Skip to content

Conversation

@coyotte508
Copy link
Member

@coyotte508 coyotte508 commented Jan 7, 2026

That makes xetchunk-wasm's chunking performance go from 5MB/s to 110MB/s... And that's after switching from a pure assembly script solution to a mix of wasm, blake3-jit and local TS code.

The performance from the rust-generated WASM is 268MB/s so will keep using that for now. Even though it's probably possible to take example from blake3-jit to match the performance.

Fix #1898

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

@huggingface/blake3-wasm is slow

2 participants