trainer_state.json

{
  "best_metric": 0.8573333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0046/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 127.92997741699219,
      "learning_rate": 5.6e-05,
      "loss": 3.3425,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.656,
      "eval_loss": 1.993105411529541,
      "eval_runtime": 14.2699,
      "eval_samples_per_second": 262.791,
      "eval_steps_per_second": 4.135,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 249.5431671142578,
      "learning_rate": 4.2e-05,
      "loss": 1.0269,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8064,
      "eval_loss": 0.6610711812973022,
      "eval_runtime": 14.3291,
      "eval_samples_per_second": 261.704,
      "eval_steps_per_second": 4.117,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 161.84127807617188,
      "learning_rate": 2.8e-05,
      "loss": 0.5545,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8432,
      "eval_loss": 0.5175615549087524,
      "eval_runtime": 14.4333,
      "eval_samples_per_second": 259.817,
      "eval_steps_per_second": 4.088,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 120.10369110107422,
      "learning_rate": 1.4e-05,
      "loss": 0.4146,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8498666666666667,
      "eval_loss": 0.48323217034339905,
      "eval_runtime": 14.393,
      "eval_samples_per_second": 260.544,
      "eval_steps_per_second": 4.099,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 246.61317443847656,
      "learning_rate": 0.0,
      "loss": 0.3695,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8573333333333333,
      "eval_loss": 0.4700867831707001,
      "eval_runtime": 14.2312,
      "eval_samples_per_second": 263.506,
      "eval_steps_per_second": 4.146,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 4.088214619776e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}