4 ms·
A few publically shared evaluation datasets include: - HuggingFace Evaluate (and their 3 dashboards) - OpenAIEval - AnthropicEval
by jacky2wong 3y ago
A few publically shared evaluation datasets include:
- HuggingFace Evaluate (and their 3 dashboards)
- OpenAIEval
- AnthropicEval
- intended 3y agoI would not use those for production. I would assume your backend model is already tested on these. You would need a novel data set.