trainer_state.json

{
  "best_metric": 0.896,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0548/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": 165.78256225585938,
      "learning_rate": 9.619397662556435e-05,
      "loss": 2.5104,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7874666666666666,
      "eval_loss": 0.7962279319763184,
      "eval_runtime": 12.501,
      "eval_samples_per_second": 299.977,
      "eval_steps_per_second": 4.72,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 97.7348403930664,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.6035,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8616,
      "eval_loss": 0.46331414580345154,
      "eval_runtime": 12.4669,
      "eval_samples_per_second": 300.797,
      "eval_steps_per_second": 4.733,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 106.37476348876953,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.3475,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8842666666666666,
      "eval_loss": 0.3724762499332428,
      "eval_runtime": 12.5478,
      "eval_samples_per_second": 298.858,
      "eval_steps_per_second": 4.702,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 107.25939178466797,
      "learning_rate": 5e-05,
      "loss": 0.2215,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8904,
      "eval_loss": 0.34392863512039185,
      "eval_runtime": 12.4428,
      "eval_samples_per_second": 301.38,
      "eval_steps_per_second": 4.742,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 67.79757690429688,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.1488,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8944,
      "eval_loss": 0.346574991941452,
      "eval_runtime": 13.0895,
      "eval_samples_per_second": 286.49,
      "eval_steps_per_second": 4.507,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 67.08159637451172,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.1094,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8946666666666667,
      "eval_loss": 0.3477937579154968,
      "eval_runtime": 12.6286,
      "eval_samples_per_second": 296.944,
      "eval_steps_per_second": 4.672,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 44.662315368652344,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0843,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.896,
      "eval_loss": 0.3452041745185852,
      "eval_runtime": 12.6166,
      "eval_samples_per_second": 297.229,
      "eval_steps_per_second": 4.676,
      "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}