last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.11323641928079571,
  "eval_steps": 5,
  "global_step": 37,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00306044376434583,
      "eval_loss": 1.1051609516143799,
      "eval_runtime": 26.5057,
      "eval_samples_per_second": 5.206,
      "eval_steps_per_second": 2.603,
      "step": 1
    },
    {
      "epoch": 0.009181331293037491,
      "grad_norm": 0.49553486704826355,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.0599,
      "step": 3
    },
    {
      "epoch": 0.015302218821729151,
      "eval_loss": 1.0945595502853394,
      "eval_runtime": 25.9873,
      "eval_samples_per_second": 5.31,
      "eval_steps_per_second": 2.655,
      "step": 5
    },
    {
      "epoch": 0.018362662586074982,
      "grad_norm": 0.5389537811279297,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.1064,
      "step": 6
    },
    {
      "epoch": 0.02754399387911247,
      "grad_norm": 0.4423370361328125,
      "learning_rate": 0.0001,
      "loss": 0.975,
      "step": 9
    },
    {
      "epoch": 0.030604437643458302,
      "eval_loss": 0.9807908535003662,
      "eval_runtime": 25.9978,
      "eval_samples_per_second": 5.308,
      "eval_steps_per_second": 2.654,
      "step": 10
    },
    {
      "epoch": 0.036725325172149964,
      "grad_norm": 0.6366612315177917,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.9206,
      "step": 12
    },
    {
      "epoch": 0.045906656465187455,
      "grad_norm": 0.6869972944259644,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.8698,
      "step": 15
    },
    {
      "epoch": 0.045906656465187455,
      "eval_loss": 0.8133208751678467,
      "eval_runtime": 25.9999,
      "eval_samples_per_second": 5.308,
      "eval_steps_per_second": 2.654,
      "step": 15
    },
    {
      "epoch": 0.05508798775822494,
      "grad_norm": 0.7443881034851074,
      "learning_rate": 0.0002,
      "loss": 0.7708,
      "step": 18
    },
    {
      "epoch": 0.061208875286916604,
      "eval_loss": 0.7175387144088745,
      "eval_runtime": 26.0285,
      "eval_samples_per_second": 5.302,
      "eval_steps_per_second": 2.651,
      "step": 20
    },
    {
      "epoch": 0.06426931905126243,
      "grad_norm": 0.667968213558197,
      "learning_rate": 0.0001879473751206489,
      "loss": 0.7045,
      "step": 21
    },
    {
      "epoch": 0.07345065034429993,
      "grad_norm": 0.618311882019043,
      "learning_rate": 0.00015469481581224272,
      "loss": 0.6921,
      "step": 24
    },
    {
      "epoch": 0.07651109410864575,
      "eval_loss": 0.6743873953819275,
      "eval_runtime": 26.0301,
      "eval_samples_per_second": 5.302,
      "eval_steps_per_second": 2.651,
      "step": 25
    },
    {
      "epoch": 0.08263198163733741,
      "grad_norm": 0.48745396733283997,
      "learning_rate": 0.00010825793454723325,
      "loss": 0.6098,
      "step": 27
    },
    {
      "epoch": 0.09181331293037491,
      "grad_norm": 0.530785083770752,
      "learning_rate": 5.983045753470308e-05,
      "loss": 0.611,
      "step": 30
    },
    {
      "epoch": 0.09181331293037491,
      "eval_loss": 0.6592618227005005,
      "eval_runtime": 26.2376,
      "eval_samples_per_second": 5.26,
      "eval_steps_per_second": 2.63,
      "step": 30
    },
    {
      "epoch": 0.1009946442234124,
      "grad_norm": 0.5798205733299255,
      "learning_rate": 2.1085949060360654e-05,
      "loss": 0.7136,
      "step": 33
    },
    {
      "epoch": 0.10711553175210406,
      "eval_loss": 0.6530157923698425,
      "eval_runtime": 26.1128,
      "eval_samples_per_second": 5.285,
      "eval_steps_per_second": 2.642,
      "step": 35
    },
    {
      "epoch": 0.11017597551644988,
      "grad_norm": 0.5891212821006775,
      "learning_rate": 1.3638696597277679e-06,
      "loss": 0.6765,
      "step": 36
    }
  ],
  "logging_steps": 3,
  "max_steps": 37,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 18,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3779253648687104e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}