trainer_state.json

{
  "best_global_step": 1218,
  "best_metric": 0.5196570893167077,
  "best_model_checkpoint": "outputs/mbert/checkpoint-1218",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 1218,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9942528735632183,
      "grad_norm": 5.0743865966796875,
      "learning_rate": 6.88e-06,
      "loss": 1.822,
      "step": 173
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3877551020408163,
      "eval_loss": 1.6465110778808594,
      "eval_macro_f1": 0.18429186983968865,
      "eval_runtime": 2.8526,
      "eval_samples_per_second": 240.484,
      "eval_steps_per_second": 7.712,
      "step": 174
    },
    {
      "epoch": 1.9885057471264367,
      "grad_norm": 10.819275856018066,
      "learning_rate": 1.38e-05,
      "loss": 1.5516,
      "step": 346
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3965014577259475,
      "eval_loss": 1.5053271055221558,
      "eval_macro_f1": 0.23687479202866749,
      "eval_runtime": 2.8479,
      "eval_samples_per_second": 240.882,
      "eval_steps_per_second": 7.725,
      "step": 348
    },
    {
      "epoch": 2.9827586206896552,
      "grad_norm": 16.215076446533203,
      "learning_rate": 1.999994401896218e-05,
      "loss": 1.2991,
      "step": 519
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5276967930029155,
      "eval_loss": 1.306789517402649,
      "eval_macro_f1": 0.3774308237979942,
      "eval_runtime": 2.8463,
      "eval_samples_per_second": 241.018,
      "eval_steps_per_second": 7.729,
      "step": 522
    },
    {
      "epoch": 3.9770114942528734,
      "grad_norm": 12.098478317260742,
      "learning_rate": 1.9993697433304146e-05,
      "loss": 1.0762,
      "step": 692
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5116618075801749,
      "eval_loss": 1.2950795888900757,
      "eval_macro_f1": 0.4219223955702657,
      "eval_runtime": 2.8679,
      "eval_samples_per_second": 239.197,
      "eval_steps_per_second": 7.671,
      "step": 696
    },
    {
      "epoch": 4.971264367816092,
      "grad_norm": 20.27077293395996,
      "learning_rate": 1.9977115923137912e-05,
      "loss": 0.8392,
      "step": 865
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5364431486880467,
      "eval_loss": 1.3656411170959473,
      "eval_macro_f1": 0.4472023218969972,
      "eval_runtime": 2.8594,
      "eval_samples_per_second": 239.907,
      "eval_steps_per_second": 7.694,
      "step": 870
    },
    {
      "epoch": 5.9655172413793105,
      "grad_norm": 12.2693452835083,
      "learning_rate": 1.995021663613649e-05,
      "loss": 0.6457,
      "step": 1038
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.543731778425656,
      "eval_loss": 1.3349292278289795,
      "eval_macro_f1": 0.4885938989698125,
      "eval_runtime": 2.8587,
      "eval_samples_per_second": 239.968,
      "eval_steps_per_second": 7.696,
      "step": 1044
    },
    {
      "epoch": 6.959770114942529,
      "grad_norm": 17.85072135925293,
      "learning_rate": 1.991302739004225e-05,
      "loss": 0.4684,
      "step": 1211
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5626822157434402,
      "eval_loss": 1.4547728300094604,
      "eval_macro_f1": 0.5196570893167077,
      "eval_runtime": 2.8647,
      "eval_samples_per_second": 239.471,
      "eval_steps_per_second": 7.68,
      "step": 1218
    }
  ],
  "logging_steps": 173,
  "max_steps": 17400,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4071489243746400.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}