trainer_state.json

{
  "best_metric": 0.8898666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0838/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": 171.59881591796875,
      "learning_rate": 8.571428571428571e-05,
      "loss": 2.9072,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6781333333333334,
      "eval_loss": 1.3353562355041504,
      "eval_runtime": 9.1091,
      "eval_samples_per_second": 411.676,
      "eval_steps_per_second": 6.477,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 106.56681823730469,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.8456,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8306666666666667,
      "eval_loss": 0.6036928296089172,
      "eval_runtime": 8.0823,
      "eval_samples_per_second": 463.977,
      "eval_steps_per_second": 7.3,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 82.1512451171875,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.501,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8597333333333333,
      "eval_loss": 0.4776412546634674,
      "eval_runtime": 8.026,
      "eval_samples_per_second": 467.231,
      "eval_steps_per_second": 7.351,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 147.7592315673828,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.3849,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8717333333333334,
      "eval_loss": 0.44812992215156555,
      "eval_runtime": 8.0869,
      "eval_samples_per_second": 463.71,
      "eval_steps_per_second": 7.296,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 200.98806762695312,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.3041,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8848,
      "eval_loss": 0.40260910987854004,
      "eval_runtime": 8.0503,
      "eval_samples_per_second": 465.819,
      "eval_steps_per_second": 7.329,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 59.4629020690918,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.2589,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8773333333333333,
      "eval_loss": 0.41696789860725403,
      "eval_runtime": 8.0775,
      "eval_samples_per_second": 464.255,
      "eval_steps_per_second": 7.304,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 52.43292999267578,
      "learning_rate": 0.0,
      "loss": 0.239,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8898666666666667,
      "eval_loss": 0.39091062545776367,
      "eval_runtime": 8.0432,
      "eval_samples_per_second": 466.234,
      "eval_steps_per_second": 7.335,
      "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
}