run-vrziupab/checkpoint-100/trainer_state.json

{
  "best_metric": 0.8130081300813008,
  "best_model_checkpoint": "distilbert-base-multilingual-cased-hyper-matt/run-vrziupab/checkpoint-100",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4,
      "grad_norm": 1.6790717840194702,
      "learning_rate": 3.3566715164263906e-05,
      "loss": 0.5543,
      "step": 10
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.9814144372940063,
      "learning_rate": 2.983708014601236e-05,
      "loss": 0.4322,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.85,
      "eval_f1": 0.7744360902255639,
      "eval_loss": 0.3429238200187683,
      "eval_precision": 0.7202797202797203,
      "eval_recall": 0.8373983739837398,
      "eval_runtime": 1.5447,
      "eval_samples_per_second": 258.95,
      "eval_steps_per_second": 16.184,
      "step": 25
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.034728050231934,
      "learning_rate": 2.6107445127760812e-05,
      "loss": 0.3387,
      "step": 30
    },
    {
      "epoch": 1.6,
      "grad_norm": 5.807296276092529,
      "learning_rate": 2.237781010950927e-05,
      "loss": 0.3155,
      "step": 40
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.17315673828125,
      "learning_rate": 1.8648175091257724e-05,
      "loss": 0.2955,
      "step": 50
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.865,
      "eval_f1": 0.7969924812030075,
      "eval_loss": 0.3180018663406372,
      "eval_precision": 0.7412587412587412,
      "eval_recall": 0.8617886178861789,
      "eval_runtime": 1.515,
      "eval_samples_per_second": 264.023,
      "eval_steps_per_second": 16.501,
      "step": 50
    },
    {
      "epoch": 2.4,
      "grad_norm": 3.6483638286590576,
      "learning_rate": 1.491854007300618e-05,
      "loss": 0.2305,
      "step": 60
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.6533458232879639,
      "learning_rate": 1.1188905054754635e-05,
      "loss": 0.2633,
      "step": 70
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.88,
      "eval_f1": 0.8048780487804879,
      "eval_loss": 0.29483410716056824,
      "eval_precision": 0.8048780487804879,
      "eval_recall": 0.8048780487804879,
      "eval_runtime": 1.5239,
      "eval_samples_per_second": 262.477,
      "eval_steps_per_second": 16.405,
      "step": 75
    },
    {
      "epoch": 3.2,
      "grad_norm": 3.5790436267852783,
      "learning_rate": 7.45927003650309e-06,
      "loss": 0.229,
      "step": 80
    },
    {
      "epoch": 3.6,
      "grad_norm": 3.870107889175415,
      "learning_rate": 3.729635018251545e-06,
      "loss": 0.1866,
      "step": 90
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.5545942783355713,
      "learning_rate": 0.0,
      "loss": 0.1838,
      "step": 100
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.885,
      "eval_f1": 0.8130081300813008,
      "eval_loss": 0.29451921582221985,
      "eval_precision": 0.8130081300813008,
      "eval_recall": 0.8130081300813008,
      "eval_runtime": 1.5767,
      "eval_samples_per_second": 253.702,
      "eval_steps_per_second": 15.856,
      "step": 100
    }
  ],
  "logging_steps": 10,
  "max_steps": 100,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 847261481803776.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": {
    "_wandb": {},
    "assignments": {},
    "learning_rate": 3.729635018251545e-05,
    "metric": "eval/loss",
    "num_train_epochs": 4,
    "per_device_train_batch_size": 64,
    "seed": 27
  }
}