nodes/hao-ai-lab-FastVideo
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

[bugfix]: install the FA4 CUDA 13 runtime

Browse Source 
Select flash-attn-4's cu13 extra on Blackwell install surfaces so revision 82d6441 receives its exact CuTe DSL 4.6.0.dev0 runtime. Enable the opt-in overlay for arm64 CUDA 13 images, including GB200, while CUDA 12.6 continues to use FA3/FA2.
This commit is contained in:
SolitaryThinker
2026-07-21 04:33:39 -07:00
parent 4943d43fb9
commit 8f670ce294
6 changed files with 41 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/inference/optimizations.md
+2 -2
View File
@@ -78,8 +78,8 @@ python setup.py install
FastVideo never auto-selects FlashAttention-4 (`flash_attn.cute`) just because it
is installed: its CuTeDSL kernels JIT-compile per shape family and can fail at
runtime on some GPU/shape combinations. To use FA4, install the pinned
`flash-attn-4` build (see the `flash-attn-4` source in `pyproject.toml`) and set:
runtime on some GPU/shape combinations. To use FA4 on CUDA 13, install the pinned
`flash-attn-4[cu13]` build (see the `flash-attn-4` source in `pyproject.toml`) and set:
```bash
export FASTVIDEO_FA4=1