{
  "best_metric": 0.8132780082987552,
  "best_model_checkpoint": "distilbert-base-multilingual-cased-hyper-matt/run-gz99jeit/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": 2.258552074432373,
      "learning_rate": 8.238402226435047e-05,
      "loss": 0.592,
      "step": 10
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.1681671142578125,
      "learning_rate": 7.64994492454683e-05,
      "loss": 0.3711,
      "step": 20
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.2577924728393555,
      "learning_rate": 7.061487622658612e-05,
      "loss": 0.3785,
      "step": 30
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.4071826934814453,
      "learning_rate": 6.473030320770394e-05,
      "loss": 0.3976,
      "step": 40
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.397700786590576,
      "learning_rate": 5.8845730188821764e-05,
      "loss": 0.3152,
      "step": 50
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8875,
      "eval_f1": 0.8132780082987552,
      "eval_loss": 0.2851196527481079,
      "eval_precision": 0.8305084745762712,
      "eval_recall": 0.7967479674796748,
      "eval_runtime": 1.5239,
      "eval_samples_per_second": 262.49,
      "eval_steps_per_second": 16.406,
      "step": 50
    }
  ],
  "logging_steps": 10,
  "max_steps": 150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 8.826859528323265e-05,
    "metric": "eval/loss",
    "num_train_epochs": 3,
    "per_device_train_batch_size": 32,
    "seed": 26
  }
}