Add ONNX weights (+transformers.js support)

#2
by Xenova HF staff - opened

Hi there! First of all, really great work with the model... I love seeing the MTEB leaderboard being switched up.

This PR adds ONNX weights for the model, both full-precision (fp32; model.onnx) as well as 8-bit precision (int8; model_quantized.onnx), so that it can be used with Transformers.js. Example usage can be found here.

Xenova changed pull request title from Add ONNX weights to Add ONNX weights (+transformers.js support)

I also added the transformers.js tag for discoverability.

WhereIsAI org

It is great! Thanks for your contribution!

SeanLee97 changed pull request status to merged

Thanks! I’ll open up another PR so it shows both transformers and transformers.js library tags.

Sign up or log in to comment