trainer_state.json

{
  "best_metric": 0.8408,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0491/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 94.95722961425781,
      "learning_rate": 2.625e-05,
      "loss": 3.844,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.40586666666666665,
      "eval_loss": 3.703624963760376,
      "eval_runtime": 9.2348,
      "eval_samples_per_second": 406.074,
      "eval_steps_per_second": 6.389,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 207.0720672607422,
      "learning_rate": 2.25e-05,
      "loss": 3.2745,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5757333333333333,
      "eval_loss": 2.7443695068359375,
      "eval_runtime": 7.2796,
      "eval_samples_per_second": 515.14,
      "eval_steps_per_second": 8.105,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 187.76119995117188,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 2.0933,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7173333333333334,
      "eval_loss": 1.5060018301010132,
      "eval_runtime": 7.3931,
      "eval_samples_per_second": 507.23,
      "eval_steps_per_second": 7.98,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 224.58616638183594,
      "learning_rate": 1.5e-05,
      "loss": 1.2274,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7837333333333333,
      "eval_loss": 0.9947818517684937,
      "eval_runtime": 7.3759,
      "eval_samples_per_second": 508.41,
      "eval_steps_per_second": 7.999,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 104.12053680419922,
      "learning_rate": 1.125e-05,
      "loss": 0.8696,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.812,
      "eval_loss": 0.8177639245986938,
      "eval_runtime": 7.6995,
      "eval_samples_per_second": 487.048,
      "eval_steps_per_second": 7.663,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 252.95147705078125,
      "learning_rate": 7.5e-06,
      "loss": 0.7181,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8306666666666667,
      "eval_loss": 0.6887484192848206,
      "eval_runtime": 7.4783,
      "eval_samples_per_second": 501.452,
      "eval_steps_per_second": 7.89,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 97.23067474365234,
      "learning_rate": 3.75e-06,
      "loss": 0.6425,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8389333333333333,
      "eval_loss": 0.6621872782707214,
      "eval_runtime": 7.401,
      "eval_samples_per_second": 506.685,
      "eval_steps_per_second": 7.972,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 136.90359497070312,
      "learning_rate": 0.0,
      "loss": 0.6038,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8408,
      "eval_loss": 0.6270493268966675,
      "eval_runtime": 7.5651,
      "eval_samples_per_second": 495.698,
      "eval_steps_per_second": 7.799,
      "step": 2664
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.5411433916416e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}