Safetensors
Not-For-All-Audiences
xzuyn's picture
Update README.md
411ef3d verified
metadata
datasets:
  - >-
    PJMixers/NobodyExistsOnTheInternet_ToxicQAFinal-L3-Instruct-8B-PreferenceShareGPT
  - NobodyExistsOnTheInternet/ToxicQAFinal
tags:
  - not-for-all-audiences

Trained on NobodyExistsOnTheInternet/ToxicQAFinal. I converted the set to a preference dataset using refusals generated from LLaMa-3-Instruct-8B.

train/rewards train/logits train/logps train