AI/diffusers
1
0
Fork 0
mirror of https://github.com/huggingface/diffusers.git synced 2026-10-26 08:59:24 +03:00
Code Activity
6,986 Commits 932 Branches 101 Tags
diffusers-serve
Commit Graph

3 Commits

Author SHA1 Message Date
Wang, Yi
839af3a043 Update FlashAttention Hub integration and guard CUDA-only tests (#15024) 
* Update FlashAttention Hub integration and guard CUDA-only tests

Signed-off-by: Wang, Yi <yi.a.wang@intel.com>

* fmt

Signed-off-by: Wang, Yi <yi.a.wang@intel.com>

---------

Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
 2026-10-14 14:05:19 +05:30
Shivam Shrirao
80c7ed262a Fix flash/sage varlen prep under torch.compile with dynamic shapes (#14568) 
* Fix flash/sage varlen prep under torch.compile with dynamic shapes

* Handle zero-length sequences in varlen offsets

* Add varlen dynamic compile test

* Remove private helper test
 2026-09-20 11:21:20 +05:30
Sayak Paul
192cf685ec [distributed] fix corrupted gradient problem under ring CP. (#14274) 
* fix corrupted gradient problem under ring CP.

* address dhruv's comments.

* handle torch version

* varlens too
 2026-08-20 18:16:17 +05:30