{
  "best_metric": 0.7947742682321812,
  "best_model_checkpoint": "./XLM-V_64-multi-outputs/checkpoint-15000",
  "epoch": 14.8619957537155,
  "eval_steps": 1000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.7077140835102619,
      "grad_norm": 1.493615746498108,
      "learning_rate": 9.433962264150943e-07,
      "loss": 0.6936,
      "step": 1000
    },
    {
      "epoch": 0.7077140835102619,
      "eval_accuracy": 0.5659663447177138,
      "eval_f1": 0.6995657867530498,
      "eval_loss": 0.682445764541626,
      "eval_precision": 0.5400659785037778,
      "eval_recall": 0.9927621283255086,
      "eval_runtime": 57.8197,
      "eval_samples_per_second": 173.695,
      "eval_steps_per_second": 2.715,
      "step": 1000
    },
    {
      "epoch": 1.4154281670205238,
      "grad_norm": 3.4982151985168457,
      "learning_rate": 1.8867924528301887e-06,
      "loss": 0.672,
      "step": 2000
    },
    {
      "epoch": 1.4154281670205238,
      "eval_accuracy": 0.6204321417903017,
      "eval_f1": 0.7176714560805807,
      "eval_loss": 0.6340225338935852,
      "eval_precision": 0.5774731823599524,
      "eval_recall": 0.9477699530516432,
      "eval_runtime": 57.5505,
      "eval_samples_per_second": 174.508,
      "eval_steps_per_second": 2.728,
      "step": 2000
    },
    {
      "epoch": 2.1231422505307855,
      "grad_norm": 12.846685409545898,
      "learning_rate": 1.907732634338139e-06,
      "loss": 0.6537,
      "step": 3000
    },
    {
      "epoch": 2.1231422505307855,
      "eval_accuracy": 0.6434332370805537,
      "eval_f1": 0.7224676431837557,
      "eval_loss": 0.6225757002830505,
      "eval_precision": 0.5982543960980619,
      "eval_recall": 0.9117762128325508,
      "eval_runtime": 57.5731,
      "eval_samples_per_second": 174.439,
      "eval_steps_per_second": 2.727,
      "step": 3000
    },
    {
      "epoch": 2.8308563340410475,
      "grad_norm": 4.933294296264648,
      "learning_rate": 1.8028833551769331e-06,
      "loss": 0.6499,
      "step": 4000
    },
    {
      "epoch": 2.8308563340410475,
      "eval_accuracy": 0.6594643034949716,
      "eval_f1": 0.743858597962852,
      "eval_loss": 0.604403555393219,
      "eval_precision": 0.6026699029126213,
      "eval_recall": 0.9714397496087637,
      "eval_runtime": 57.5924,
      "eval_samples_per_second": 174.381,
      "eval_steps_per_second": 2.726,
      "step": 4000
    },
    {
      "epoch": 3.538570417551309,
      "grad_norm": 9.257452011108398,
      "learning_rate": 1.6980340760157273e-06,
      "loss": 0.6389,
      "step": 5000
    },
    {
      "epoch": 3.538570417551309,
      "eval_accuracy": 0.660460021905805,
      "eval_f1": 0.6269146608315098,
      "eval_loss": 0.6047727465629578,
      "eval_precision": 0.7112711022840119,
      "eval_recall": 0.5604460093896714,
      "eval_runtime": 57.6043,
      "eval_samples_per_second": 174.344,
      "eval_steps_per_second": 2.725,
      "step": 5000
    },
    {
      "epoch": 4.246284501061571,
      "grad_norm": 8.192702293395996,
      "learning_rate": 1.5931847968545215e-06,
      "loss": 0.6187,
      "step": 6000
    },
    {
      "epoch": 4.246284501061571,
      "eval_accuracy": 0.6978990341531415,
      "eval_f1": 0.7157046476761619,
      "eval_loss": 0.5651090741157532,
      "eval_precision": 0.6868705035971223,
      "eval_recall": 0.7470657276995305,
      "eval_runtime": 57.5852,
      "eval_samples_per_second": 174.402,
      "eval_steps_per_second": 2.726,
      "step": 6000
    },
    {
      "epoch": 4.953998584571833,
      "grad_norm": 49.664493560791016,
      "learning_rate": 1.488335517693316e-06,
      "loss": 0.5962,
      "step": 7000
    },
    {
      "epoch": 4.953998584571833,
      "eval_accuracy": 0.7021806233197252,
      "eval_f1": 0.7534009398961168,
      "eval_loss": 0.5552804470062256,
      "eval_precision": 0.6511329628046174,
      "eval_recall": 0.8937793427230047,
      "eval_runtime": 57.4997,
      "eval_samples_per_second": 174.662,
      "eval_steps_per_second": 2.73,
      "step": 7000
    },
    {
      "epoch": 5.661712668082095,
      "grad_norm": 7.591745376586914,
      "learning_rate": 1.38348623853211e-06,
      "loss": 0.5716,
      "step": 8000
    },
    {
      "epoch": 5.661712668082095,
      "eval_accuracy": 0.7096485114009758,
      "eval_f1": 0.7601184600197434,
      "eval_loss": 0.5378949046134949,
      "eval_precision": 0.65587734241908,
      "eval_recall": 0.903755868544601,
      "eval_runtime": 57.5168,
      "eval_samples_per_second": 174.61,
      "eval_steps_per_second": 2.73,
      "step": 8000
    },
    {
      "epoch": 6.369426751592357,
      "grad_norm": 7.424753189086914,
      "learning_rate": 1.2786369593709043e-06,
      "loss": 0.531,
      "step": 9000
    },
    {
      "epoch": 6.369426751592357,
      "eval_accuracy": 0.7241860001991437,
      "eval_f1": 0.746800731261426,
      "eval_loss": 0.4708074629306793,
      "eval_precision": 0.7009265614275909,
      "eval_recall": 0.7991001564945227,
      "eval_runtime": 57.6158,
      "eval_samples_per_second": 174.31,
      "eval_steps_per_second": 2.725,
      "step": 9000
    },
    {
      "epoch": 7.077140835102618,
      "grad_norm": 9.683096885681152,
      "learning_rate": 1.1737876802096983e-06,
      "loss": 0.4858,
      "step": 10000
    },
    {
      "epoch": 7.077140835102618,
      "eval_accuracy": 0.7385243453151449,
      "eval_f1": 0.7664947536902009,
      "eval_loss": 0.44091591238975525,
      "eval_precision": 0.7026410172807304,
      "eval_recall": 0.843114241001565,
      "eval_runtime": 57.5121,
      "eval_samples_per_second": 174.624,
      "eval_steps_per_second": 2.73,
      "step": 10000
    },
    {
      "epoch": 7.78485491861288,
      "grad_norm": 25.252531051635742,
      "learning_rate": 1.0689384010484928e-06,
      "loss": 0.4577,
      "step": 11000
    },
    {
      "epoch": 7.78485491861288,
      "eval_accuracy": 0.7428059344817286,
      "eval_f1": 0.7718399434678915,
      "eval_loss": 0.42118868231773376,
      "eval_precision": 0.7036559832501208,
      "eval_recall": 0.8546557120500783,
      "eval_runtime": 57.4963,
      "eval_samples_per_second": 174.672,
      "eval_steps_per_second": 2.731,
      "step": 11000
    },
    {
      "epoch": 8.492569002123142,
      "grad_norm": 9.067980766296387,
      "learning_rate": 9.64089121887287e-07,
      "loss": 0.4404,
      "step": 12000
    },
    {
      "epoch": 8.492569002123142,
      "eval_accuracy": 0.7539579806830629,
      "eval_f1": 0.7588091752074183,
      "eval_loss": 0.40373048186302185,
      "eval_precision": 0.75725696473797,
      "eval_recall": 0.7603677621283255,
      "eval_runtime": 57.7852,
      "eval_samples_per_second": 173.799,
      "eval_steps_per_second": 2.717,
      "step": 12000
    },
    {
      "epoch": 9.200283085633403,
      "grad_norm": 6.412005424499512,
      "learning_rate": 8.592398427260812e-07,
      "loss": 0.42,
      "step": 13000
    },
    {
      "epoch": 9.200283085633403,
      "eval_accuracy": 0.7534601214776461,
      "eval_f1": 0.7862938028655274,
      "eval_loss": 0.394222617149353,
      "eval_precision": 0.7035835650293482,
      "eval_recall": 0.8910406885758998,
      "eval_runtime": 57.6787,
      "eval_samples_per_second": 174.12,
      "eval_steps_per_second": 2.722,
      "step": 13000
    },
    {
      "epoch": 9.907997169143666,
      "grad_norm": 9.248723983764648,
      "learning_rate": 7.543905635648754e-07,
      "loss": 0.4061,
      "step": 14000
    },
    {
      "epoch": 9.907997169143666,
      "eval_accuracy": 0.7524644030668127,
      "eval_f1": 0.7793752218672346,
      "eval_loss": 0.3883645236492157,
      "eval_precision": 0.7132878492527616,
      "eval_recall": 0.8589593114241002,
      "eval_runtime": 57.5171,
      "eval_samples_per_second": 174.609,
      "eval_steps_per_second": 2.73,
      "step": 14000
    },
    {
      "epoch": 10.615711252653927,
      "grad_norm": 8.524171829223633,
      "learning_rate": 6.495412844036698e-07,
      "loss": 0.4007,
      "step": 15000
    },
    {
      "epoch": 10.615711252653927,
      "eval_accuracy": 0.7528626904311461,
      "eval_f1": 0.7947742682321812,
      "eval_loss": 0.38538095355033875,
      "eval_precision": 0.6883414494414208,
      "eval_recall": 0.9401408450704225,
      "eval_runtime": 57.8102,
      "eval_samples_per_second": 173.724,
      "eval_steps_per_second": 2.716,
      "step": 15000
    },
    {
      "epoch": 11.32342533616419,
      "grad_norm": 4.774406909942627,
      "learning_rate": 5.44692005242464e-07,
      "loss": 0.3936,
      "step": 16000
    },
    {
      "epoch": 11.32342533616419,
      "eval_accuracy": 0.7545554117295629,
      "eval_f1": 0.793083186434987,
      "eval_loss": 0.38386669754981995,
      "eval_precision": 0.6946037347448905,
      "eval_recall": 0.9241001564945227,
      "eval_runtime": 57.8533,
      "eval_samples_per_second": 173.594,
      "eval_steps_per_second": 2.714,
      "step": 16000
    },
    {
      "epoch": 12.031139419674451,
      "grad_norm": 4.318003177642822,
      "learning_rate": 4.398427260812582e-07,
      "loss": 0.3915,
      "step": 17000
    },
    {
      "epoch": 12.031139419674451,
      "eval_accuracy": 0.7562481330279797,
      "eval_f1": 0.7931034482758621,
      "eval_loss": 0.3860309422016144,
      "eval_precision": 0.6982142857142857,
      "eval_recall": 0.9178403755868545,
      "eval_runtime": 57.8089,
      "eval_samples_per_second": 173.728,
      "eval_steps_per_second": 2.716,
      "step": 17000
    },
    {
      "epoch": 12.738853503184714,
      "grad_norm": 15.07564926147461,
      "learning_rate": 3.3499344692005245e-07,
      "loss": 0.3888,
      "step": 18000
    },
    {
      "epoch": 12.738853503184714,
      "eval_accuracy": 0.7568455640744798,
      "eval_f1": 0.7816523605150214,
      "eval_loss": 0.3812848627567291,
      "eval_precision": 0.7198616600790514,
      "eval_recall": 0.8550469483568075,
      "eval_runtime": 57.8126,
      "eval_samples_per_second": 173.716,
      "eval_steps_per_second": 2.716,
      "step": 18000
    },
    {
      "epoch": 13.446567586694975,
      "grad_norm": 10.069337844848633,
      "learning_rate": 2.3014416775884665e-07,
      "loss": 0.3832,
      "step": 19000
    },
    {
      "epoch": 13.446567586694975,
      "eval_accuracy": 0.7569451359155631,
      "eval_f1": 0.7900576244947106,
      "eval_loss": 0.37964996695518494,
      "eval_precision": 0.7049884881043745,
      "eval_recall": 0.8984741784037559,
      "eval_runtime": 57.6196,
      "eval_samples_per_second": 174.298,
      "eval_steps_per_second": 2.725,
      "step": 19000
    },
    {
      "epoch": 14.154281670205236,
      "grad_norm": 8.055023193359375,
      "learning_rate": 1.252948885976409e-07,
      "loss": 0.383,
      "step": 20000
    },
    {
      "epoch": 14.154281670205236,
      "eval_accuracy": 0.7573434232798965,
      "eval_f1": 0.7904016513288037,
      "eval_loss": 0.3802996873855591,
      "eval_precision": 0.7052954719877207,
      "eval_recall": 0.8988654147104851,
      "eval_runtime": 57.777,
      "eval_samples_per_second": 173.824,
      "eval_steps_per_second": 2.717,
      "step": 20000
    },
    {
      "epoch": 14.8619957537155,
      "grad_norm": 2.661689281463623,
      "learning_rate": 2.0445609436435123e-08,
      "loss": 0.3823,
      "step": 21000
    },
    {
      "epoch": 14.8619957537155,
      "eval_accuracy": 0.7576421388031465,
      "eval_f1": 0.7906056434962148,
      "eval_loss": 0.37897399067878723,
      "eval_precision": 0.7056203931203932,
      "eval_recall": 0.8988654147104851,
      "eval_runtime": 57.6891,
      "eval_samples_per_second": 174.088,
      "eval_steps_per_second": 2.721,
      "step": 21000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 21195,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.173409489802339e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}