last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0031854742374771046,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00031854742374771045,
      "grad_norm": 3.8480565547943115,
      "learning_rate": 2e-05,
      "loss": 1.8005,
      "step": 1
    },
    {
      "epoch": 0.00031854742374771045,
      "eval_loss": 3.871753454208374,
      "eval_runtime": 39.3461,
      "eval_samples_per_second": 16.8,
      "eval_steps_per_second": 16.8,
      "step": 1
    },
    {
      "epoch": 0.0006370948474954209,
      "grad_norm": NaN,
      "learning_rate": 2e-05,
      "loss": 2.1265,
      "step": 2
    },
    {
      "epoch": 0.0009556422712431313,
      "grad_norm": NaN,
      "learning_rate": 2e-05,
      "loss": 3.9734,
      "step": 3
    },
    {
      "epoch": 0.0009556422712431313,
      "eval_loss": 3.871753454208374,
      "eval_runtime": 38.8119,
      "eval_samples_per_second": 17.031,
      "eval_steps_per_second": 17.031,
      "step": 3
    },
    {
      "epoch": 0.0012741896949908418,
      "grad_norm": 9.316009521484375,
      "learning_rate": 4e-05,
      "loss": 1.7856,
      "step": 4
    },
    {
      "epoch": 0.0015927371187385523,
      "grad_norm": 11.193156242370605,
      "learning_rate": 6e-05,
      "loss": 4.3318,
      "step": 5
    },
    {
      "epoch": 0.0019112845424862626,
      "grad_norm": 7.8824567794799805,
      "learning_rate": 8e-05,
      "loss": 2.8762,
      "step": 6
    },
    {
      "epoch": 0.0019112845424862626,
      "eval_loss": 3.5064570903778076,
      "eval_runtime": 38.8548,
      "eval_samples_per_second": 17.012,
      "eval_steps_per_second": 17.012,
      "step": 6
    },
    {
      "epoch": 0.002229831966233973,
      "grad_norm": 13.027796745300293,
      "learning_rate": 0.0001,
      "loss": 2.2495,
      "step": 7
    },
    {
      "epoch": 0.0025483793899816836,
      "grad_norm": 18.181108474731445,
      "learning_rate": 0.00012,
      "loss": 3.5551,
      "step": 8
    },
    {
      "epoch": 0.002866926813729394,
      "grad_norm": 9.172633171081543,
      "learning_rate": 0.00014,
      "loss": 1.5026,
      "step": 9
    },
    {
      "epoch": 0.002866926813729394,
      "eval_loss": 2.45552396774292,
      "eval_runtime": 38.888,
      "eval_samples_per_second": 16.998,
      "eval_steps_per_second": 16.998,
      "step": 9
    },
    {
      "epoch": 0.0031854742374771046,
      "grad_norm": 7.70857048034668,
      "learning_rate": 0.00016,
      "loss": 1.0961,
      "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": 360477150412800.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}