trainer_state.json

{
  "best_metric": 0.8442666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0456/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": 141.430908203125,
      "learning_rate": 4.849231551964771e-05,
      "loss": 3.6683,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4530666666666667,
      "eval_loss": 3.001512289047241,
      "eval_runtime": 8.0165,
      "eval_samples_per_second": 467.784,
      "eval_steps_per_second": 7.36,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 164.22503662109375,
      "learning_rate": 4.415111107797445e-05,
      "loss": 1.7505,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7197333333333333,
      "eval_loss": 1.0588290691375732,
      "eval_runtime": 8.0393,
      "eval_samples_per_second": 466.456,
      "eval_steps_per_second": 7.339,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 201.58348083496094,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.8967,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7866666666666666,
      "eval_loss": 0.7354041934013367,
      "eval_runtime": 8.0238,
      "eval_samples_per_second": 467.361,
      "eval_steps_per_second": 7.353,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 178.6982421875,
      "learning_rate": 2.9341204441673266e-05,
      "loss": 0.6863,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8165333333333333,
      "eval_loss": 0.5949721336364746,
      "eval_runtime": 8.04,
      "eval_samples_per_second": 466.417,
      "eval_steps_per_second": 7.338,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 171.48495483398438,
      "learning_rate": 2.0658795558326743e-05,
      "loss": 0.5798,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8269333333333333,
      "eval_loss": 0.571071445941925,
      "eval_runtime": 7.9401,
      "eval_samples_per_second": 472.285,
      "eval_steps_per_second": 7.431,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 156.7442626953125,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 0.5305,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8368,
      "eval_loss": 0.5271461009979248,
      "eval_runtime": 8.1347,
      "eval_samples_per_second": 460.99,
      "eval_steps_per_second": 7.253,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 61.661930084228516,
      "learning_rate": 5.848888922025553e-06,
      "loss": 0.4848,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8442666666666667,
      "eval_loss": 0.4976595640182495,
      "eval_runtime": 7.9971,
      "eval_samples_per_second": 468.919,
      "eval_steps_per_second": 7.378,
      "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
}