CUTLASS 3.0 is now available! · Discussion #787 · NVIDIA/cutlass · GitHub - discu.eu

Linking pages

FlashAttention 2: making Transformers 800% faster w/o approximation - with Tri Dao of Together AI https://www.latent.space/p/flashattention 0 comments

Related searches:

Search whole site: site:github.com

Search title: CUTLASS 3.0 is now available! · Discussion #787 · NVIDIA/cutlass · GitHub

See how to search.

Submit link to: