run-47/checkpoint-384/trainer_state.json

{
  "best_metric": 0.824853228962818,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-47/checkpoint-384",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 384,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.9338581562042236,
      "learning_rate": 0.0001802406049279757,
      "loss": 0.5354,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7671232876712328,
      "eval_f1": 0.8023255813953489,
      "eval_loss": 0.4617577791213989,
      "eval_precision": 0.696969696969697,
      "eval_recall": 0.9452054794520548,
      "eval_runtime": 29.4455,
      "eval_samples_per_second": 34.708,
      "eval_steps_per_second": 1.087,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.99806547164917,
      "learning_rate": 0.00015771052931197875,
      "loss": 0.4503,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8043052837573386,
      "eval_f1": 0.8242530755711774,
      "eval_loss": 0.42502960562705994,
      "eval_precision": 0.748006379585327,
      "eval_recall": 0.9178082191780822,
      "eval_runtime": 28.5574,
      "eval_samples_per_second": 35.788,
      "eval_steps_per_second": 1.121,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.956643581390381,
      "learning_rate": 0.00013518045369598176,
      "loss": 0.4243,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8238747553816047,
      "eval_f1": 0.8351648351648352,
      "eval_loss": 0.4111813008785248,
      "eval_precision": 0.7848537005163512,
      "eval_recall": 0.8923679060665362,
      "eval_runtime": 27.7197,
      "eval_samples_per_second": 36.869,
      "eval_steps_per_second": 1.154,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.8551297187805176,
      "learning_rate": 0.00011265037807998481,
      "loss": 0.409,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.824853228962818,
      "eval_f1": 0.8306527909176915,
      "eval_loss": 0.40299269556999207,
      "eval_precision": 0.8040293040293041,
      "eval_recall": 0.8590998043052838,
      "eval_runtime": 27.9157,
      "eval_samples_per_second": 36.61,
      "eval_steps_per_second": 1.146,
      "step": 384
    }
  ],
  "logging_steps": 500,
  "max_steps": 864,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 942780789120.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8114321139277666,
    "learning_rate": 0.00020277068054397267,
    "num_train_epochs": 9,
    "temperature": 7
  }
}