trainer_state.json

{
  "best_metric": 0.3162662386894226,
  "best_model_checkpoint": "glot500-all-full-finetuned-toxicity-classification/checkpoint-4016",
  "epoch": 2.9987546699875467,
  "eval_steps": 500,
  "global_step": 6021,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.24906600249066002,
      "grad_norm": 9.292550086975098,
      "learning_rate": 1.83391463212091e-05,
      "loss": 0.6475,
      "step": 500
    },
    {
      "epoch": 0.49813200498132004,
      "grad_norm": 9.401397705078125,
      "learning_rate": 1.6678292642418204e-05,
      "loss": 0.4811,
      "step": 1000
    },
    {
      "epoch": 0.7471980074719801,
      "grad_norm": 9.068319320678711,
      "learning_rate": 1.5017438963627307e-05,
      "loss": 0.4014,
      "step": 1500
    },
    {
      "epoch": 0.9962640099626401,
      "grad_norm": 8.684527397155762,
      "learning_rate": 1.3356585284836408e-05,
      "loss": 0.3774,
      "step": 2000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.3449280261993408,
      "eval_runtime": 46.4179,
      "eval_samples_per_second": 153.798,
      "eval_steps_per_second": 9.63,
      "step": 2008
    },
    {
      "epoch": 1.2450809464508095,
      "grad_norm": 9.13967514038086,
      "learning_rate": 1.1695731606045508e-05,
      "loss": 0.3165,
      "step": 2500
    },
    {
      "epoch": 1.4941469489414696,
      "grad_norm": 5.335002422332764,
      "learning_rate": 1.0034877927254609e-05,
      "loss": 0.2968,
      "step": 3000
    },
    {
      "epoch": 1.7432129514321295,
      "grad_norm": 11.03541088104248,
      "learning_rate": 8.374024248463712e-06,
      "loss": 0.2965,
      "step": 3500
    },
    {
      "epoch": 1.9922789539227894,
      "grad_norm": 8.862164497375488,
      "learning_rate": 6.7131705696728125e-06,
      "loss": 0.283,
      "step": 4000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.3162662386894226,
      "eval_runtime": 46.3896,
      "eval_samples_per_second": 153.892,
      "eval_steps_per_second": 9.636,
      "step": 4016
    },
    {
      "epoch": 2.241095890410959,
      "grad_norm": 19.95586585998535,
      "learning_rate": 5.0523168908819146e-06,
      "loss": 0.2286,
      "step": 4500
    },
    {
      "epoch": 2.490161892901619,
      "grad_norm": 7.364178657531738,
      "learning_rate": 3.391463212091015e-06,
      "loss": 0.2284,
      "step": 5000
    },
    {
      "epoch": 2.739227895392279,
      "grad_norm": 9.687633514404297,
      "learning_rate": 1.7306095333001162e-06,
      "loss": 0.2174,
      "step": 5500
    },
    {
      "epoch": 2.988293897882939,
      "grad_norm": 7.806052207946777,
      "learning_rate": 6.975585450921775e-08,
      "loss": 0.2228,
      "step": 6000
    },
    {
      "epoch": 2.9987546699875467,
      "eval_loss": 0.3281228244304657,
      "eval_runtime": 46.4096,
      "eval_samples_per_second": 153.826,
      "eval_steps_per_second": 9.632,
      "step": 6021
    }
  ],
  "logging_steps": 500,
  "max_steps": 6021,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.06830825939968e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}