You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
vladmandic
changed the title
[Issue]: Crash when loading the FLUX.1-dev model on AMD GPU
[Issue]: Crash when loading the FLUX.1 with ZLUDA
Dec 18, 2024
Hello,
I’ve searched for solutions on my side but couldn’t find anything. Do you have any idea where the problem might come from on your end?
Thank you in advance.
Issue Description
System Configuration:
System: Windows
Python: 3.10.11
Git version: 2.47.1.windows.1
CPU: AMD Ryzen 7 7800X3D
GPU: AMD Radeon RX 7900 XTX (24GB VRAM)
RAM: 32 GB
GPU Drivers: Adrenalin version 24.12.1
Installation Steps:
webui.bat --use-zluda --debug --autolaunch
Steps to Reproduce the Issue:
webui.bat --use-zluda --debug --autolaunch
Expected Behavior:
The FLUX.1-dev model should load correctly.
Actual Behavior:
The application closes immediately without an explicit error message.
Additional Information:
Version Platform Description
D:\Desktop\automatic>webui.bat --use-zluda --debug --autolaunch
Using VENV: D:\Desktop\automatic\venv
00:22:00-932979 INFO Starting SD.Next
00:22:00-935977 INFO Logger: file="D:\Desktop\automatic\sdnext.log" level=DEBUG size=65 mode=create
00:22:00-937978 INFO Python: version=3.10.11 platform=Windows bin="D:\Desktop\automatic\venv\Scripts\python.exe" venv="D:\Desktop\automatic\venv"
00:22:01-048454 INFO Version: app=sd.next updated=2024-12-09 hash=a2f5a39a branch=master url=https://github.com/vladmandic/automatic/tree/master ui=main
00:22:01-350463 INFO Platform: arch=AMD64 cpu=AMD64 Family 25 Model 97 Stepping 2, AuthenticAMD system=Windows release=Windows-10-10.0.22631-SP0 python=3.10.11 docker=False
00:22:01-352687 DEBUG Packages: venv=venv site=['venv', 'venv\lib\site-packages']
00:22:01-354095 INFO Args: ['--use-zluda', '--debug', '--autolaunch']
00:22:01-356104 DEBUG Setting environment tuning
00:22:01-357105 DEBUG Torch allocator: "garbage_collection_threshold:0.80,max_split_size_mb:512"
00:22:01-360104 DEBUG Torch overrides: cuda=False rocm=False ipex=False directml=False openvino=False zluda=True
00:22:01-365109 INFO Python: version=3.10.11 platform=Windows bin="D:\Desktop\automatic\venv\Scripts\python.exe" venv="D:\Desktop\automatic\venv"
00:22:01-368105 INFO ROCm: AMD toolkit detected
00:22:02-096492 INFO ROCm: agents=['gfx1100', 'gfx1036']
00:22:02-098337 INFO ROCm: version=6.1, using agent gfx1100
00:22:02-132403 INFO Using ZLUDA in D:\Desktop\automatic.zluda
00:22:03-250939 INFO Install: verifying requirements
00:22:03-254634 INFO Verifying packages
00:22:03-285213 DEBUG Timestamp repository update time: Mon Dec 9 09:27:14 2024
00:22:03-286213 INFO Startup: standard
00:22:03-287213 INFO Verifying submodules
00:22:04-910392 DEBUG Git submodule: extensions-builtin/sd-extension-chainner / main
00:22:04-982394 DEBUG Git detached head detected: folder="extensions-builtin/sd-extension-system-info" reattach=main
00:22:04-984405 DEBUG Git submodule: extensions-builtin/sd-extension-system-info / main
00:22:05-032404 DEBUG Git submodule: extensions-builtin/sd-webui-agent-scheduler / main
00:22:05-101185 DEBUG Git detached head detected: folder="extensions-builtin/sdnext-modernui" reattach=main
00:22:05-102713 DEBUG Git submodule: extensions-builtin/sdnext-modernui / main
00:22:05-150886 DEBUG Git submodule: extensions-builtin/stable-diffusion-webui-rembg / master
00:22:05-199891 DEBUG Git submodule: modules/k-diffusion / master
00:22:05-266457 DEBUG Git detached head detected: folder="wiki" reattach=master
00:22:05-268466 DEBUG Git submodule: wiki / master
00:22:05-301342 DEBUG Register paths
00:22:05-331340 DEBUG Installed packages: 182
00:22:05-332065 DEBUG Extensions all: ['Lora', 'sd-extension-chainner', 'sd-extension-system-info', 'sd-webui-agent-scheduler', 'sdnext-modernui', 'stable-diffusion-webui-rembg']
00:22:05-422078 DEBUG Extension installer: D:\Desktop\automatic\extensions-builtin\sd-webui-agent-scheduler\install.py
00:22:06-829595 DEBUG Extension installer: D:\Desktop\automatic\extensions-builtin\stable-diffusion-webui-rembg\install.py
00:22:10-958678 DEBUG Extensions all: []
00:22:10-960676 INFO Extensions enabled: ['Lora', 'sd-extension-chainner', 'sd-extension-system-info', 'sd-webui-agent-scheduler', 'sdnext-modernui', 'stable-diffusion-webui-rembg']
00:22:10-962297 INFO Install: verifying requirements
00:22:10-963197 DEBUG Setup complete without errors: 1734477731
00:22:10-966209 DEBUG Extension preload: {'extensions-builtin': 0.0, 'extensions': 0.0}
00:22:10-967205 INFO Command line args: ['--use-zluda', '--debug', '--autolaunch'] autolaunch=True use_zluda=True debug=True
00:22:10-969205 DEBUG Env flags: []
00:22:10-970204 DEBUG Starting module: <module 'webui' from 'D:\Desktop\automatic\webui.py'>
00:22:13-490536 INFO Torch: torch==2.3.1+cu118 torchvision==0.18.1+cu118
00:22:13-494532 INFO Packages: diffusers==0.32.0.dev0 transformers==4.46.2 accelerate==1.1.1 gradio==3.43.2
00:22:13-784146 DEBUG Huggingface cache: folder="C:\Users\Max.cache\huggingface\hub"
00:22:14-164583 INFO Device detect: memory=24.0 optimization=none
00:22:14-166582 DEBUG Read: file="config.json" json=38 bytes=1710 time=0.000
00:22:14-268675 INFO Engine: backend=Backend.DIFFUSERS compute=zluda device=cuda attention="Dynamic Attention BMM" mode=no_grad
00:22:14-270674 DEBUG Read: file="html\reference.json" json=59 bytes=31585 time=0.000
00:22:14-388037 INFO Torch parameters: backend=zluda device=cuda config=Auto dtype=torch.bfloat16 vae=torch.bfloat16 unet=torch.bfloat16 context=no_grad nohalf=False nohalfvae=False upscast=False deterministic=False test-fp16=True test-bf16=True optimization="Dynamic Attention BMM"
00:22:14-612728 DEBUG ONNX: version=1.20.1 provider=CPUExecutionProvider, available=['AzureExecutionProvider', 'CPUExecutionProvider']
00:22:14-656072 INFO Device: device=AMD Radeon RX 7900 XTX [ZLUDA] n=1 arch=compute_37 capability=(8, 8) cuda=11.8 cudnn=8700 driver=
Relevant log output
Backend
Diffusers
UI
Standard
Branch
Master
Model
Other
Acknowledgements
The text was updated successfully, but these errors were encountered: