README.md

metadata

datasets:
  - aligner/aligner-20K
language:
  - en
tags:
  - safety
  - ai-safety
  - aligner
  - llama

Aligner Model Card

Model details

Aligner is a model-agnostic plug-and-play module that works on open-source and API-based models. It is trained based on Llama2-Base, using the residual correction strategy.

Model Sources

• Repository: https://github.com/Aligner2024/aligner
• Dataset:https://huggingface.co/datasets/aligner/aligner-20K

More Details

• Website:https://aligner2024.github.io/

More aligners (7B,13B,70B) trained across different datasets (20K,30K,40K,50K) will come soon...