alexwb's picture
End of training
6b77313 verified
metadata
license: other
base_model: facebook/opt-350m
tags:
  - trl
  - reward-trainer
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: reward_modeling_anthropic_hh_rm1e-6
    results: []

reward_modeling_anthropic_hh_rm1e-6

This model is a fine-tuned version of facebook/opt-350m on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6426
  • Accuracy: 0.6277

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 1e-06
  • train_batch_size: 16
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 1.0

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.6812 0.1087 500 0.6965 0.5592
0.6554 0.2174 1000 0.6798 0.5765
0.6715 0.3262 1500 0.6699 0.5850
0.6401 0.4349 2000 0.6635 0.5932
0.6741 0.5436 2500 0.6555 0.6093
0.653 0.6523 3000 0.6504 0.6185
0.6446 0.7610 3500 0.6456 0.6227
0.6449 0.8698 4000 0.6434 0.6250
0.6597 0.9785 4500 0.6426 0.6277

Framework versions

  • Transformers 4.40.2
  • Pytorch 2.4.0+cu121
  • Datasets 2.19.1
  • Tokenizers 0.19.1