lllyasviel / lllyasviel/stable-diffusion-webui-forge
Fresh Forge Install Torch 2.4.0 and CUDA 2.4 failing to launch after installing Adetailer
- Dominant language
- Python
- Stars
- 13k
- Forks
- 1.7k
- PR merge metrics
- No merged PRs in 30d
Description
Hello All!
Longtime user of A111 and Forge and I wanted to get it update to try and get it to be faster, even wanted to figure out how to get Triton working. However, after a fresh Forge install I went through the GUI and installed Adetailer and subsequently restarted Forge as directed. Now I am getting this error when trying to run the .bat file where before it was working fine:
```
Python 3.10.6 (tags/v3.10.6:9c7b4bd, Aug 1 2022, 21:53:49) [MSC v.1932 64 bit (AMD64)]
Version: f2.0.1v1.10.1-previous-633-ge073e4ec
Commit hash: e073e4ec581c803cbc71003f6d3261d37ec43840
Traceback (most recent call last):
File "E:\Stable Diffusion\FORGE 2.0\webui_forge_cu124_torch24\webui\launch.py", line 54, in
main()
File "E:\Stable Diffusion\FORGE 2.0\webui_forge_cu124_torch24\webui\launch.py", line 42, in main
prepare_environment()
File "E:\Stable Diffusion\FORGE 2.0\webui_forge_cu124_torch24\webui\modules\launch_utils.py", line 436, in prepare_environment
raise RuntimeError(
RuntimeError: Your device does not support the current version of Torch/CUDA! Consider download another version:
https://github.com/lllyasviel/stable-diffusion-webui-forge/releases/tag/latest
Press any key to continue . . .
```
EDIT: I did a fresh install with "CUDA 12.1 + Pytorch 2.3.1 <- Recommended" and performed the exact same steps and it works just fine. Adetailer installed via GUI, performed the reset, and then the GUI reloaded as normal. I am able to generate images like before. It appears this is isolated to "Forge with CUDA 12.4 + Pytorch 2.4"
Contributor guide
No contributing guide indexed for this repository
Research direction
Start with webui/modules/launch_utils.py around prepare_environment at line 436 and webui/launch.py. Reproduce the failure with CUDA 12.4 and PyTorch 2.4 after installing Adetailer, then compare it with the working CUDA 12.1 and PyTorch 2.3.1 setup. Done means the affected installation launches successfully after the same steps.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 30/100