trainer_state.json

{
  "best_metric": 0.904,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0897/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": Infinity,
      "learning_rate": 9.621650897886543e-05,
      "loss": 1.4429,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7946666666666666,
      "eval_loss": 0.7274636030197144,
      "eval_runtime": 6.9804,
      "eval_samples_per_second": 537.218,
      "eval_steps_per_second": 8.452,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.0918819904327393,
      "learning_rate": 8.539700818671957e-05,
      "loss": 0.4746,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.844,
      "eval_loss": 0.507297933101654,
      "eval_runtime": 7.2528,
      "eval_samples_per_second": 517.039,
      "eval_steps_per_second": 8.135,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 6.924306924189557e-05,
      "loss": 0.2758,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8594666666666667,
      "eval_loss": 0.45288288593292236,
      "eval_runtime": 6.687,
      "eval_samples_per_second": 560.787,
      "eval_steps_per_second": 8.823,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.26704248785972595,
      "learning_rate": 5.011792754282769e-05,
      "loss": 0.1547,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8776,
      "eval_loss": 0.42758631706237793,
      "eval_runtime": 6.6247,
      "eval_samples_per_second": 566.064,
      "eval_steps_per_second": 8.906,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.648855447769165,
      "learning_rate": 3.0974832444380136e-05,
      "loss": 0.0723,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8850666666666667,
      "eval_loss": 0.3976556956768036,
      "eval_runtime": 6.561,
      "eval_samples_per_second": 571.563,
      "eval_steps_per_second": 8.993,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.9727654457092285,
      "learning_rate": 1.472814664270234e-05,
      "loss": 0.0312,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.896,
      "eval_loss": 0.37830182909965515,
      "eval_runtime": 6.633,
      "eval_samples_per_second": 565.356,
      "eval_steps_per_second": 8.895,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.06778880208730698,
      "learning_rate": 3.851280774526222e-06,
      "loss": 0.013,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.904,
      "eval_loss": 0.37247687578201294,
      "eval_runtime": 6.5673,
      "eval_samples_per_second": 571.012,
      "eval_steps_per_second": 8.984,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}