{
  "best_metric": 0.5007363557815552,
  "best_model_checkpoint": "/data/jcanete/all_results/xnli/albeto_base2/epochs_2_bs_16_lr_5e-5/checkpoint-44000",
  "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": 4.949478487614081e-05,
      "loss": 0.8958,
      "step": 500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.8985495436766624e-05,
      "loss": 0.7815,
      "step": 1000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.847722457627119e-05,
      "loss": 0.7453,
      "step": 1500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.7967935136897005e-05,
      "loss": 0.7032,
      "step": 2000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.7104417670682731,
      "eval_loss": 0.6837603449821472,
      "eval_runtime": 3.0721,
      "eval_samples_per_second": 810.518,
      "eval_steps_per_second": 50.779,
      "step": 2000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.745864569752282e-05,
      "loss": 0.6916,
      "step": 2500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.694935625814863e-05,
      "loss": 0.678,
      "step": 3000
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.644006681877445e-05,
      "loss": 0.6673,
      "step": 3500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.593077737940026e-05,
      "loss": 0.6418,
      "step": 4000
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.7413654618473896,
      "eval_loss": 0.6337006092071533,
      "eval_runtime": 1.3627,
      "eval_samples_per_second": 1827.229,
      "eval_steps_per_second": 114.477,
      "step": 4000
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.5421487940026075e-05,
      "loss": 0.6562,
      "step": 4500
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.491219850065189e-05,
      "loss": 0.6461,
      "step": 5000
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.440290906127771e-05,
      "loss": 0.648,
      "step": 5500
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.389361962190352e-05,
      "loss": 0.6207,
      "step": 6000
    },
    {
      "epoch": 0.24,
      "eval_accuracy": 0.7421686746987952,
      "eval_loss": 0.6351802945137024,
      "eval_runtime": 3.0952,
      "eval_samples_per_second": 804.479,
      "eval_steps_per_second": 50.401,
      "step": 6000
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.3385348761408085e-05,
      "loss": 0.6335,
      "step": 6500
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.28760593220339e-05,
      "loss": 0.633,
      "step": 7000
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.236676988265971e-05,
      "loss": 0.6173,
      "step": 7500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.185849902216428e-05,
      "loss": 0.6146,
      "step": 8000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.7534136546184739,
      "eval_loss": 0.6005685329437256,
      "eval_runtime": 1.3569,
      "eval_samples_per_second": 1835.093,
      "eval_steps_per_second": 114.97,
      "step": 8000
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.1349209582790094e-05,
      "loss": 0.6117,
      "step": 8500
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.083992014341591e-05,
      "loss": 0.6029,
      "step": 9000
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.033063070404172e-05,
      "loss": 0.607,
      "step": 9500
    },
    {
      "epoch": 0.41,
      "learning_rate": 3.9821341264667536e-05,
      "loss": 0.593,
      "step": 10000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.7706827309236948,
      "eval_loss": 0.5669934749603271,
      "eval_runtime": 1.3711,
      "eval_samples_per_second": 1816.031,
      "eval_steps_per_second": 113.775,
      "step": 10000
    },
    {
      "epoch": 0.43,
      "learning_rate": 3.931205182529335e-05,
      "loss": 0.5915,
      "step": 10500
    },
    {
      "epoch": 0.45,
      "learning_rate": 3.8802762385919165e-05,
      "loss": 0.593,
      "step": 11000
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.8293472946544985e-05,
      "loss": 0.5951,
      "step": 11500
    },
    {
      "epoch": 0.49,
      "learning_rate": 3.778418350717079e-05,
      "loss": 0.5794,
      "step": 12000
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.770281124497992,
      "eval_loss": 0.5802505016326904,
      "eval_runtime": 3.0751,
      "eval_samples_per_second": 809.728,
      "eval_steps_per_second": 50.73,
      "step": 12000
    },
    {
      "epoch": 0.51,
      "learning_rate": 3.7274894067796613e-05,
      "loss": 0.5824,
      "step": 12500
    },
    {
      "epoch": 0.53,
      "learning_rate": 3.676560462842243e-05,
      "loss": 0.6039,
      "step": 13000
    },
    {
      "epoch": 0.55,
      "learning_rate": 3.625631518904824e-05,
      "loss": 0.5873,
      "step": 13500
    },
    {
      "epoch": 0.57,
      "learning_rate": 3.5747025749674056e-05,
      "loss": 0.5857,
      "step": 14000
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.7690763052208835,
      "eval_loss": 0.5731454491615295,
      "eval_runtime": 1.374,
      "eval_samples_per_second": 1812.235,
      "eval_steps_per_second": 113.538,
      "step": 14000
    },
    {
      "epoch": 0.59,
      "learning_rate": 3.5237736310299876e-05,
      "loss": 0.5673,
      "step": 14500
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.4728446870925684e-05,
      "loss": 0.5966,
      "step": 15000
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.422017601043025e-05,
      "loss": 0.5709,
      "step": 15500
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.371190514993481e-05,
      "loss": 0.573,
      "step": 16000
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.791566265060241,
      "eval_loss": 0.5376591086387634,
      "eval_runtime": 1.3604,
      "eval_samples_per_second": 1830.288,
      "eval_steps_per_second": 114.669,
      "step": 16000
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.3202615710560626e-05,
      "loss": 0.569,
      "step": 16500
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.2694344850065186e-05,
      "loss": 0.5577,
      "step": 17000
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.218505541069101e-05,
      "loss": 0.5709,
      "step": 17500
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.1675765971316814e-05,
      "loss": 0.5475,
      "step": 18000
    },
    {
      "epoch": 0.73,
      "eval_accuracy": 0.7803212851405622,
      "eval_loss": 0.5562066435813904,
      "eval_runtime": 1.3712,
      "eval_samples_per_second": 1815.926,
      "eval_steps_per_second": 113.769,
      "step": 18000
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.1166476531942635e-05,
      "loss": 0.5466,
      "step": 18500
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.065718709256845e-05,
      "loss": 0.5616,
      "step": 19000
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.0147897653194263e-05,
      "loss": 0.563,
      "step": 19500
    },
    {
      "epoch": 0.81,
      "learning_rate": 2.963860821382008e-05,
      "loss": 0.5438,
      "step": 20000
    },
    {
      "epoch": 0.81,
      "eval_accuracy": 0.7843373493975904,
      "eval_loss": 0.533427894115448,
      "eval_runtime": 3.0567,
      "eval_samples_per_second": 814.606,
      "eval_steps_per_second": 51.036,
      "step": 20000
    },
    {
      "epoch": 0.84,
      "learning_rate": 2.9129318774445895e-05,
      "loss": 0.5547,
      "step": 20500
    },
    {
      "epoch": 0.86,
      "learning_rate": 2.862002933507171e-05,
      "loss": 0.5488,
      "step": 21000
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.8110739895697523e-05,
      "loss": 0.5357,
      "step": 21500
    },
    {
      "epoch": 0.9,
      "learning_rate": 2.760145045632334e-05,
      "loss": 0.5522,
      "step": 22000
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.7927710843373494,
      "eval_loss": 0.5198507905006409,
      "eval_runtime": 1.361,
      "eval_samples_per_second": 1829.583,
      "eval_steps_per_second": 114.624,
      "step": 22000
    },
    {
      "epoch": 0.92,
      "learning_rate": 2.709216101694915e-05,
      "loss": 0.5422,
      "step": 22500
    },
    {
      "epoch": 0.94,
      "learning_rate": 2.6582871577574968e-05,
      "loss": 0.5481,
      "step": 23000
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.6073582138200786e-05,
      "loss": 0.5376,
      "step": 23500
    },
    {
      "epoch": 0.98,
      "learning_rate": 2.5564292698826596e-05,
      "loss": 0.5401,
      "step": 24000
    },
    {
      "epoch": 0.98,
      "eval_accuracy": 0.7799196787148595,
      "eval_loss": 0.5350679159164429,
      "eval_runtime": 3.1079,
      "eval_samples_per_second": 801.173,
      "eval_steps_per_second": 50.194,
      "step": 24000
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.5055003259452414e-05,
      "loss": 0.5375,
      "step": 24500
    },
    {
      "epoch": 1.02,
      "learning_rate": 2.4545713820078228e-05,
      "loss": 0.4512,
      "step": 25000
    },
    {
      "epoch": 1.04,
      "learning_rate": 2.403744295958279e-05,
      "loss": 0.4575,
      "step": 25500
    },
    {
      "epoch": 1.06,
      "learning_rate": 2.3528153520208606e-05,
      "loss": 0.4434,
      "step": 26000
    },
    {
      "epoch": 1.06,
      "eval_accuracy": 0.797991967871486,
      "eval_loss": 0.5054460763931274,
      "eval_runtime": 1.3491,
      "eval_samples_per_second": 1845.65,
      "eval_steps_per_second": 115.631,
      "step": 26000
    },
    {
      "epoch": 1.08,
      "learning_rate": 2.301886408083442e-05,
      "loss": 0.4493,
      "step": 26500
    },
    {
      "epoch": 1.1,
      "learning_rate": 2.2509574641460234e-05,
      "loss": 0.4281,
      "step": 27000
    },
    {
      "epoch": 1.12,
      "learning_rate": 2.2001303780964798e-05,
      "loss": 0.4491,
      "step": 27500
    },
    {
      "epoch": 1.14,
      "learning_rate": 2.1492014341590612e-05,
      "loss": 0.4483,
      "step": 28000
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.7959839357429719,
      "eval_loss": 0.5253944396972656,
      "eval_runtime": 1.3702,
      "eval_samples_per_second": 1817.309,
      "eval_steps_per_second": 113.856,
      "step": 28000
    },
    {
      "epoch": 1.16,
      "learning_rate": 2.098374348109518e-05,
      "loss": 0.4543,
      "step": 28500
    },
    {
      "epoch": 1.18,
      "learning_rate": 2.0474454041720993e-05,
      "loss": 0.4405,
      "step": 29000
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.9965164602346807e-05,
      "loss": 0.4563,
      "step": 29500
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.945587516297262e-05,
      "loss": 0.4488,
      "step": 30000
    },
    {
      "epoch": 1.22,
      "eval_accuracy": 0.7995983935742972,
      "eval_loss": 0.5188587307929993,
      "eval_runtime": 1.3722,
      "eval_samples_per_second": 1814.556,
      "eval_steps_per_second": 113.683,
      "step": 30000
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.8946585723598436e-05,
      "loss": 0.4416,
      "step": 30500
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.843729628422425e-05,
      "loss": 0.4289,
      "step": 31000
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.7928006844850064e-05,
      "loss": 0.4315,
      "step": 31500
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.741871740547588e-05,
      "loss": 0.4348,
      "step": 32000
    },
    {
      "epoch": 1.3,
      "eval_accuracy": 0.8008032128514057,
      "eval_loss": 0.5141127705574036,
      "eval_runtime": 1.3695,
      "eval_samples_per_second": 1818.24,
      "eval_steps_per_second": 113.914,
      "step": 32000
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.6909427966101695e-05,
      "loss": 0.4364,
      "step": 32500
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.640115710560626e-05,
      "loss": 0.4398,
      "step": 33000
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.5891867666232073e-05,
      "loss": 0.4492,
      "step": 33500
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.5382578226857887e-05,
      "loss": 0.4339,
      "step": 34000
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.802008032128514,
      "eval_loss": 0.5085433721542358,
      "eval_runtime": 3.0786,
      "eval_samples_per_second": 808.808,
      "eval_steps_per_second": 50.672,
      "step": 34000
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.4873288787483703e-05,
      "loss": 0.4313,
      "step": 34500
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.4363999348109519e-05,
      "loss": 0.4334,
      "step": 35000
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.3854709908735333e-05,
      "loss": 0.4294,
      "step": 35500
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.3345420469361147e-05,
      "loss": 0.4227,
      "step": 36000
    },
    {
      "epoch": 1.47,
      "eval_accuracy": 0.7943775100401607,
      "eval_loss": 0.5210641622543335,
      "eval_runtime": 1.3741,
      "eval_samples_per_second": 1812.132,
      "eval_steps_per_second": 113.531,
      "step": 36000
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.2837149608865712e-05,
      "loss": 0.4224,
      "step": 36500
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.2327860169491526e-05,
      "loss": 0.4398,
      "step": 37000
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.1819589308996089e-05,
      "loss": 0.4264,
      "step": 37500
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.1310299869621905e-05,
      "loss": 0.4379,
      "step": 38000
    },
    {
      "epoch": 1.55,
      "eval_accuracy": 0.802008032128514,
      "eval_loss": 0.5011274814605713,
      "eval_runtime": 3.0414,
      "eval_samples_per_second": 818.69,
      "eval_steps_per_second": 51.291,
      "step": 38000
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.0801010430247719e-05,
      "loss": 0.4328,
      "step": 38500
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.0291720990873534e-05,
      "loss": 0.4166,
      "step": 39000
    },
    {
      "epoch": 1.61,
      "learning_rate": 9.78243155149935e-06,
      "loss": 0.4294,
      "step": 39500
    },
    {
      "epoch": 1.63,
      "learning_rate": 9.273142112125164e-06,
      "loss": 0.4301,
      "step": 40000
    },
    {
      "epoch": 1.63,
      "eval_accuracy": 0.802008032128514,
      "eval_loss": 0.5149931311607361,
      "eval_runtime": 1.375,
      "eval_samples_per_second": 1810.901,
      "eval_steps_per_second": 113.454,
      "step": 40000
    },
    {
      "epoch": 1.65,
      "learning_rate": 8.763852672750978e-06,
      "loss": 0.4199,
      "step": 40500
    },
    {
      "epoch": 1.67,
      "learning_rate": 8.255581812255542e-06,
      "loss": 0.43,
      "step": 41000
    },
    {
      "epoch": 1.69,
      "learning_rate": 7.747310951760104e-06,
      "loss": 0.4209,
      "step": 41500
    },
    {
      "epoch": 1.71,
      "learning_rate": 7.23802151238592e-06,
      "loss": 0.4235,
      "step": 42000
    },
    {
      "epoch": 1.71,
      "eval_accuracy": 0.8028112449799196,
      "eval_loss": 0.51395583152771,
      "eval_runtime": 1.3678,
      "eval_samples_per_second": 1820.447,
      "eval_steps_per_second": 114.052,
      "step": 42000
    },
    {
      "epoch": 1.73,
      "learning_rate": 6.728732073011734e-06,
      "loss": 0.4249,
      "step": 42500
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.219442633637549e-06,
      "loss": 0.4102,
      "step": 43000
    },
    {
      "epoch": 1.77,
      "learning_rate": 5.710153194263364e-06,
      "loss": 0.4177,
      "step": 43500
    },
    {
      "epoch": 1.79,
      "learning_rate": 5.200863754889179e-06,
      "loss": 0.4122,
      "step": 44000
    },
    {
      "epoch": 1.79,
      "eval_accuracy": 0.8068273092369478,
      "eval_loss": 0.5007363557815552,
      "eval_runtime": 1.3698,
      "eval_samples_per_second": 1817.792,
      "eval_steps_per_second": 113.886,
      "step": 44000
    },
    {
      "epoch": 1.81,
      "learning_rate": 4.691574315514994e-06,
      "loss": 0.4256,
      "step": 44500
    },
    {
      "epoch": 1.83,
      "learning_rate": 4.182284876140809e-06,
      "loss": 0.4202,
      "step": 45000
    },
    {
      "epoch": 1.85,
      "learning_rate": 3.6729954367666235e-06,
      "loss": 0.4073,
      "step": 45500
    },
    {
      "epoch": 1.87,
      "learning_rate": 3.1637059973924383e-06,
      "loss": 0.4105,
      "step": 46000
    },
    {
      "epoch": 1.87,
      "eval_accuracy": 0.8100401606425702,
      "eval_loss": 0.5013556480407715,
      "eval_runtime": 1.3597,
      "eval_samples_per_second": 1831.353,
      "eval_steps_per_second": 114.735,
      "step": 46000
    },
    {
      "epoch": 1.89,
      "learning_rate": 2.654416558018253e-06,
      "loss": 0.4171,
      "step": 46500
    },
    {
      "epoch": 1.91,
      "learning_rate": 2.1451271186440677e-06,
      "loss": 0.4062,
      "step": 47000
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.6358376792698828e-06,
      "loss": 0.4137,
      "step": 47500
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.127566818774446e-06,
      "loss": 0.3977,
      "step": 48000
    },
    {
      "epoch": 1.96,
      "eval_accuracy": 0.804417670682731,
      "eval_loss": 0.5082036852836609,
      "eval_runtime": 1.4228,
      "eval_samples_per_second": 1750.022,
      "eval_steps_per_second": 109.64,
      "step": 48000
    },
    {
      "epoch": 1.98,
      "learning_rate": 6.192959582790092e-07,
      "loss": 0.4197,
      "step": 48500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.1000651890482399e-07,
      "loss": 0.4186,
      "step": 49000
    },
    {
      "epoch": 2.0,
      "step": 49088,
      "total_flos": 3200911484154624.0,
      "train_loss": 0.5175694002632678,
      "train_runtime": 2940.9235,
      "train_samples_per_second": 267.06,
      "train_steps_per_second": 16.691
    }
  ],
  "max_steps": 49088,
  "num_train_epochs": 2,
  "total_flos": 3200911484154624.0,
  "trial_name": null,
  "trial_params": null
}