trainer_state.json

{
  "best_metric": 0.8858666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0528/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": 743.876220703125,
      "learning_rate": 8.554359905560886e-05,
      "loss": 2.7892,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7544,
      "eval_loss": 0.9682375192642212,
      "eval_runtime": 8.1458,
      "eval_samples_per_second": 460.361,
      "eval_steps_per_second": 7.243,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 108.92616271972656,
      "learning_rate": 7.305704108364301e-05,
      "loss": 0.6658,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.848,
      "eval_loss": 0.5306808948516846,
      "eval_runtime": 7.9762,
      "eval_samples_per_second": 470.147,
      "eval_steps_per_second": 7.397,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 184.5489044189453,
      "learning_rate": 5.501344202803415e-05,
      "loss": 0.3904,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8717333333333334,
      "eval_loss": 0.43771159648895264,
      "eval_runtime": 8.0117,
      "eval_samples_per_second": 468.068,
      "eval_steps_per_second": 7.364,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 57.365901947021484,
      "learning_rate": 3.498655797196586e-05,
      "loss": 0.2536,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8762666666666666,
      "eval_loss": 0.40537744760513306,
      "eval_runtime": 7.9059,
      "eval_samples_per_second": 474.33,
      "eval_steps_per_second": 7.463,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 97.658447265625,
      "learning_rate": 1.6942958916356995e-05,
      "loss": 0.1931,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8829333333333333,
      "eval_loss": 0.4033312499523163,
      "eval_runtime": 7.9767,
      "eval_samples_per_second": 470.117,
      "eval_steps_per_second": 7.397,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 71.80194091796875,
      "learning_rate": 4.456400944391144e-06,
      "loss": 0.1535,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8842666666666666,
      "eval_loss": 0.40199601650238037,
      "eval_runtime": 7.8535,
      "eval_samples_per_second": 477.494,
      "eval_steps_per_second": 7.513,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 87.62568664550781,
      "learning_rate": 0.0,
      "loss": 0.1386,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8858666666666667,
      "eval_loss": 0.38963568210601807,
      "eval_runtime": 7.9253,
      "eval_samples_per_second": 473.166,
      "eval_steps_per_second": 7.444,
      "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
}