trainer_state.json

{
  "best_metric": 0.8853333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0127/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": 68.2916030883789,
      "learning_rate": 7.714285714285714e-05,
      "loss": 2.978,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7325333333333334,
      "eval_loss": 1.1154764890670776,
      "eval_runtime": 5.3414,
      "eval_samples_per_second": 702.061,
      "eval_steps_per_second": 11.046,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 68.89349365234375,
      "learning_rate": 6.428571428571429e-05,
      "loss": 0.755,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8434666666666667,
      "eval_loss": 0.5194705128669739,
      "eval_runtime": 5.3178,
      "eval_samples_per_second": 705.184,
      "eval_steps_per_second": 11.095,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 48.08671188354492,
      "learning_rate": 5.142857142857143e-05,
      "loss": 0.4725,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8634666666666667,
      "eval_loss": 0.43179377913475037,
      "eval_runtime": 5.3865,
      "eval_samples_per_second": 696.182,
      "eval_steps_per_second": 10.953,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 46.980934143066406,
      "learning_rate": 3.857142857142857e-05,
      "loss": 0.3597,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8733333333333333,
      "eval_loss": 0.39097362756729126,
      "eval_runtime": 5.3702,
      "eval_samples_per_second": 698.302,
      "eval_steps_per_second": 10.987,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 87.41561889648438,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 0.2933,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8794666666666666,
      "eval_loss": 0.3770959973335266,
      "eval_runtime": 5.2772,
      "eval_samples_per_second": 710.599,
      "eval_steps_per_second": 11.18,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 58.91222381591797,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 0.2542,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8813333333333333,
      "eval_loss": 0.3833366334438324,
      "eval_runtime": 5.3707,
      "eval_samples_per_second": 698.232,
      "eval_steps_per_second": 10.986,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 94.67659759521484,
      "learning_rate": 0.0,
      "loss": 0.2403,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8853333333333333,
      "eval_loss": 0.36212292313575745,
      "eval_runtime": 5.3522,
      "eval_samples_per_second": 700.641,
      "eval_steps_per_second": 11.023,
      "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
}