CNSeniorious000 / free-chat

An elegant LLM chat UI forked from chatgpt-demo of @anse-app. Index site at https://free-chat.asia
https://endless-chat.vercel.app
MIT License
177 stars 51 forks source link

Improve performance by rendering markdown partially #68

Open CNSeniorious000 opened 1 month ago

CNSeniorious000 commented 1 month ago

Use MarkdownIt.parse to implement this.

linear[bot] commented 1 month ago

PROM-31 Improve performance by rendering markdown partially

CNSeniorious000 commented 1 month ago

Or we can try using another markdown parser like marked.

This is because MarkdownIt.parse's result is somehow not ready to render: it produces a list of tokens, but we want a nested tree. Although transforming that should be easy enough, it may hurt performance.