‼️ If you want to work on this issue: please comment below and wait until a maintainer assigns this issue to you before opening a PR to avoid several contributions on the same issue. Thanks! 😊

✨ What You’ll Do

Right now, we have quantizers like Quanto, torchao or HQQ. It would be nice to integrate GANQ as a new quantizer. The final interface to apply the quantization method should look as follows:

model = ...

smash_config = SmashConfig()
smash_config["algo_group_1"] = "algo_1"
smash_config["algo_group_2"] = "algo_2"
smashed_model = smash(model=model, smash_config=smash_config)

🤖 Useful Resources

• Check the detailed contributions guidelines with steps to integrate a new algorithm
• Check the paper and github repo
• Developp the integration.
• Test it and write the documentation.

✅ Acceptance Criteria

• Tests & Docs: All existing and new unit tests pass, and the documentation is updated

And don’t forget to give us a ⭐️!

❓ Questions?

Feel free to jump into the #contributing Discord channel if you hit any roadblocks. Can’t wait to see your contribution! 🚀

Share on Socials