trainer_state.json

{
  "best_metric": 0.7674666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0602/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": 339.4329833984375,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 3.864,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3045333333333333,
      "eval_loss": 3.773716926574707,
      "eval_runtime": 5.5567,
      "eval_samples_per_second": 674.864,
      "eval_steps_per_second": 10.618,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 263.06085205078125,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 3.4914,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4704,
      "eval_loss": 3.247539520263672,
      "eval_runtime": 5.4875,
      "eval_samples_per_second": 683.372,
      "eval_steps_per_second": 10.752,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 427.2501525878906,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 2.5959,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.616,
      "eval_loss": 2.013122320175171,
      "eval_runtime": 5.6435,
      "eval_samples_per_second": 664.486,
      "eval_steps_per_second": 10.455,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 168.39389038085938,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 1.7143,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7021333333333334,
      "eval_loss": 1.36212956905365,
      "eval_runtime": 5.6083,
      "eval_samples_per_second": 668.651,
      "eval_steps_per_second": 10.52,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 138.50547790527344,
      "learning_rate": 8.571428571428571e-06,
      "loss": 1.2959,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7424,
      "eval_loss": 1.1159014701843262,
      "eval_runtime": 5.5072,
      "eval_samples_per_second": 680.929,
      "eval_steps_per_second": 10.713,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 250.3559112548828,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 1.1102,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7621333333333333,
      "eval_loss": 1.032724142074585,
      "eval_runtime": 5.5088,
      "eval_samples_per_second": 680.734,
      "eval_steps_per_second": 10.71,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 320.6709899902344,
      "learning_rate": 0.0,
      "loss": 1.0318,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7674666666666666,
      "eval_loss": 0.9580545425415039,
      "eval_runtime": 5.6257,
      "eval_samples_per_second": 666.583,
      "eval_steps_per_second": 10.488,
      "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
}