{
  "best_global_step": 2000,
  "best_metric": 0.0627085492014885,
  "best_model_checkpoint": "OpenAHA-Calvin-1p-3b/checkpoint-2000",
  "epoch": 2.6496519721577725,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.6629101756711966,
      "grad_norm": 4.84523344039917,
      "learning_rate": 1.733071638861629e-05,
      "loss": 0.5578,
      "step": 500
    },
    {
      "epoch": 0.6629101756711966,
      "eval_loss": 0.1356603056192398,
      "eval_runtime": 31.1266,
      "eval_samples_per_second": 13.044,
      "eval_steps_per_second": 1.638,
      "step": 500
    },
    {
      "epoch": 1.3248259860788862,
      "grad_norm": 7.176058769226074,
      "learning_rate": 1.2423945044160943e-05,
      "loss": 0.0908,
      "step": 1000
    },
    {
      "epoch": 1.3248259860788862,
      "eval_loss": 0.08840304613113403,
      "eval_runtime": 29.435,
      "eval_samples_per_second": 13.793,
      "eval_steps_per_second": 1.733,
      "step": 1000
    },
    {
      "epoch": 1.9877361617500828,
      "grad_norm": 2.4896891117095947,
      "learning_rate": 7.517173699705595e-06,
      "loss": 0.0603,
      "step": 1500
    },
    {
      "epoch": 1.9877361617500828,
      "eval_loss": 0.07601085305213928,
      "eval_runtime": 29.6732,
      "eval_samples_per_second": 13.682,
      "eval_steps_per_second": 1.719,
      "step": 1500
    },
    {
      "epoch": 2.6496519721577725,
      "grad_norm": 5.802169322967529,
      "learning_rate": 2.6104023552502455e-06,
      "loss": 0.0406,
      "step": 2000
    },
    {
      "epoch": 2.6496519721577725,
      "eval_loss": 0.0627085492014885,
      "eval_runtime": 29.4236,
      "eval_samples_per_second": 13.798,
      "eval_steps_per_second": 1.733,
      "step": 2000
    }
  ],
  "logging_steps": 500,
  "max_steps": 2265,
  "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": 9.832880380047903e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}