{
  "best_metric": 0.7953667953667953,
  "best_model_checkpoint": "distilbert-base-multilingual-cased-hyper-matt/run-r7r8youc/checkpoint-50",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 50,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "grad_norm": 1.2081632614135742,
      "learning_rate": 6.699943884317094e-05,
      "loss": 0.5387,
      "step": 10
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.086586952209473,
      "learning_rate": 6.347315258826721e-05,
      "loss": 0.4384,
      "step": 20
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.491854190826416,
      "learning_rate": 5.994686633336347e-05,
      "loss": 0.4018,
      "step": 30
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.669224739074707,
      "learning_rate": 5.6420580078459746e-05,
      "loss": 0.3826,
      "step": 40
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.3707859516143799,
      "learning_rate": 5.2894293823556006e-05,
      "loss": 0.4844,
      "step": 50
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8675,
      "eval_f1": 0.7953667953667953,
      "eval_loss": 0.33214980363845825,
      "eval_precision": 0.7573529411764706,
      "eval_recall": 0.8373983739837398,
      "eval_runtime": 1.5208,
      "eval_samples_per_second": 263.015,
      "eval_steps_per_second": 16.438,
      "step": 50
    }
  ],
  "logging_steps": 10,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 211815370450944.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "_wandb": {},
    "assignments": {},
    "learning_rate": 7.052572509807468e-05,
    "metric": "eval/loss",
    "num_train_epochs": 4,
    "per_device_train_batch_size": 32,
    "seed": 29
  }
}