Skip to content

Issues: Dao-AILab/flash-attention

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

646 Open 567 Closed
646 Open 567 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

FA3 KV Cache is slower than FA2 KV Cache
#1465 opened Jan 26, 2025 by DD-DuDa
2
Add support for Cuda 12.8 and B200 GPUs
#1464 opened Jan 25, 2025 by ofirkris
fused dense lib warning
#1462 opened Jan 24, 2025 by YuyueminAustin
BUG? get the wrong value when logit_scale is 0
#1461 opened Jan 24, 2025 by shunshen93
Conflict When Installing flash-attn 2.7.3 and 3.0.0b1 Together
#1459 opened Jan 24, 2025 by quanta42
using Flash Attention version 2.5.7, upgraded CUTLASS to version 3.5 then encountered the following compilation error.
#1458 opened Jan 23, 2025 by ccccjunkang
dropout_layer_norm
#1456 opened Jan 21, 2025 by ADiko1997
Main branch compilation on nvcc 12.6
#1453 opened Jan 21, 2025 by roded2
2
v2.7.3 build failed in NGC pytorch:24.12-py3
#1452 opened Jan 21, 2025 by xuchunmei000
2
FA3 consecutive failing tests after first failure
#1451 opened Jan 21, 2025 by benjamin-kroeger
BUG？ static_assert(!(!Mma1_is_RS && !IntraWGOverlap), "Mma1 must be RS if IntraWGOverlap is enabled");
#1450 opened Jan 20, 2025 by ziyuhuang123
[QST] masking steps in flash decoding
#1449 opened Jan 17, 2025 by aws-jiadingg
1
Clarification on MMA0 Results Handling in the Latest Code
#1448 opened Jan 16, 2025 by ziyuhuang123
1
subprocess.CalledProcessError: Command '['path/to/cuda-11.7/bin/nvcc', '-V']' returned non-zero exit status 255
#1447 opened Jan 16, 2025 by ChosenOne-xx
Wheel names and version inconsitency.
#1444 opened Jan 14, 2025 by sfc-gh-mhazy
2
Error when importing dropout_layer_norm
#1441 opened Jan 13, 2025 by anfortas337
1
Running flash_attn/flash_attn_triton_amd/bench.py with sequence length > 4096 causes RuntimeError: Triton Error [CUDA]: an illegal memory access was encountered
#1440 opened Jan 12, 2025 by jiqimaoke
5
IncompatibleTypeErrorImpl('invalid operands of type pointer<int64> and triton.language.int32')
#1439 opened Jan 11, 2025 by wuyouliaoxi
FA3 forward performance regression on H200
#1438 opened Jan 10, 2025 by complexfilter
7
FA3 does not work with torch.compile
#1435 opened Jan 10, 2025 by nighting0le01
FA3 regression on H100 80GB?
#1432 opened Jan 9, 2025 by bastianhagedorn
8
[flash attn v2] Why V uses no-swizzle layout for registers?
#1429 opened Jan 8, 2025 by phantaurus
1
version `GLIBCXX_3.4.29' not found
#1428 opened Jan 8, 2025 by zhanghanxing2022
ERROR: No matching distribution found for flash-attn==2.6.3+cu123torch2.4cxx11abifalse
#1423 opened Jan 6, 2025 by carolynsoo
1
Unable to install flash_attn on H100 with CUDA 12.5
#1422 opened Jan 6, 2025 by ghadiaravi13
ProTip! Exclude everything labeled bug with -label:bug.