trainer_state.json

{
  "best_metric": 0.8170666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0916/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": 204.1887664794922,
      "learning_rate": 2.9095389311788626e-05,
      "loss": 3.8513,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3328,
      "eval_loss": 3.771674394607544,
      "eval_runtime": 8.3366,
      "eval_samples_per_second": 449.824,
      "eval_steps_per_second": 7.077,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 282.0765075683594,
      "learning_rate": 2.649066664678467e-05,
      "loss": 3.2543,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5546666666666666,
      "eval_loss": 2.5727603435516357,
      "eval_runtime": 8.5155,
      "eval_samples_per_second": 440.372,
      "eval_steps_per_second": 6.929,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 107.82434844970703,
      "learning_rate": 2.25e-05,
      "loss": 1.8694,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7045333333333333,
      "eval_loss": 1.3677384853363037,
      "eval_runtime": 8.5334,
      "eval_samples_per_second": 439.451,
      "eval_steps_per_second": 6.914,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 832.704833984375,
      "learning_rate": 1.760472266500396e-05,
      "loss": 1.1456,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7594666666666666,
      "eval_loss": 0.9404367208480835,
      "eval_runtime": 8.6525,
      "eval_samples_per_second": 433.398,
      "eval_steps_per_second": 6.819,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 168.1475830078125,
      "learning_rate": 1.2395277334996045e-05,
      "loss": 0.874,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7909333333333334,
      "eval_loss": 0.7899569272994995,
      "eval_runtime": 8.4142,
      "eval_samples_per_second": 445.676,
      "eval_steps_per_second": 7.012,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 278.5458679199219,
      "learning_rate": 7.500000000000004e-06,
      "loss": 0.7518,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8072,
      "eval_loss": 0.7163684368133545,
      "eval_runtime": 8.7579,
      "eval_samples_per_second": 428.186,
      "eval_steps_per_second": 6.737,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 225.07772827148438,
      "learning_rate": 3.5093333532153316e-06,
      "loss": 0.701,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8170666666666667,
      "eval_loss": 0.6719250082969666,
      "eval_runtime": 8.5028,
      "eval_samples_per_second": 441.029,
      "eval_steps_per_second": 6.939,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}