trainer_state.json

{
  "best_metric": 0.8893333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0946/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": 191.3410186767578,
      "learning_rate": 9.698463103929542e-05,
      "loss": 2.7119,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7541333333333333,
      "eval_loss": 0.9297605752944946,
      "eval_runtime": 4.8537,
      "eval_samples_per_second": 772.6,
      "eval_steps_per_second": 12.156,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 62.90114974975586,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.6434,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8498666666666667,
      "eval_loss": 0.49980631470680237,
      "eval_runtime": 4.8723,
      "eval_samples_per_second": 769.659,
      "eval_steps_per_second": 12.109,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 95.13540649414062,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.3774,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8741333333333333,
      "eval_loss": 0.432142049074173,
      "eval_runtime": 4.8662,
      "eval_samples_per_second": 770.627,
      "eval_steps_per_second": 12.125,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 330.0178527832031,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.2398,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8754666666666666,
      "eval_loss": 0.417522132396698,
      "eval_runtime": 5.0195,
      "eval_samples_per_second": 747.09,
      "eval_steps_per_second": 11.754,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 277.2745666503906,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.1585,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8778666666666667,
      "eval_loss": 0.42143574357032776,
      "eval_runtime": 4.8655,
      "eval_samples_per_second": 770.727,
      "eval_steps_per_second": 12.126,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 206.205078125,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.1183,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8893333333333333,
      "eval_loss": 0.41205281019210815,
      "eval_runtime": 4.9052,
      "eval_samples_per_second": 764.496,
      "eval_steps_per_second": 12.028,
      "step": 1998
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}