last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.02336448598130841,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002336448598130841,
      "grad_norm": 0.5417346954345703,
      "learning_rate": 2e-05,
      "loss": 1.7637,
      "step": 1
    },
    {
      "epoch": 0.002336448598130841,
      "eval_loss": 1.971550464630127,
      "eval_runtime": 6.0193,
      "eval_samples_per_second": 119.781,
      "eval_steps_per_second": 15.118,
      "step": 1
    },
    {
      "epoch": 0.004672897196261682,
      "grad_norm": 0.5608940124511719,
      "learning_rate": 4e-05,
      "loss": 2.0274,
      "step": 2
    },
    {
      "epoch": 0.007009345794392523,
      "grad_norm": 0.7115104794502258,
      "learning_rate": 6e-05,
      "loss": 2.1016,
      "step": 3
    },
    {
      "epoch": 0.007009345794392523,
      "eval_loss": 1.9603239297866821,
      "eval_runtime": 6.0569,
      "eval_samples_per_second": 119.038,
      "eval_steps_per_second": 15.024,
      "step": 3
    },
    {
      "epoch": 0.009345794392523364,
      "grad_norm": 0.5761290788650513,
      "learning_rate": 8e-05,
      "loss": 1.607,
      "step": 4
    },
    {
      "epoch": 0.011682242990654205,
      "grad_norm": NaN,
      "learning_rate": 8e-05,
      "loss": 1.8908,
      "step": 5
    },
    {
      "epoch": 0.014018691588785047,
      "grad_norm": 0.5327461957931519,
      "learning_rate": 0.0001,
      "loss": 1.5124,
      "step": 6
    },
    {
      "epoch": 0.014018691588785047,
      "eval_loss": 1.9273180961608887,
      "eval_runtime": 6.098,
      "eval_samples_per_second": 118.235,
      "eval_steps_per_second": 14.923,
      "step": 6
    },
    {
      "epoch": 0.016355140186915886,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 2.6024,
      "step": 7
    },
    {
      "epoch": 0.018691588785046728,
      "grad_norm": 0.9092730283737183,
      "learning_rate": 0.00012,
      "loss": 2.1422,
      "step": 8
    },
    {
      "epoch": 0.02102803738317757,
      "grad_norm": 1.0777069330215454,
      "learning_rate": 0.00014,
      "loss": 2.1089,
      "step": 9
    },
    {
      "epoch": 0.02102803738317757,
      "eval_loss": 1.878110647201538,
      "eval_runtime": 6.0972,
      "eval_samples_per_second": 118.252,
      "eval_steps_per_second": 14.925,
      "step": 9
    },
    {
      "epoch": 0.02336448598130841,
      "grad_norm": 0.48678022623062134,
      "learning_rate": 0.00016,
      "loss": 1.6706,
      "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": 8234950176276480.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}