Comment by siwakotisaurav

You’re using slightly improved qwen and llama not r1. R1 only has 600b model