Hacker News new | ask | show | jobs
by mluo 488 days ago
For quantization, very big impact for small models, can drop at much as 10% on AIME. Our model does best on bfloat16 ;)

Come checkout our repo at: https://github.com/agentica-project/deepscaler