Is your summer reading list still empty? Curious if an LLM can generate a book blurb you'd enjoy and help build a KTO preference dataset at the same time?

A demo using @huggingface Spaces and @gradio to collect LLM output preferences: https://huggingface.co/spaces/davanstrien/would-you-read-it

Would You Read It - a Hugging Face Space by davanstrien

Discover amazing ML apps made by the community

@danielvanstrien hmm, i feel the same way i do when i rate images on Midjourney -- concerned about the criteria. In some cases, I would read that kind of content, but not in the style of the LLMs text blurb, which is almost always awful. On Amazon that's a sign the book text itself is probably badly written.
@arnicas Occasionally the books sounds interesting but often the blurbs are not very good. Think LLMs are still very lacking in this kind of task tbh.
@danielvanstrien they're uniformly awful text stylists