last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.014858841010401188,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0014858841010401188,
      "grad_norm": 23.471858978271484,
      "learning_rate": 2e-05,
      "loss": 10.0714,
      "step": 1
    },
    {
      "epoch": 0.0014858841010401188,
      "eval_loss": 9.684135437011719,
      "eval_runtime": 31.905,
      "eval_samples_per_second": 8.901,
      "eval_steps_per_second": 4.451,
      "step": 1
    },
    {
      "epoch": 0.0029717682020802376,
      "grad_norm": 24.766386032104492,
      "learning_rate": 4e-05,
      "loss": 9.0429,
      "step": 2
    },
    {
      "epoch": 0.004457652303120356,
      "grad_norm": 21.197526931762695,
      "learning_rate": 6e-05,
      "loss": 9.8085,
      "step": 3
    },
    {
      "epoch": 0.004457652303120356,
      "eval_loss": 9.287724494934082,
      "eval_runtime": 30.5477,
      "eval_samples_per_second": 9.297,
      "eval_steps_per_second": 4.648,
      "step": 3
    },
    {
      "epoch": 0.005943536404160475,
      "grad_norm": 24.982576370239258,
      "learning_rate": 8e-05,
      "loss": 9.791,
      "step": 4
    },
    {
      "epoch": 0.007429420505200594,
      "grad_norm": 21.579788208007812,
      "learning_rate": 0.0001,
      "loss": 8.7649,
      "step": 5
    },
    {
      "epoch": 0.008915304606240713,
      "grad_norm": 20.995973587036133,
      "learning_rate": 0.00012,
      "loss": 8.723,
      "step": 6
    },
    {
      "epoch": 0.008915304606240713,
      "eval_loss": 6.387414932250977,
      "eval_runtime": 29.634,
      "eval_samples_per_second": 9.584,
      "eval_steps_per_second": 4.792,
      "step": 6
    },
    {
      "epoch": 0.010401188707280832,
      "grad_norm": 36.22989273071289,
      "learning_rate": 0.00014,
      "loss": 6.0263,
      "step": 7
    },
    {
      "epoch": 0.01188707280832095,
      "grad_norm": 17.54269790649414,
      "learning_rate": 0.00016,
      "loss": 5.0566,
      "step": 8
    },
    {
      "epoch": 0.01337295690936107,
      "grad_norm": 24.35970115661621,
      "learning_rate": 0.00018,
      "loss": 4.6364,
      "step": 9
    },
    {
      "epoch": 0.01337295690936107,
      "eval_loss": 3.430145025253296,
      "eval_runtime": 28.208,
      "eval_samples_per_second": 10.068,
      "eval_steps_per_second": 5.034,
      "step": 9
    },
    {
      "epoch": 0.014858841010401188,
      "grad_norm": 14.752691268920898,
      "learning_rate": 0.0002,
      "loss": 2.9616,
      "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": 720954300825600.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}