checkpoint-2975/trainer_state.json

{
  "best_global_step": 2975,
  "best_metric": 0.12348528951406479,
  "best_model_checkpoint": "metaclip-2-image-classification/checkpoint-2975",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 2975,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8403361344537815,
      "grad_norm": 36.58893585205078,
      "learning_rate": 1.6929914529914532e-05,
      "loss": 0.8451,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8110012620950778,
      "eval_loss": 0.4803926944732666,
      "eval_model_preparation_time": 0.0159,
      "eval_runtime": 184.4364,
      "eval_samples_per_second": 103.103,
      "eval_steps_per_second": 12.888,
      "step": 595
    },
    {
      "epoch": 1.680672268907563,
      "grad_norm": 25.577220916748047,
      "learning_rate": 1.3511111111111112e-05,
      "loss": 0.5616,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8604859066049643,
      "eval_loss": 0.35589027404785156,
      "eval_model_preparation_time": 0.0159,
      "eval_runtime": 182.6442,
      "eval_samples_per_second": 104.115,
      "eval_steps_per_second": 13.014,
      "step": 1190
    },
    {
      "epoch": 2.5210084033613445,
      "grad_norm": 22.501468658447266,
      "learning_rate": 1.0092307692307693e-05,
      "loss": 0.4402,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8958245687841817,
      "eval_loss": 0.26440638303756714,
      "eval_model_preparation_time": 0.0159,
      "eval_runtime": 185.7803,
      "eval_samples_per_second": 102.357,
      "eval_steps_per_second": 12.795,
      "step": 1785
    },
    {
      "epoch": 3.361344537815126,
      "grad_norm": 27.099328994750977,
      "learning_rate": 6.673504273504275e-06,
      "loss": 0.3172,
      "step": 2000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9453092132940681,
      "eval_loss": 0.15674497187137604,
      "eval_model_preparation_time": 0.0159,
      "eval_runtime": 181.8977,
      "eval_samples_per_second": 104.542,
      "eval_steps_per_second": 13.068,
      "step": 2380
    },
    {
      "epoch": 4.201680672268908,
      "grad_norm": 28.719318389892578,
      "learning_rate": 3.2547008547008553e-06,
      "loss": 0.2164,
      "step": 2500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9559318468657972,
      "eval_loss": 0.12348528951406479,
      "eval_model_preparation_time": 0.0159,
      "eval_runtime": 186.47,
      "eval_samples_per_second": 101.979,
      "eval_steps_per_second": 12.747,
      "step": 2975
    }
  ],
  "logging_steps": 500,
  "max_steps": 2975,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 1.854202333390848e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}