OPENGIT/stable-diffusion-webui
1
0
Fork 0
mirror of https://github.com/AUTOMATIC1111/stable-diffusion-webui.git synced 2024-12-29 19:05:05 +08:00
Code Issues Packages Projects Releases Wiki Activity

Updated Xformers (markdown)

AUTOMATIC1111 2022-10-08 19:53:14 +03:00
parent 03cfb4c5d5
commit 468ec85373
1 changed files with 15 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
Xformers.md

@ -1,24 +1,23 @@
Xformers library is an optional way to speedup your image generation.
There are not binaries for Windows except for one specific configuration, but you can build it yourself. A guide from an anonymous user:
There are no binaries for Windows except for one specific configuration, but you can build it yourself.
A guide from an anonymous user, although I think it is for building on Linux:
GUIDE ON HOW TO BUILD XFORMERS
also includes how to uncuck yourself from sm86 restriction on voldy's new commit
1. go to the webui directory
2. source ./venv/bin/activate
3. cd repositories
3. git clone https://github.com/facebookresearch/xformers.git
4. cd xformers
5. git submodule update --init --recursive
6. pip install -r requirements.txt
7. pip install -e .
1. go to the webui directory
2. `source ./venv/bin/activate`
3. `cd repositories`
3. `git clone https://github.com/facebookresearch/xformers.git`
4. `cd xformers`
5. `git submodule update --init --recursive`
6. `pip install -r requirements.txt`
7. `pip install -e .`
If you encounter some error about torch not being built with your cuda version blah blah, then try:
pip install setuptools==49.6.0
`pip install setuptools==49.6.0`
After step 7, just wait like 30 minutes for everything to build and you're done.
Then go back to modules/sd_[hijack.py](http://hijack.py/) and search for this text and delete it:
```python
and torch.cuda.get_device_capability(shared.device) == (8, 6)
```
8. wait like 30 minutes for everything to build
9. launch with `--force-enable-xformers` commandline argument