trainer_state.json

{
  "best_metric": 0.7088,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0806/checkpoints/checkpoint-999",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 999,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 111.59214782714844,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 3.637,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5512,
      "eval_loss": 3.133552074432373,
      "eval_runtime": 14.022,
      "eval_samples_per_second": 267.438,
      "eval_steps_per_second": 4.208,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 700.5872802734375,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 2.0706,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.684,
      "eval_loss": 1.4993630647659302,
      "eval_runtime": 14.0379,
      "eval_samples_per_second": 267.135,
      "eval_steps_per_second": 4.203,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 368.4249267578125,
      "learning_rate": 0.0,
      "loss": 1.3921,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7088,
      "eval_loss": 1.3639311790466309,
      "eval_runtime": 14.0757,
      "eval_samples_per_second": 266.416,
      "eval_steps_per_second": 4.192,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 999,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 2.4529287718656e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}