last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0001391715811631265,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.391715811631265e-05,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 4.5194,
      "step": 1
    },
    {
      "epoch": 1.391715811631265e-05,
      "eval_loss": 5.923115253448486,
      "eval_runtime": 1534.7897,
      "eval_samples_per_second": 9.857,
      "eval_steps_per_second": 9.857,
      "step": 1
    },
    {
      "epoch": 2.78343162326253e-05,
      "grad_norm": 3.8339357376098633,
      "learning_rate": 2e-05,
      "loss": 6.9966,
      "step": 2
    },
    {
      "epoch": 4.1751474348937944e-05,
      "grad_norm": 1.9755796194076538,
      "learning_rate": 4e-05,
      "loss": 3.8321,
      "step": 3
    },
    {
      "epoch": 4.1751474348937944e-05,
      "eval_loss": 5.91879415512085,
      "eval_runtime": 1532.6296,
      "eval_samples_per_second": 9.871,
      "eval_steps_per_second": 9.871,
      "step": 3
    },
    {
      "epoch": 5.56686324652506e-05,
      "grad_norm": 4.169151782989502,
      "learning_rate": 6e-05,
      "loss": 5.5239,
      "step": 4
    },
    {
      "epoch": 6.958579058156325e-05,
      "grad_norm": NaN,
      "learning_rate": 6e-05,
      "loss": 5.6624,
      "step": 5
    },
    {
      "epoch": 8.350294869787589e-05,
      "grad_norm": 2.504304885864258,
      "learning_rate": 8e-05,
      "loss": 3.8521,
      "step": 6
    },
    {
      "epoch": 8.350294869787589e-05,
      "eval_loss": 5.870975494384766,
      "eval_runtime": 1533.6973,
      "eval_samples_per_second": 9.864,
      "eval_steps_per_second": 9.864,
      "step": 6
    },
    {
      "epoch": 9.742010681418854e-05,
      "grad_norm": 2.4749975204467773,
      "learning_rate": 0.0001,
      "loss": 3.6626,
      "step": 7
    },
    {
      "epoch": 0.0001113372649305012,
      "grad_norm": 3.602562427520752,
      "learning_rate": 0.00012,
      "loss": 5.4928,
      "step": 8
    },
    {
      "epoch": 0.00012525442304681385,
      "grad_norm": 4.245274066925049,
      "learning_rate": 0.00014,
      "loss": 3.3841,
      "step": 9
    },
    {
      "epoch": 0.00012525442304681385,
      "eval_loss": 5.604666233062744,
      "eval_runtime": 1537.609,
      "eval_samples_per_second": 9.839,
      "eval_steps_per_second": 9.839,
      "step": 9
    },
    {
      "epoch": 0.0001391715811631265,
      "grad_norm": NaN,
      "learning_rate": 0.00014,
      "loss": 6.5318,
      "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": 3173290868736000.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}