checkpoint-3000/trainer_state.json

{
  "best_metric": 0.7647623419761658,
  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-3000",
  "epoch": 0.3199658703071672,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05332764505119454,
      "grad_norm": 414.2552795410156,
      "learning_rate": 4.911120591581342e-07,
      "loss": 0.3809,
      "step": 500
    },
    {
      "epoch": 0.05332764505119454,
      "eval_loss": 1.1109352111816406,
      "eval_runtime": 63.6734,
      "eval_samples_per_second": 15.501,
      "eval_steps_per_second": 1.947,
      "step": 500
    },
    {
      "epoch": 0.10665529010238908,
      "grad_norm": 45.880027770996094,
      "learning_rate": 4.822241183162685e-07,
      "loss": 0.2649,
      "step": 1000
    },
    {
      "epoch": 0.10665529010238908,
      "eval_loss": 0.9588962197303772,
      "eval_runtime": 62.6825,
      "eval_samples_per_second": 15.746,
      "eval_steps_per_second": 1.978,
      "step": 1000
    },
    {
      "epoch": 0.1599829351535836,
      "grad_norm": 432.9915466308594,
      "learning_rate": 4.733361774744027e-07,
      "loss": 0.2104,
      "step": 1500
    },
    {
      "epoch": 0.1599829351535836,
      "eval_loss": 0.90110844373703,
      "eval_runtime": 62.6757,
      "eval_samples_per_second": 15.748,
      "eval_steps_per_second": 1.978,
      "step": 1500
    },
    {
      "epoch": 0.21331058020477817,
      "grad_norm": 0.014208819717168808,
      "learning_rate": 4.6444823663253695e-07,
      "loss": 0.1849,
      "step": 2000
    },
    {
      "epoch": 0.21331058020477817,
      "eval_loss": 0.8570474982261658,
      "eval_runtime": 63.7296,
      "eval_samples_per_second": 15.487,
      "eval_steps_per_second": 1.946,
      "step": 2000
    },
    {
      "epoch": 0.2666382252559727,
      "grad_norm": 174.73324584960938,
      "learning_rate": 4.5556029579067116e-07,
      "loss": 0.2056,
      "step": 2500
    },
    {
      "epoch": 0.2666382252559727,
      "eval_loss": 0.7895939350128174,
      "eval_runtime": 63.8387,
      "eval_samples_per_second": 15.461,
      "eval_steps_per_second": 1.942,
      "step": 2500
    },
    {
      "epoch": 0.3199658703071672,
      "grad_norm": 9.852073823424234e-09,
      "learning_rate": 4.4667235494880547e-07,
      "loss": 0.159,
      "step": 3000
    },
    {
      "epoch": 0.3199658703071672,
      "eval_loss": 0.7647623419761658,
      "eval_runtime": 63.9155,
      "eval_samples_per_second": 15.442,
      "eval_steps_per_second": 1.94,
      "step": 3000
    }
  ],
  "logging_steps": 500,
  "max_steps": 28128,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 1079994570482520.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}