trainer_state.json

{
  "best_metric": 0.9114666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0691/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 54.23231506347656,
      "learning_rate": 0.00042857142857142855,
      "loss": 1.2861,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8269333333333333,
      "eval_loss": 0.5942741632461548,
      "eval_runtime": 13.9678,
      "eval_samples_per_second": 268.475,
      "eval_steps_per_second": 4.224,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 47.50257873535156,
      "learning_rate": 0.00035714285714285714,
      "loss": 0.4192,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8704,
      "eval_loss": 0.4213908314704895,
      "eval_runtime": 13.3314,
      "eval_samples_per_second": 281.29,
      "eval_steps_per_second": 4.426,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 23.521650314331055,
      "learning_rate": 0.0002857142857142857,
      "loss": 0.2589,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.3902606964111328,
      "eval_runtime": 13.1164,
      "eval_samples_per_second": 285.902,
      "eval_steps_per_second": 4.498,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 16.23404884338379,
      "learning_rate": 0.00021428571428571427,
      "loss": 0.169,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8914666666666666,
      "eval_loss": 0.3886300027370453,
      "eval_runtime": 13.0801,
      "eval_samples_per_second": 286.696,
      "eval_steps_per_second": 4.511,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 42.43682861328125,
      "learning_rate": 0.00014285714285714284,
      "loss": 0.1124,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9093333333333333,
      "eval_loss": 0.33056527376174927,
      "eval_runtime": 13.1212,
      "eval_samples_per_second": 285.797,
      "eval_steps_per_second": 4.497,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 30.834156036376953,
      "learning_rate": 7.142857142857142e-05,
      "loss": 0.0671,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9114666666666666,
      "eval_loss": 0.34325259923934937,
      "eval_runtime": 13.1649,
      "eval_samples_per_second": 284.847,
      "eval_steps_per_second": 4.482,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.9058575437312e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}