Comment on DeepSeek's distilled new R1 AI model can run on a single GPU | TechCrunch

<- View Parent
vhstape@lemmy.sdf.org ⁨2⁩ ⁨weeks⁩ ago

On my Mac mini running LM Studio, it managed 1702 tokens at 17.19 tok/sec and thought for 1 minute

source
Sort:hotnewtop