trainer_state.json

{
  "best_metric": 0.8336,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0757/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 86.14022827148438,
      "learning_rate": 2.8514533018536286e-05,
      "loss": 3.8058,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.48,
      "eval_loss": 3.57427716255188,
      "eval_runtime": 5.0083,
      "eval_samples_per_second": 748.754,
      "eval_steps_per_second": 11.78,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 107.25267028808594,
      "learning_rate": 2.4352347027881003e-05,
      "loss": 2.8935,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.66,
      "eval_loss": 2.069218635559082,
      "eval_runtime": 4.9924,
      "eval_samples_per_second": 751.144,
      "eval_steps_per_second": 11.818,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 98.55683135986328,
      "learning_rate": 1.8337814009344716e-05,
      "loss": 1.4133,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7816,
      "eval_loss": 0.9759421944618225,
      "eval_runtime": 5.0106,
      "eval_samples_per_second": 748.412,
      "eval_steps_per_second": 11.775,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 106.32801818847656,
      "learning_rate": 1.1662185990655285e-05,
      "loss": 0.8554,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8173333333333334,
      "eval_loss": 0.7378785014152527,
      "eval_runtime": 5.2099,
      "eval_samples_per_second": 719.785,
      "eval_steps_per_second": 11.325,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 217.58209228515625,
      "learning_rate": 5.647652972118998e-06,
      "loss": 0.6959,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.828,
      "eval_loss": 0.6526791453361511,
      "eval_runtime": 4.9918,
      "eval_samples_per_second": 751.235,
      "eval_steps_per_second": 11.819,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 193.79254150390625,
      "learning_rate": 1.4854669814637145e-06,
      "loss": 0.6416,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8242666666666667,
      "eval_loss": 0.6352399587631226,
      "eval_runtime": 5.0215,
      "eval_samples_per_second": 746.794,
      "eval_steps_per_second": 11.75,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 136.4894561767578,
      "learning_rate": 0.0,
      "loss": 0.617,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8336,
      "eval_loss": 0.6287975311279297,
      "eval_runtime": 5.0124,
      "eval_samples_per_second": 748.152,
      "eval_steps_per_second": 11.771,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}