I’ve often wanted to pair a smart orchestrator / architect model with a super fast worker model. Diffusion LLM could fill the worker role
Tri Dao suggests diffusion LLMs as fast worker models
AISummary
Tri Dao says he has wanted to pair a smart orchestrator or architect model with a very fast worker model, and suggests a diffusion LLM could fill the worker role. The post cites Stefano Ermon's Mercury 3 release, which Ermon says runs at 1000+ tokens per second.
Post on XView on X
Tri DaoVerified on X
@tri_dao
1/4 Today, we’re releasing Mercury 3. Mercury 3 is by far our most powerful model and the most intelligent diffusion LLM ever released. It outperforms GPT-6 Luna and Haiku 5.5 on quality while running at 1000+ tok/sec.
Source: Tri Dao · x.comPublished · added here