The only way to do speech-to-speech with an open source model is this Hugging Face demo

So I think i'll make a better one

The only way to do speech-to-speech with an open source model is this Hugging Face demo

So I think i'll make a better one

https://huggingface.co/spaces/Qwen/Qwen2.5-Omni-7B-Demo

Nice. If you’re going to deploy it on @base and use a qwen variant fill out the @flock-io form on Base Services Hub to get half off qwen tokens. If you end up fill out the form DC me and I’ll get your application prioritized.

Decentralized open source consciousness | Values-aligned communities | citizen of Black Rock City

a jiraffe on the internet • nf.td/morereese

Totally depends on what you’re trying to use the model for but some initial ideas:
-fine tuning competition on Base project for model training
-store model weights on Base
-integrate with existing web3 agents on Base for voice controlled dApps
-async voice-based DAO proposal discussions with multi-lingual support

What part of it do you think should go on Base?

Nice. If you’re going to deploy it on @base.base.eth and use a qwen variant fill out the @flock-io form on Base Services Hub to get half off qwen tokens. If you end up fill out the form DC me and I’ll get your application prioritized. 

https://docs.base.org/get-started/base-services-hub

https://docs.google.com/forms/d/e/1FAIpQLSdAi_RBUVaBUakg3yJdpnO_uQhj5p2lY4U7jStpPXAIva0TnQ/viewform