trainer_state.json

{
  "best_metric": 0.8733333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0407/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": 220.46487426757812,
      "learning_rate": 3e-05,
      "loss": 3.8328,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4325333333333333,
      "eval_loss": 3.657475709915161,
      "eval_runtime": 12.7319,
      "eval_samples_per_second": 294.535,
      "eval_steps_per_second": 4.634,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 102.22537994384766,
      "learning_rate": 3e-05,
      "loss": 3.071,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.62,
      "eval_loss": 2.281257390975952,
      "eval_runtime": 12.4794,
      "eval_samples_per_second": 300.495,
      "eval_steps_per_second": 4.728,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 127.67041778564453,
      "learning_rate": 3e-05,
      "loss": 1.4915,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7672,
      "eval_loss": 0.9350252747535706,
      "eval_runtime": 12.4649,
      "eval_samples_per_second": 300.844,
      "eval_steps_per_second": 4.733,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 514.3515625,
      "learning_rate": 3e-05,
      "loss": 0.7953,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8149333333333333,
      "eval_loss": 0.6676617860794067,
      "eval_runtime": 12.5944,
      "eval_samples_per_second": 297.752,
      "eval_steps_per_second": 4.685,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 151.02589416503906,
      "learning_rate": 3e-05,
      "loss": 0.5736,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8432,
      "eval_loss": 0.5224517583847046,
      "eval_runtime": 12.7868,
      "eval_samples_per_second": 293.27,
      "eval_steps_per_second": 4.614,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 140.41590881347656,
      "learning_rate": 3e-05,
      "loss": 0.4406,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8661333333333333,
      "eval_loss": 0.45962339639663696,
      "eval_runtime": 12.5008,
      "eval_samples_per_second": 299.98,
      "eval_steps_per_second": 4.72,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 148.19427490234375,
      "learning_rate": 3e-05,
      "loss": 0.3546,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8733333333333333,
      "eval_loss": 0.4310742914676666,
      "eval_runtime": 13.107,
      "eval_samples_per_second": 286.106,
      "eval_steps_per_second": 4.501,
      "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
}