last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.00016234293321211728,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.6234293321211728e-05,
      "grad_norm": 9.819344520568848,
      "learning_rate": 2e-05,
      "loss": 3.1041,
      "step": 1
    },
    {
      "epoch": 1.6234293321211728e-05,
      "eval_loss": NaN,
      "eval_runtime": 1684.6563,
      "eval_samples_per_second": 7.698,
      "eval_steps_per_second": 7.698,
      "step": 1
    },
    {
      "epoch": 3.2468586642423456e-05,
      "grad_norm": 11.458588600158691,
      "learning_rate": 4e-05,
      "loss": 3.7572,
      "step": 2
    },
    {
      "epoch": 4.8702879963635184e-05,
      "grad_norm": 11.369792938232422,
      "learning_rate": 6e-05,
      "loss": 2.8828,
      "step": 3
    },
    {
      "epoch": 4.8702879963635184e-05,
      "eval_loss": NaN,
      "eval_runtime": 1684.5918,
      "eval_samples_per_second": 7.698,
      "eval_steps_per_second": 7.698,
      "step": 3
    },
    {
      "epoch": 6.493717328484691e-05,
      "grad_norm": 9.130757331848145,
      "learning_rate": 8e-05,
      "loss": 3.286,
      "step": 4
    },
    {
      "epoch": 8.117146660605864e-05,
      "grad_norm": 9.889158248901367,
      "learning_rate": 0.0001,
      "loss": 2.3931,
      "step": 5
    },
    {
      "epoch": 9.740575992727037e-05,
      "grad_norm": 7.035650253295898,
      "learning_rate": 0.00012,
      "loss": 2.4323,
      "step": 6
    },
    {
      "epoch": 9.740575992727037e-05,
      "eval_loss": NaN,
      "eval_runtime": 1684.7423,
      "eval_samples_per_second": 7.697,
      "eval_steps_per_second": 7.697,
      "step": 6
    },
    {
      "epoch": 0.0001136400532484821,
      "grad_norm": 9.116958618164062,
      "learning_rate": 0.00014,
      "loss": 2.7834,
      "step": 7
    },
    {
      "epoch": 0.00012987434656969382,
      "grad_norm": 6.480123996734619,
      "learning_rate": 0.00016,
      "loss": 2.6653,
      "step": 8
    },
    {
      "epoch": 0.00014610863989090555,
      "grad_norm": 6.486927032470703,
      "learning_rate": 0.00018,
      "loss": 2.2272,
      "step": 9
    },
    {
      "epoch": 0.00014610863989090555,
      "eval_loss": NaN,
      "eval_runtime": 1684.8043,
      "eval_samples_per_second": 7.697,
      "eval_steps_per_second": 7.697,
      "step": 9
    },
    {
      "epoch": 0.00016234293321211728,
      "grad_norm": 5.726334571838379,
      "learning_rate": 0.0002,
      "loss": 2.212,
      "step": 10
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2010596581048320.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}