Activity
Stars and mentions · Last 30 days, Oct 6 – Nov 4
Top repositories
Dao-AILab’s repositories ranked by new forks, as on trending.
Contributes to
Repositories outside Dao-AILab’s account where Dao-AILab opened or reviewed pull requests in the past 365 days.
Who noticed Dao-AILab
Everyone who starred or forked Dao-AILab’s repositories since the archive began in 2011.
Notable accounts
Ordered by the stars their own repositories have received.
Biggest fans
Accounts that starred the most of these repositories.
Breakout weeks
Each repository’s best week of new stars, with the release published around it.
Achievements
Awards earned by the repositories Dao-AILab owns. Select an award to embed its badge.
No awards yet
Repositories earn awards by reaching the top three in a completed trending day or week.
A repository earns an award for finishing first, second or third in a completed trending day or week, ranked by new stars, new forks or pull request authors. Awards follow a repository through renames and transfers and count toward its current owner. Weeks run Monday to Sunday, UTC.
Mentions
resources: free: the original paper, "flashattention: fast and memory-efficient exact attention with io-awareness" by tri dao et al. arxiv.org free: the official code from tri dao's lab Dao-AILab/flash-attention
@LunixA380 @osanseviero That is correct, simple architecture especially for the smaller models. Would be also nice if the attention head size be compatible with Flash Attention. Related: flash-attention#2427
@maharshii Many primitives in Dao-AILab/quack
7/8 Attention kernels FlashAttention (Tri Dao). The official repository of FlashAttention, used inside nearly every training and serving system. The README links to the papers and blog posts for each version. Dao-AILab/flash-attention t.co
@realcezarc @StasBekman Maybe a better way to test is to find a kernel that supports both mxfp4 and nvfp4 so that the only things that really change between the two are the formats and not the kernel implementation details themselves. I think Quack supports both? Dao-AILab/quack