{
  "best_metric": 0.4270777404308319,
  "best_model_checkpoint": "/data/jcanete/all_results/xnli/albeto_xxlarge/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": 0.8888,
      "step": 500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.898549543676663e-06,
      "loss": 0.7298,
      "step": 1000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.847620599739244e-06,
      "loss": 0.6804,
      "step": 1500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.796691655801826e-06,
      "loss": 0.6317,
      "step": 2000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.7558232931726908,
      "eval_loss": 0.598174512386322,
      "eval_runtime": 84.2575,
      "eval_samples_per_second": 29.552,
      "eval_steps_per_second": 3.703,
      "step": 2000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.745762711864408e-06,
      "loss": 0.6141,
      "step": 2500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.694833767926988e-06,
      "loss": 0.5958,
      "step": 3000
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.644006681877445e-06,
      "loss": 0.5947,
      "step": 3500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.593179595827901e-06,
      "loss": 0.5644,
      "step": 4000
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.7831325301204819,
      "eval_loss": 0.5494194626808167,
      "eval_runtime": 84.263,
      "eval_samples_per_second": 29.55,
      "eval_steps_per_second": 3.703,
      "step": 4000
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.5423525097783575e-06,
      "loss": 0.5719,
      "step": 4500
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.491423565840939e-06,
      "loss": 0.5633,
      "step": 5000
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.440494621903521e-06,
      "loss": 0.5673,
      "step": 5500
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.389565677966102e-06,
      "loss": 0.5429,
      "step": 6000
    },
    {
      "epoch": 0.24,
      "eval_accuracy": 0.7959839357429719,
      "eval_loss": 0.527809202671051,
      "eval_runtime": 84.2589,
      "eval_samples_per_second": 29.552,
      "eval_steps_per_second": 3.703,
      "step": 6000
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.338636734028683e-06,
      "loss": 0.5402,
      "step": 6500
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.287707790091265e-06,
      "loss": 0.5457,
      "step": 7000
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.236778846153847e-06,
      "loss": 0.5426,
      "step": 7500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.185849902216428e-06,
      "loss": 0.5279,
      "step": 8000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.7991967871485943,
      "eval_loss": 0.5055843591690063,
      "eval_runtime": 84.2703,
      "eval_samples_per_second": 29.548,
      "eval_steps_per_second": 3.702,
      "step": 8000
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.134920958279009e-06,
      "loss": 0.5204,
      "step": 8500
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.0841957301173405e-06,
      "loss": 0.5139,
      "step": 9000
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.033266786179922e-06,
      "loss": 0.5189,
      "step": 9500
    },
    {
      "epoch": 0.41,
      "learning_rate": 3.982337842242504e-06,
      "loss": 0.5149,
      "step": 10000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.814859437751004,
      "eval_loss": 0.4850701093673706,
      "eval_runtime": 84.292,
      "eval_samples_per_second": 29.54,
      "eval_steps_per_second": 3.701,
      "step": 10000
    },
    {
      "epoch": 0.43,
      "learning_rate": 3.931408898305085e-06,
      "loss": 0.5143,
      "step": 10500
    },
    {
      "epoch": 0.45,
      "learning_rate": 3.880479954367666e-06,
      "loss": 0.5011,
      "step": 11000
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.829551010430248e-06,
      "loss": 0.5051,
      "step": 11500
    },
    {
      "epoch": 0.49,
      "learning_rate": 3.7786220664928295e-06,
      "loss": 0.4991,
      "step": 12000
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.8116465863453816,
      "eval_loss": 0.4848240613937378,
      "eval_runtime": 84.2552,
      "eval_samples_per_second": 29.553,
      "eval_steps_per_second": 3.703,
      "step": 12000
    },
    {
      "epoch": 0.51,
      "learning_rate": 3.7276931225554107e-06,
      "loss": 0.4942,
      "step": 12500
    },
    {
      "epoch": 0.53,
      "learning_rate": 3.676866036505867e-06,
      "loss": 0.515,
      "step": 13000
    },
    {
      "epoch": 0.55,
      "learning_rate": 3.625937092568449e-06,
      "loss": 0.5015,
      "step": 13500
    },
    {
      "epoch": 0.57,
      "learning_rate": 3.5750081486310302e-06,
      "loss": 0.4943,
      "step": 14000
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.8232931726907631,
      "eval_loss": 0.4758189022541046,
      "eval_runtime": 84.249,
      "eval_samples_per_second": 29.555,
      "eval_steps_per_second": 3.703,
      "step": 14000
    },
    {
      "epoch": 0.59,
      "learning_rate": 3.5240792046936118e-06,
      "loss": 0.4901,
      "step": 14500
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.4731502607561934e-06,
      "loss": 0.5031,
      "step": 15000
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.4222213168187745e-06,
      "loss": 0.4825,
      "step": 15500
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.371292372881356e-06,
      "loss": 0.4925,
      "step": 16000
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.8240963855421687,
      "eval_loss": 0.45331594347953796,
      "eval_runtime": 84.2469,
      "eval_samples_per_second": 29.556,
      "eval_steps_per_second": 3.703,
      "step": 16000
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.320363428943938e-06,
      "loss": 0.4896,
      "step": 16500
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.2695363428943936e-06,
      "loss": 0.4854,
      "step": 17000
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.2189129726205997e-06,
      "loss": 0.4973,
      "step": 17500
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.1679840286831813e-06,
      "loss": 0.4695,
      "step": 18000
    },
    {
      "epoch": 0.73,
      "eval_accuracy": 0.821285140562249,
      "eval_loss": 0.4604874551296234,
      "eval_runtime": 84.2777,
      "eval_samples_per_second": 29.545,
      "eval_steps_per_second": 3.702,
      "step": 18000
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.1170550847457633e-06,
      "loss": 0.4713,
      "step": 18500
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.0661261408083444e-06,
      "loss": 0.4788,
      "step": 19000
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.015197196870926e-06,
      "loss": 0.4718,
      "step": 19500
    },
    {
      "epoch": 0.81,
      "learning_rate": 2.9642682529335076e-06,
      "loss": 0.469,
      "step": 20000
    },
    {
      "epoch": 0.81,
      "eval_accuracy": 0.8244979919678714,
      "eval_loss": 0.44962260127067566,
      "eval_runtime": 84.2647,
      "eval_samples_per_second": 29.55,
      "eval_steps_per_second": 3.703,
      "step": 20000
    },
    {
      "epoch": 0.84,
      "learning_rate": 2.9133393089960887e-06,
      "loss": 0.4782,
      "step": 20500
    },
    {
      "epoch": 0.86,
      "learning_rate": 2.8624103650586703e-06,
      "loss": 0.4783,
      "step": 21000
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.8114814211212523e-06,
      "loss": 0.4619,
      "step": 21500
    },
    {
      "epoch": 0.9,
      "learning_rate": 2.7605524771838334e-06,
      "loss": 0.4671,
      "step": 22000
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.8373493975903614,
      "eval_loss": 0.4270777404308319,
      "eval_runtime": 84.2512,
      "eval_samples_per_second": 29.554,
      "eval_steps_per_second": 3.703,
      "step": 22000
    },
    {
      "epoch": 0.92,
      "learning_rate": 2.709623533246415e-06,
      "loss": 0.4665,
      "step": 22500
    },
    {
      "epoch": 0.94,
      "learning_rate": 2.658694589308996e-06,
      "loss": 0.4703,
      "step": 23000
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.6077656453715777e-06,
      "loss": 0.4559,
      "step": 23500
    },
    {
      "epoch": 0.98,
      "learning_rate": 2.5568367014341593e-06,
      "loss": 0.4709,
      "step": 24000
    },
    {
      "epoch": 0.98,
      "eval_accuracy": 0.8337349397590361,
      "eval_loss": 0.436306357383728,
      "eval_runtime": 84.2829,
      "eval_samples_per_second": 29.543,
      "eval_steps_per_second": 3.702,
      "step": 24000
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.5059077574967404e-06,
      "loss": 0.4581,
      "step": 24500
    },
    {
      "epoch": 1.02,
      "learning_rate": 2.454978813559322e-06,
      "loss": 0.3387,
      "step": 25000
    },
    {
      "epoch": 1.04,
      "learning_rate": 2.4040498696219036e-06,
      "loss": 0.3281,
      "step": 25500
    },
    {
      "epoch": 1.06,
      "learning_rate": 2.353120925684485e-06,
      "loss": 0.3167,
      "step": 26000
    },
    {
      "epoch": 1.06,
      "eval_accuracy": 0.8289156626506025,
      "eval_loss": 0.47485458850860596,
      "eval_runtime": 84.2522,
      "eval_samples_per_second": 29.554,
      "eval_steps_per_second": 3.703,
      "step": 26000
    },
    {
      "epoch": 1.08,
      "learning_rate": 2.3021919817470667e-06,
      "loss": 0.3267,
      "step": 26500
    },
    {
      "epoch": 1.1,
      "learning_rate": 2.2512630378096483e-06,
      "loss": 0.3058,
      "step": 27000
    },
    {
      "epoch": 1.12,
      "learning_rate": 2.2003340938722295e-06,
      "loss": 0.3249,
      "step": 27500
    },
    {
      "epoch": 1.14,
      "learning_rate": 2.149405149934811e-06,
      "loss": 0.3214,
      "step": 28000
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.8325301204819278,
      "eval_loss": 0.47795429825782776,
      "eval_runtime": 84.2738,
      "eval_samples_per_second": 29.547,
      "eval_steps_per_second": 3.702,
      "step": 28000
    },
    {
      "epoch": 1.16,
      "learning_rate": 2.0985780638852674e-06,
      "loss": 0.3289,
      "step": 28500
    },
    {
      "epoch": 1.18,
      "learning_rate": 2.047649119947849e-06,
      "loss": 0.3176,
      "step": 29000
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.9967201760104306e-06,
      "loss": 0.3292,
      "step": 29500
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.9457912320730117e-06,
      "loss": 0.326,
      "step": 30000
    },
    {
      "epoch": 1.22,
      "eval_accuracy": 0.8309236947791164,
      "eval_loss": 0.46767711639404297,
      "eval_runtime": 84.2707,
      "eval_samples_per_second": 29.548,
      "eval_steps_per_second": 3.702,
      "step": 30000
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.8948622881355933e-06,
      "loss": 0.3164,
      "step": 30500
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.8439333441981749e-06,
      "loss": 0.3151,
      "step": 31000
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.7930044002607562e-06,
      "loss": 0.3196,
      "step": 31500
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.7420754563233378e-06,
      "loss": 0.3192,
      "step": 32000
    },
    {
      "epoch": 1.3,
      "eval_accuracy": 0.8353413654618473,
      "eval_loss": 0.4579215347766876,
      "eval_runtime": 84.2732,
      "eval_samples_per_second": 29.547,
      "eval_steps_per_second": 3.702,
      "step": 32000
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.6911465123859194e-06,
      "loss": 0.3148,
      "step": 32500
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.6402175684485007e-06,
      "loss": 0.3126,
      "step": 33000
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.5892886245110823e-06,
      "loss": 0.3106,
      "step": 33500
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.5384615384615387e-06,
      "loss": 0.3239,
      "step": 34000
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.8333333333333334,
      "eval_loss": 0.4818671643733978,
      "eval_runtime": 84.2403,
      "eval_samples_per_second": 29.558,
      "eval_steps_per_second": 3.704,
      "step": 34000
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.48753259452412e-06,
      "loss": 0.3131,
      "step": 34500
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.4366036505867014e-06,
      "loss": 0.3057,
      "step": 35000
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.3856747066492832e-06,
      "loss": 0.3032,
      "step": 35500
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.3347457627118646e-06,
      "loss": 0.3026,
      "step": 36000
    },
    {
      "epoch": 1.47,
      "eval_accuracy": 0.8389558232931726,
      "eval_loss": 0.47428080439567566,
      "eval_runtime": 84.2812,
      "eval_samples_per_second": 29.544,
      "eval_steps_per_second": 3.702,
      "step": 36000
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.2839186766623208e-06,
      "loss": 0.3007,
      "step": 36500
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.2329897327249023e-06,
      "loss": 0.3148,
      "step": 37000
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.1820607887874837e-06,
      "loss": 0.299,
      "step": 37500
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.1311318448500653e-06,
      "loss": 0.3176,
      "step": 38000
    },
    {
      "epoch": 1.55,
      "eval_accuracy": 0.8437751004016064,
      "eval_loss": 0.46083542704582214,
      "eval_runtime": 84.2921,
      "eval_samples_per_second": 29.54,
      "eval_steps_per_second": 3.701,
      "step": 38000
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.0803047588005217e-06,
      "loss": 0.3099,
      "step": 38500
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.029375814863103e-06,
      "loss": 0.2976,
      "step": 39000
    },
    {
      "epoch": 1.61,
      "learning_rate": 9.785487288135595e-07,
      "loss": 0.2978,
      "step": 39500
    },
    {
      "epoch": 1.63,
      "learning_rate": 9.276197848761409e-07,
      "loss": 0.309,
      "step": 40000
    },
    {
      "epoch": 1.63,
      "eval_accuracy": 0.8409638554216867,
      "eval_loss": 0.46184200048446655,
      "eval_runtime": 84.2612,
      "eval_samples_per_second": 29.551,
      "eval_steps_per_second": 3.703,
      "step": 40000
    },
    {
      "epoch": 1.65,
      "learning_rate": 8.766908409387223e-07,
      "loss": 0.2989,
      "step": 40500
    },
    {
      "epoch": 1.67,
      "learning_rate": 8.257618970013039e-07,
      "loss": 0.309,
      "step": 41000
    },
    {
      "epoch": 1.69,
      "learning_rate": 7.749348109517601e-07,
      "loss": 0.3016,
      "step": 41500
    },
    {
      "epoch": 1.71,
      "learning_rate": 7.240058670143417e-07,
      "loss": 0.2999,
      "step": 42000
    },
    {
      "epoch": 1.71,
      "eval_accuracy": 0.838152610441767,
      "eval_loss": 0.4617726802825928,
      "eval_runtime": 84.2688,
      "eval_samples_per_second": 29.548,
      "eval_steps_per_second": 3.702,
      "step": 42000
    },
    {
      "epoch": 1.73,
      "learning_rate": 6.730769230769231e-07,
      "loss": 0.3081,
      "step": 42500
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.221479791395047e-07,
      "loss": 0.3019,
      "step": 43000
    },
    {
      "epoch": 1.77,
      "learning_rate": 5.712190352020861e-07,
      "loss": 0.3049,
      "step": 43500
    },
    {
      "epoch": 1.79,
      "learning_rate": 5.202900912646676e-07,
      "loss": 0.2941,
      "step": 44000
    },
    {
      "epoch": 1.79,
      "eval_accuracy": 0.8449799196787149,
      "eval_loss": 0.4609120190143585,
      "eval_runtime": 84.2453,
      "eval_samples_per_second": 29.557,
      "eval_steps_per_second": 3.703,
      "step": 44000
    },
    {
      "epoch": 1.81,
      "learning_rate": 4.6936114732724906e-07,
      "loss": 0.3002,
      "step": 44500
    },
    {
      "epoch": 1.83,
      "learning_rate": 4.184322033898305e-07,
      "loss": 0.3008,
      "step": 45000
    },
    {
      "epoch": 1.85,
      "learning_rate": 3.676051173402868e-07,
      "loss": 0.296,
      "step": 45500
    },
    {
      "epoch": 1.87,
      "learning_rate": 3.166761734028683e-07,
      "loss": 0.2946,
      "step": 46000
    },
    {
      "epoch": 1.87,
      "eval_accuracy": 0.8461847389558232,
      "eval_loss": 0.460882306098938,
      "eval_runtime": 84.2508,
      "eval_samples_per_second": 29.555,
      "eval_steps_per_second": 3.703,
      "step": 46000
    },
    {
      "epoch": 1.89,
      "learning_rate": 2.657472294654498e-07,
      "loss": 0.2888,
      "step": 46500
    },
    {
      "epoch": 1.91,
      "learning_rate": 2.1492014341590616e-07,
      "loss": 0.2907,
      "step": 47000
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.6399119947848762e-07,
      "loss": 0.3067,
      "step": 47500
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.130622555410691e-07,
      "loss": 0.2757,
      "step": 48000
    },
    {
      "epoch": 1.96,
      "eval_accuracy": 0.8437751004016064,
      "eval_loss": 0.4675902724266052,
      "eval_runtime": 84.2793,
      "eval_samples_per_second": 29.545,
      "eval_steps_per_second": 3.702,
      "step": 48000
    },
    {
      "epoch": 1.98,
      "learning_rate": 6.21333116036506e-08,
      "loss": 0.3026,
      "step": 48500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.1204367666232074e-08,
      "loss": 0.309,
      "step": 49000
    },
    {
      "epoch": 2.0,
      "step": 49088,
      "total_flos": 7.899132922600589e+16,
      "train_loss": 0.4176930578855368,
      "train_runtime": 61258.2129,
      "train_samples_per_second": 12.821,
      "train_steps_per_second": 0.801
    }
  ],
  "max_steps": 49088,
  "num_train_epochs": 2,
  "total_flos": 7.899132922600589e+16,
  "trial_name": null,
  "trial_params": null
}