[Media] fast-h2m: High-performance HTML to Markdown converter
![[Media] fast-h2m: High-performance HTML to Markdown converter](https://i.redd.it/iowsym9hr75h1.jpeg)
It is a converter from HTML to Markdown (what LLMs eating). Orignally from https://crates.io/crates/html-to-markdown-rs which is a vibe-coded repo with some of inefficient decisions inside. My fork is mostly a clean up.
To make this library even faster I also forked astral-tl to make zero-copy parsing and enhance that code to latest Rust.
- It has python 3.8+ bindings under the same name on PyPI.
- Library supports SIMD if you really need gain of 5-10% in parsing