trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.989010989010989,
  "eval_steps": 500,
  "global_step": 102,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.29304029304029305,
      "grad_norm": 6.00564017416128,
      "learning_rate": 5e-06,
      "loss": 0.8612,
      "step": 10
    },
    {
      "epoch": 0.5860805860805861,
      "grad_norm": 3.810412857546984,
      "learning_rate": 5e-06,
      "loss": 0.7583,
      "step": 20
    },
    {
      "epoch": 0.8791208791208791,
      "grad_norm": 2.835346031353567,
      "learning_rate": 5e-06,
      "loss": 0.7171,
      "step": 30
    },
    {
      "epoch": 0.9963369963369964,
      "eval_loss": 0.6949356198310852,
      "eval_runtime": 23.9449,
      "eval_samples_per_second": 38.338,
      "eval_steps_per_second": 0.626,
      "step": 34
    },
    {
      "epoch": 1.1721611721611722,
      "grad_norm": 3.6787612930551394,
      "learning_rate": 5e-06,
      "loss": 0.7449,
      "step": 40
    },
    {
      "epoch": 1.4652014652014653,
      "grad_norm": 0.8606934602884543,
      "learning_rate": 5e-06,
      "loss": 0.6554,
      "step": 50
    },
    {
      "epoch": 1.7582417582417582,
      "grad_norm": 0.9805632267452244,
      "learning_rate": 5e-06,
      "loss": 0.6415,
      "step": 60
    },
    {
      "epoch": 1.9926739926739927,
      "eval_loss": 0.6563424468040466,
      "eval_runtime": 24.5646,
      "eval_samples_per_second": 37.371,
      "eval_steps_per_second": 0.611,
      "step": 68
    },
    {
      "epoch": 2.051282051282051,
      "grad_norm": 1.0030553009772163,
      "learning_rate": 5e-06,
      "loss": 0.6776,
      "step": 70
    },
    {
      "epoch": 2.3443223443223444,
      "grad_norm": 1.1368386124641172,
      "learning_rate": 5e-06,
      "loss": 0.5809,
      "step": 80
    },
    {
      "epoch": 2.6373626373626373,
      "grad_norm": 1.2620401338366838,
      "learning_rate": 5e-06,
      "loss": 0.5752,
      "step": 90
    },
    {
      "epoch": 2.9304029304029307,
      "grad_norm": 0.8239444630205299,
      "learning_rate": 5e-06,
      "loss": 0.5713,
      "step": 100
    },
    {
      "epoch": 2.989010989010989,
      "eval_loss": 0.6484237909317017,
      "eval_runtime": 22.7074,
      "eval_samples_per_second": 40.427,
      "eval_steps_per_second": 0.661,
      "step": 102
    },
    {
      "epoch": 2.989010989010989,
      "step": 102,
      "total_flos": 170644419379200.0,
      "train_loss": 0.6762744293493383,
      "train_runtime": 3624.8354,
      "train_samples_per_second": 14.425,
      "train_steps_per_second": 0.028
    }
  ],
  "logging_steps": 10,
  "max_steps": 102,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 170644419379200.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}