Skip to content

Comment on How to run DeepSeek R1 locallyparent

Comments

I haven't played with the distillations extensively, but I think the internal monologue helps the LLM produce a higher quality output, even if its just Qwen or Llama

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.