checkpoint-21/trainer_state.json

{
  "best_metric": 0.4889541566371918,
  "best_model_checkpoint": "./xlam_1epoch_pythonic5e-6/checkpoint-21",
  "epoch": 0.345679012345679,
  "eval_steps": 7,
  "global_step": 21,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11522633744855967,
      "eval_loss": 0.4905904531478882,
      "eval_runtime": 40.3827,
      "eval_samples_per_second": 1.882,
      "eval_steps_per_second": 1.882,
      "step": 7
    },
    {
      "epoch": 0.1646090534979424,
      "grad_norm": 6.2620391845703125,
      "learning_rate": 4.7689385491773934e-06,
      "loss": 8.4783,
      "step": 10
    },
    {
      "epoch": 0.23045267489711935,
      "eval_loss": 0.4900592863559723,
      "eval_runtime": 40.7901,
      "eval_samples_per_second": 1.863,
      "eval_steps_per_second": 1.863,
      "step": 14
    },
    {
      "epoch": 0.3292181069958848,
      "grad_norm": 6.400896072387695,
      "learning_rate": 3.9029676634059565e-06,
      "loss": 8.8473,
      "step": 20
    },
    {
      "epoch": 0.345679012345679,
      "eval_loss": 0.4889541566371918,
      "eval_runtime": 40.6095,
      "eval_samples_per_second": 1.871,
      "eval_steps_per_second": 1.871,
      "step": 21
    }
  ],
  "logging_steps": 10,
  "max_steps": 60,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 7,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.996724248941691e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}