Search results

[How-to]How to get Flash-Attention under windows 11 CUDA

Status: Open.

#1469 In Dao-AILab/flash-attention;

· mytait opened on Jan 30, 2025

FA3 package is missing padding utilities

Status: Open.

#1467 In Dao-AILab/flash-attention;

· tmm1 opened on Jan 28, 2025

What is seqused_q and seqused_k?

Status: Open.

#1466 In Dao-AILab/flash-attention;

· cassanof opened on Jan 27, 2025

FA3 KV Cache is slower than FA2 KV Cache

Status: Open.

#1465 In Dao-AILab/flash-attention;

· DD-DuDa opened on Jan 26, 2025

Add support for Cuda 12.8 and B200 GPUs

Status: Open.

#1464 In Dao-AILab/flash-attention;

· ofirkris opened on Jan 25, 2025

fused dense lib warning

Status: Open.

#1462 In Dao-AILab/flash-attention;

· YuyueminAustin opened on Jan 24, 2025

BUG? get the wrong value when logit_scale is 0

Status: Open.

#1461 In Dao-AILab/flash-attention;

· shunshen93 opened on Jan 24, 2025

Conflict When Installing flash-attn 2.7.3 and 3.0.0b1 Together

Status: Open.

#1459 In Dao-AILab/flash-attention;

· quanta42 opened on Jan 24, 2025

using Flash Attention version 2.5.7, upgraded CUTLASS to version 3.5 then encountered the following compilation error.

Status: Open.

#1458 In Dao-AILab/flash-attention;

· ccccjunkang opened on Jan 23, 2025

Main branch compilation on nvcc 12.6

Status: Open.

#1453 In Dao-AILab/flash-attention;

· roded2 opened on Jan 21, 2025

v2.7.3 build failed in NGC pytorch:24.12-py3

Status: Open.

#1452 In Dao-AILab/flash-attention;

· xuchunmei000 opened on Jan 21, 2025

FA3 consecutive failing tests after first failure

Status: Open.

#1451 In Dao-AILab/flash-attention;

· benjamin-kroeger opened on Jan 21, 2025