3 ms·
yeah I feel the same, I think even having a screenshot of part of rendered page or full page can be useful even for machines considering how heavy those HTML ca
by pzo 9mo ago
yeah I feel the same, I think even having a screenshot of part of rendered page or full page can be useful even for machines considering how heavy those HTML can be to parse and expensive for LLM context. Sometimes (sub)screenshot is just a better kind of compression
- fbouvier 9mo agoYes HTML is too heavy and too expensive for LLM. We are working on a text-based format more suitable for AI.
- httpteapot 9mo agoWhat do you think of the DeepSeek OCR approach where they say that vision tokens might better compress a document than its pure text representation? https://news.ycombinator.com/item?id=45640594 https://news.ycombinator.com/item?id=45640594 I've spent some time feeding llm with scrapped web pages and I've found that retaining some style information (text size, visibility, decoration image content) is non trivial.
- fbouvier 9mo agoKeeping some kind of style information is definitely important to understand the semantics of the webpage.