      "eval_accuracy": 0.8068273092369478,
      "eval_loss": 0.49413371086120605,
      "eval_runtime": 16.1684,
      "eval_samples_per_second": 154.004,
      "eval_steps_per_second": 9.648,
      "step": 42000
    },
    {
      "epoch": 1.73,
      "learning_rate": 6.729750651890484e-07,
      "loss": 0.396,
      "step": 42500
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.220461212516297e-07,
      "loss": 0.3896,
      "step": 43000
    },
    {
      "epoch": 1.77,
      "learning_rate": 5.711171773142113e-07,
      "loss": 0.3993,
      "step": 43500
    },
    {
      "epoch": 1.79,
      "learning_rate": 5.201882333767928e-07,
      "loss": 0.3895,
      "step": 44000
    },
    {
      "epoch": 1.79,
      "eval_accuracy": 0.8112449799196787,
      "eval_loss": 0.47773313522338867,
      "eval_runtime": 16.1399,
      "eval_samples_per_second": 154.276,
      "eval_steps_per_second": 9.665,
      "step": 44000
    },
    {
      "epoch": 1.81,
      "learning_rate": 4.694630052151239e-07,
      "loss": 0.3971,
      "step": 44500
    },
    {
      "epoch": 1.83,
      "learning_rate": 4.1853406127770535e-07,
      "loss": 0.3864,
      "step": 45000
    },
    {
      "epoch": 1.85,
      "learning_rate": 3.6770697522816176e-07,
      "loss": 0.3901,
      "step": 45500
    },
    {
      "epoch": 1.87,
      "learning_rate": 3.167780312907432e-07,
      "loss": 0.3806,
      "step": 46000
    },
    {
      "epoch": 1.87,
      "eval_accuracy": 0.8124497991967872,
      "eval_loss": 0.4864996075630188,
      "eval_runtime": 15.9489,
      "eval_samples_per_second": 156.123,
      "eval_steps_per_second": 9.781,
      "step": 46000
    },
    {
      "epoch": 1.89,
      "learning_rate": 2.6584908735332464e-07,
      "loss": 0.386,
      "step": 46500
    },
    {
      "epoch": 1.91,
      "learning_rate": 2.1492014341590616e-07,
      "loss": 0.3804,
      "step": 47000
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.6409305736636246e-07,
      "loss": 0.3886,
      "step": 47500
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.1316411342894395e-07,
      "loss": 0.3771,
      "step": 48000
    },
    {
      "epoch": 1.96,
      "eval_accuracy": 0.8156626506024096,
      "eval_loss": 0.4936372935771942,
      "eval_runtime": 16.2104,
      "eval_samples_per_second": 153.605,
      "eval_steps_per_second": 9.623,
      "step": 48000
    },
    {
      "epoch": 1.98,
      "learning_rate": 6.223516949152543e-08,
      "loss": 0.4011,
      "step": 48500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.130622555410691e-08,
      "loss": 0.394,
      "step": 49000
    },
    {
      "epoch": 2.0,
      "step": 49088,
      "total_flos": 2.2556354760831744e+16,
      "train_loss": 0.4839201732751133,
      "train_runtime": 15881.4239,
      "train_samples_per_second": 49.454,
      "train_steps_per_second": 3.091
    }
  ],
  "max_steps": 49088,
  "num_train_epochs": 2,
  "total_flos": 2.2556354760831744e+16,
  "trial_name": null,
  "trial_params": null
}