lllyasviel / lllyasviel/FramePack
RTX 50 Series Install Sage Attention
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 17.3k
- Forks
- 1.7k
- PR merge metrics
- No merged PRs in 30d
Description
Hi, I've got Framepack and F1 working for RTX 5070ti using the right pytorch version
1. .\python.exe -m pip install torch torchvision torchaudio --force-reinstall --index-url https://download.pytorch.org/whl/cu128
2. .\python.exe -m pip install --force-reinstall numpy==1.26.2
3. .\python.exe -m pip install flatbuffers
Now I'm trying to use install Sage Attention (I believe the review is Xformer and Flashattention is not as good as Sage attention).
But I've yet to get it to work as I lack the expertise knowledge.
Using the Package Installer (https://github.com/lllyasviel/FramePack/issues/138) doesn't work as I believe its meant for CU12.6, I tried some weird method such as reinstalling Pytorch 2.7 and it doesn't work as well
I hope there are some experts that can run me through on how to install Sage Attention on RTX 50 series using Pytorch 2.7 correctly.
Appreciate any help that I can get cause I cant really find a suitable solution for my case, thanks!
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
The issue names no repository files or tests; start by reviewing the Package Installer from linked issue #138 and reproducing the listed PyTorch and NumPy installation steps. Done means identifying a working Sage Attention installation path for an RTX 5070 Ti with PyTorch 2.7, or documenting the specific incompatibility.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning, performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100