metadata
license: apache-2.0
language:
- en
DPO from the model ranked 6th on the overall leaderboard and 1st in the 7B leaderboard - v1olet/v1olet_marcoroni-go-bruins-merge-7B.
Training data: comparison_gpt4_en,en_orca_dpo
Train for 1 epoch.
You can use alpaca template.
template_format = """{system}
### Instruction:
{prompt}
### Response:
"""
Developed by: Trong-Hieu Nguyen-Mau