last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0006276675872457946,
  "eval_steps": 1,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 6.276675872457946e-05,
      "grad_norm": 5.331059455871582,
      "learning_rate": 2e-05,
      "loss": 5.6395,
      "step": 1
    },
    {
      "epoch": 6.276675872457946e-05,
      "eval_loss": 6.696534156799316,
      "eval_runtime": 1715.9511,
      "eval_samples_per_second": 3.91,
      "eval_steps_per_second": 1.955,
      "step": 1
    },
    {
      "epoch": 0.00012553351744915893,
      "grad_norm": 6.676981449127197,
      "learning_rate": 4e-05,
      "loss": 6.8763,
      "step": 2
    },
    {
      "epoch": 0.00012553351744915893,
      "eval_loss": 6.688595771789551,
      "eval_runtime": 1724.2223,
      "eval_samples_per_second": 3.891,
      "eval_steps_per_second": 1.946,
      "step": 2
    },
    {
      "epoch": 0.0001883002761737384,
      "grad_norm": 5.51637601852417,
      "learning_rate": 6e-05,
      "loss": 6.3732,
      "step": 3
    },
    {
      "epoch": 0.0001883002761737384,
      "eval_loss": 6.655938148498535,
      "eval_runtime": 1724.3203,
      "eval_samples_per_second": 3.891,
      "eval_steps_per_second": 1.946,
      "step": 3
    },
    {
      "epoch": 0.00025106703489831785,
      "grad_norm": 8.880010604858398,
      "learning_rate": 8e-05,
      "loss": 7.5347,
      "step": 4
    },
    {
      "epoch": 0.00025106703489831785,
      "eval_loss": 6.558417320251465,
      "eval_runtime": 1724.5664,
      "eval_samples_per_second": 3.89,
      "eval_steps_per_second": 1.945,
      "step": 4
    },
    {
      "epoch": 0.0003138337936228973,
      "grad_norm": 5.816895484924316,
      "learning_rate": 0.0001,
      "loss": 6.8217,
      "step": 5
    },
    {
      "epoch": 0.0003138337936228973,
      "eval_loss": 6.30959415435791,
      "eval_runtime": 1725.1208,
      "eval_samples_per_second": 3.889,
      "eval_steps_per_second": 1.945,
      "step": 5
    },
    {
      "epoch": 0.0003766005523474768,
      "grad_norm": 7.0391926765441895,
      "learning_rate": 0.00012,
      "loss": 5.5864,
      "step": 6
    },
    {
      "epoch": 0.0003766005523474768,
      "eval_loss": 5.816605567932129,
      "eval_runtime": 1725.3472,
      "eval_samples_per_second": 3.888,
      "eval_steps_per_second": 1.945,
      "step": 6
    },
    {
      "epoch": 0.0004393673110720562,
      "grad_norm": 7.574453353881836,
      "learning_rate": 0.00014,
      "loss": 5.0298,
      "step": 7
    },
    {
      "epoch": 0.0004393673110720562,
      "eval_loss": 5.065299034118652,
      "eval_runtime": 1724.5033,
      "eval_samples_per_second": 3.89,
      "eval_steps_per_second": 1.945,
      "step": 7
    },
    {
      "epoch": 0.0005021340697966357,
      "grad_norm": 8.677300453186035,
      "learning_rate": 0.00016,
      "loss": 4.878,
      "step": 8
    },
    {
      "epoch": 0.0005021340697966357,
      "eval_loss": 4.18253755569458,
      "eval_runtime": 1724.6846,
      "eval_samples_per_second": 3.89,
      "eval_steps_per_second": 1.945,
      "step": 8
    },
    {
      "epoch": 0.0005649008285212152,
      "grad_norm": 8.120089530944824,
      "learning_rate": 0.00018,
      "loss": 4.1115,
      "step": 9
    },
    {
      "epoch": 0.0005649008285212152,
      "eval_loss": 3.3734354972839355,
      "eval_runtime": 1724.4529,
      "eval_samples_per_second": 3.891,
      "eval_steps_per_second": 1.946,
      "step": 9
    },
    {
      "epoch": 0.0006276675872457946,
      "grad_norm": 8.416716575622559,
      "learning_rate": 0.0002,
      "loss": 3.4107,
      "step": 10
    },
    {
      "epoch": 0.0006276675872457946,
      "eval_loss": 2.742058753967285,
      "eval_runtime": 1723.9173,
      "eval_samples_per_second": 3.892,
      "eval_steps_per_second": 1.946,
      "step": 10
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 25,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7398256992583680.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}