Installation Issue: xFormers wasn't build with CUDA support #2505

SamSilmarilData · 2023-11-12T23:59:10Z

SamSilmarilData
Nov 12, 2023

Hello,
i have Cuda 11.8, python version 3.10.13.
i have noticed that whenever i do a fresh install, i have to use --use-xformers, so that i get the pytorch cu118 version.
the install output looks like this:

Click to view the code

./webui.sh --share --listen --use-xformers --debug --insecure 
Create and activate python venv
Launching launch.py...
17:52:34-228562 DEBUG    Logger: file=/home/PACE/sm98741n/automatic/sdnext.log level=10 size=0 mode=create                
17:52:34-231966 INFO     Starting SD.Next                                                                                 
17:52:34-233617 INFO     Python 3.10.13 on Linux                                                                          
17:52:34-258529 INFO     Version: app=sd.next updated=2023-11-11 hash=39085ab3                                            
                         url=https://github.com/vladmandic/automatic/tree/master                                          
17:52:34-266471 INFO     Platform: arch=x86_64 cpu=x86_64 system=Linux release=3.10.0-957.el7.x86_64 python=3.10.13       
17:52:34-267509 DEBUG    Setting environment tuning                                                                       
17:52:34-268861 DEBUG    Torch overrides: cuda=False rocm=False ipex=False diml=False openvino=False                      
17:52:34-270272 DEBUG    Torch allowed: cuda=True rocm=True ipex=True diml=True openvino=True                             
17:52:34-271681 DEBUG    Package not found: onnxruntime                                                                   
17:52:34-272900 INFO     Installing package: onnxruntime                                                                  
17:52:34-274153 DEBUG    Running pip: install --upgrade onnxruntime                                                       
17:53:06-430711 INFO     nVidia CUDA toolkit detected: nvidia-smi present                                                 
17:53:06-431946 DEBUG    Package not found: onnxruntime-gpu                                                               
17:53:06-433364 INFO     Installing package: onnxruntime-gpu                                                              
17:53:06-434631 DEBUG    Running pip: install --upgrade onnxruntime-gpu                                                   
17:53:28-100565 DEBUG    Package not found: torch                                                                         
17:53:28-101813 DEBUG    Package not found: torchvision                                                                   
17:53:28-103086 INFO     Installing package: torch torchvision --index-url https://download.pytorch.org/whl/cu118         
17:53:28-104309 DEBUG    Running pip: install --upgrade torch torchvision --index-url                                     
                         https://download.pytorch.org/whl/cu118                                                           
17:57:00-432369 DEBUG    Repository update time: Sat Nov 11 11:38:07 2023                                                 
17:57:00-433492 INFO     Startup: standard                                                                                
17:57:00-434935 INFO     Verifying requirements                                                                           
17:57:00-436981 DEBUG    Package not found: addict                                                                        
17:57:00-438581 INFO     Installing package: addict                                                                       
17:57:00-440210 DEBUG    Running pip: install --upgrade addict                                                            
17:57:01-339203 DEBUG    Package not found: aenum                                                                         
17:57:01-340251 INFO     Installing package: aenum                                                                        
17:57:01-341858 DEBUG    Running pip: install --upgrade aenum                                                             
17:57:02-241516 DEBUG    Package not found: aiohttp                                                                       
17:57:02-242567 INFO     Installing package: aiohttp                                                                      
17:57:02-244091 DEBUG    Running pip: install --upgrade aiohttp                                                           
17:57:06-667609 DEBUG    Package not found: anyio                                                                         
17:57:06-668915 INFO     Installing package: anyio                                                                        
17:57:06-670515 DEBUG    Running pip: install --upgrade anyio                                                             
17:57:08-042177 DEBUG    Package not found: appdirs                                                                       
17:57:08-043239 INFO     Installing package: appdirs                                                                      
17:57:08-044704 DEBUG    Running pip: install --upgrade appdirs                                                           
17:57:08-795333 DEBUG    Package not found: astunparse                                                                    
17:57:08-796393 INFO     Installing package: astunparse                                                                   
17:57:08-797937 DEBUG    Running pip: install --upgrade astunparse                                                        
17:57:09-913485 DEBUG    Package not found: blendmodes                                                                    
17:57:09-914541 INFO     Installing package: blendmodes                                                                   
17:57:09-915976 DEBUG    Running pip: install --upgrade blendmodes                                                        
17:57:12-690304 DEBUG    Package not found: clean-fid                                                                     
17:57:12-691362 INFO     Installing package: clean-fid                                                                    
17:57:12-692827 DEBUG    Running pip: install --upgrade clean-fid                                                         
17:57:29-967756 DEBUG    Package not found: easydev                                                                       
17:57:29-969090 INFO     Installing package: easydev                                                                      
17:57:29-970642 DEBUG    Running pip: install --upgrade easydev                                                           
17:57:31-902152 DEBUG    Package not found: extcolors                                                                     
17:57:31-903205 INFO     Installing package: extcolors                                                                    
17:57:31-904807 DEBUG    Running pip: install --upgrade extcolors                                                         
17:57:32-793311 DEBUG    Package not found: facexlib                                                                      
17:57:32-794357 INFO     Installing package: facexlib                                                                     
17:57:32-795612 DEBUG    Running pip: install --upgrade facexlib                                                          
17:58:01-945493 DEBUG    Package not found: filetype                                                                      
17:58:01-946841 INFO     Installing package: filetype                                                                     
17:58:01-948274 DEBUG    Running pip: install --upgrade filetype                                                          
17:58:03-236423 DEBUG    Package not found: future                                                                        
17:58:03-237489 INFO     Installing package: future                                                                       
17:58:03-238907 DEBUG    Running pip: install --upgrade future                                                            
17:58:05-904401 DEBUG    Package not found: gdown                                                                         
17:58:05-905462 INFO     Installing package: gdown                                                                        
17:58:05-907124 DEBUG    Running pip: install --upgrade gdown                                                             
17:58:07-294274 DEBUG    Package not found: gfpgan                                                                        
17:58:07-295322 INFO     Installing package: gfpgan                                                                       
17:58:07-296813 DEBUG    Running pip: install --upgrade gfpgan                                                            
17:58:28-058820 DEBUG    Package not found: GitPython                                                                     
17:58:28-060115 INFO     Installing package: GitPython                                                                    
17:58:28-061532 DEBUG    Running pip: install --upgrade GitPython                                                         
17:58:29-853324 DEBUG    Package not found: httpcore                                                                      
17:58:29-854387 INFO     Installing package: httpcore                                                                     
17:58:29-856091 DEBUG    Running pip: install --upgrade httpcore                                                          
17:58:31-438265 DEBUG    Package not found: inflection                                                                    
17:58:31-439324 INFO     Installing package: inflection                                                                   
17:58:31-480347 DEBUG    Running pip: install --upgrade inflection                                                        
17:58:32-559880 DEBUG    Package not found: jsonmerge                                                                     
17:58:32-560947 INFO     Installing package: jsonmerge                                                                    
17:58:32-562341 DEBUG    Running pip: install --upgrade jsonmerge                                                         
17:58:34-684888 DEBUG    Package not found: kornia                                                                        
17:58:34-685944 INFO     Installing package: kornia                                                                       
17:58:34-687669 DEBUG    Running pip: install --upgrade kornia                                                            
17:58:41-051582 DEBUG    Package not found: lark                                                                          
17:58:41-052820 INFO     Installing package: lark                                                                         
17:58:41-054361 DEBUG    Running pip: install --upgrade lark                                                              
17:58:42-486963 DEBUG    Package not found: lmdb                                                                          
17:58:42-488018 INFO     Installing package: lmdb                                                                         
17:58:42-489423 DEBUG    Running pip: install --upgrade lmdb                                                              
17:58:43-603196 DEBUG    Package not found: lpips                                                                         
17:58:43-604260 INFO     Installing package: lpips                                                                        
17:58:43-605690 DEBUG    Running pip: install --upgrade lpips                                                             
17:58:44-782447 DEBUG    Package not found: omegaconf                                                                     
17:58:44-783491 INFO     Installing package: omegaconf                                                                    
17:58:44-785102 DEBUG    Running pip: install --upgrade omegaconf                                                         
17:58:46-749128 DEBUG    Package not found: open-clip-torch                                                               
17:58:46-750189 INFO     Installing package: open-clip-torch                                                              
17:58:46-751705 DEBUG    Running pip: install --upgrade open-clip-torch                                                   
17:58:55-408317 DEBUG    Package not found: opencv-contrib-python-headless                                                
17:58:55-409680 INFO     Installing package: opencv-contrib-python-headless                                               
17:58:55-411139 DEBUG    Running pip: install --upgrade opencv-contrib-python-headless                                    
17:59:00-878651 DEBUG    Package not found: piexif                                                                        
17:59:00-879890 INFO     Installing package: piexif                                                                       
17:59:00-881326 DEBUG    Running pip: install --upgrade piexif                                                            
17:59:02-585567 DEBUG    Package not found: psutil                                                                        
17:59:02-586619 INFO     Installing package: psutil                                                                       
17:59:02-588311 DEBUG    Running pip: install --upgrade psutil                                                            
17:59:04-930364 DEBUG    Package not found: pyyaml                                                                        
17:59:04-931438 INFO     Installing package: pyyaml                                                                       
17:59:04-932962 DEBUG    Running pip: install --upgrade pyyaml                                                            
17:59:06-172239 DEBUG    Package not found: resize-right                                                                  
17:59:06-173281 INFO     Installing package: resize-right                                                                 
17:59:06-174870 DEBUG    Running pip: install --upgrade resize-right                                                      
17:59:07-438323 DEBUG    Package not found: rich                                                                          
17:59:07-439386 INFO     Installing package: rich                                                                         
17:59:07-440801 DEBUG    Running pip: install --upgrade rich                                                              
17:59:08-715709 DEBUG    Package not found: safetensors                                                                   
17:59:08-716747 INFO     Installing package: safetensors                                                                  
17:59:08-718258 DEBUG    Running pip: install --upgrade safetensors                                                       
17:59:10-015743 DEBUG    Package not found: scipy                                                                         
17:59:10-016781 INFO     Installing package: scipy                                                                        
17:59:10-018215 DEBUG    Running pip: install --upgrade scipy                                                             
17:59:11-352584 DEBUG    Package not found: tb_nightly                                                                    
17:59:11-353629 INFO     Installing package: tb_nightly                                                                   
17:59:11-355297 DEBUG    Running pip: install --upgrade tb_nightly                                                        
17:59:12-966678 DEBUG    Package not found: toml                                                                          
17:59:12-967725 INFO     Installing package: toml                                                                         
17:59:12-969388 DEBUG    Running pip: install --upgrade toml                                                              
17:59:14-546806 DEBUG    Package not found: torchdiffeq                                                                   
17:59:14-547852 INFO     Installing package: torchdiffeq                                                                  
17:59:14-549453 DEBUG    Running pip: install --upgrade torchdiffeq                                                       
17:59:15-956029 DEBUG    Package not found: voluptuous                                                                    
17:59:15-957079 INFO     Installing package: voluptuous                                                                   
17:59:15-958565 DEBUG    Running pip: install --upgrade voluptuous                                                        
17:59:17-290128 DEBUG    Package not found: yapf                                                                          
17:59:17-291193 INFO     Installing package: yapf                                                                         
17:59:17-292841 DEBUG    Running pip: install --upgrade yapf                                                              
17:59:18-538822 DEBUG    Package not found: scikit-image                                                                  
17:59:18-539849 INFO     Installing package: scikit-image                                                                 
17:59:18-541513 DEBUG    Running pip: install --upgrade scikit-image                                                      
17:59:19-818770 DEBUG    Package not found: basicsr                                                                       
17:59:19-819796 INFO     Installing package: basicsr                                                                      
17:59:19-821308 DEBUG    Running pip: install --upgrade basicsr                                                           
17:59:21-153731 DEBUG    Package not found: fasteners                                                                     
17:59:21-154779 INFO     Installing package: fasteners                                                                    
17:59:21-156497 DEBUG    Running pip: install --upgrade fasteners                                                         
17:59:22-455339 DEBUG    Package not found: dctorch                                                                       
17:59:22-456410 INFO     Installing package: dctorch                                                                      
17:59:22-457756 DEBUG    Running pip: install --upgrade dctorch                                                           
17:59:23-765356 DEBUG    Package not found: httpx                                                                         
17:59:23-766408 INFO     Installing package: httpx==0.24.1                                                                
17:59:23-768112 DEBUG    Running pip: install --upgrade httpx==0.24.1                                                     
17:59:25-578771 DEBUG    Package not found: compel                                                                        
17:59:25-579825 INFO     Installing package: compel==2.0.2                                                                
17:59:25-581455 DEBUG    Running pip: install --upgrade compel==2.0.2                                                     
17:59:50-104266 DEBUG    Package not found: torchsde                                                                      
17:59:50-105604 INFO     Installing package: torchsde==0.2.6                                                              
17:59:50-107296 DEBUG    Running pip: install --upgrade torchsde==0.2.6                                                   
17:59:52-147877 DEBUG    Package not found: clip-interrogator                                                             
17:59:52-148939 INFO     Installing package: clip-interrogator==0.6.0                                                     
17:59:52-150484 DEBUG    Running pip: install --upgrade clip-interrogator==0.6.0                                          
17:59:54-893161 DEBUG    Package not found: antlr4-python3-runtime                                                        
17:59:54-894219 INFO     Installing package: antlr4-python3-runtime==4.9.3                                                
17:59:54-895916 DEBUG    Running pip: install --upgrade antlr4-python3-runtime==4.9.3                                     
17:59:56-540593 DEBUG    Package not found: requests                                                                      
17:59:56-541643 INFO     Installing package: requests==2.31.0                                                             
17:59:56-543216 DEBUG    Running pip: install --upgrade requests==2.31.0                                                  
17:59:58-437801 DEBUG    Package not found: tqdm                                                                          
17:59:58-438862 INFO     Installing package: tqdm==4.66.1                                                                 
17:59:58-440637 DEBUG    Running pip: install --upgrade tqdm==4.66.1                                                      
18:00:00-108222 DEBUG    Package not found: accelerate                                                                    
18:00:00-109289 INFO     Installing package: accelerate==0.20.3                                                           
18:00:00-110876 DEBUG    Running pip: install --upgrade accelerate==0.20.3                                                
18:00:03-254214 DEBUG    Package not found: opencv-python-headless                                                        
18:00:03-255455 INFO     Installing package: opencv-python-headless==4.7.0.72                                             
18:00:03-256822 DEBUG    Running pip: install --upgrade opencv-python-headless==4.7.0.72                                  
18:00:11-220364 DEBUG    Package not found: diffusers                                                                     
18:00:11-221688 INFO     Installing package: diffusers==0.23.0                                                            
18:00:11-223105 DEBUG    Running pip: install --upgrade diffusers==0.23.0                                                 
18:00:12-899938 DEBUG    Package not found: einops                                                                        
18:00:12-900980 INFO     Installing package: einops==0.4.1                                                                
18:00:12-902808 DEBUG    Running pip: install --upgrade einops==0.4.1                                                     
18:00:14-756994 DEBUG    Package not found: gradio                                                                        
18:00:14-758025 INFO     Installing package: gradio==3.43.2                                                               
18:00:14-759568 DEBUG    Running pip: install --upgrade gradio==3.43.2                                                    
18:00:54-749916 DEBUG    Package not found: huggingface_hub                                                               
18:00:54-751269 INFO     Installing package: huggingface_hub==0.18.0                                                      
18:00:54-753056 DEBUG    Running pip: install --upgrade huggingface_hub==0.18.0                                           
18:00:59-453942 DEBUG    Package not found: numexpr                                                                       
18:00:59-455226 INFO     Installing package: numexpr==2.8.4                                                               
18:00:59-456809 DEBUG    Running pip: install --upgrade numexpr==2.8.4                                                    
18:01:01-515062 DEBUG    Package not found: numpy                                                                         
18:01:01-516130 INFO     Installing package: numpy==1.24.4                                                                
18:01:01-517642 DEBUG    Running pip: install --upgrade numpy==1.24.4                                                     
18:01:11-664546 DEBUG    Package not found: numba                                                                         
18:01:11-665891 INFO     Installing package: numba==0.57.1                                                                
18:01:11-667447 DEBUG    Running pip: install --upgrade numba==0.57.1                                                     
18:01:29-462712 DEBUG    Package not found: pandas                                                                        
18:01:29-464182 INFO     Installing package: pandas==1.5.3                                                                
18:01:29-465738 DEBUG    Running pip: install --upgrade pandas==1.5.3                                                     
18:01:50-285873 DEBUG    Package not found: protobuf                                                                      
18:01:50-287213 INFO     Installing package: protobuf==3.20.3                                                             
18:01:50-288688 DEBUG    Running pip: install --upgrade protobuf==3.20.3                                                  
18:01:52-943103 DEBUG    Package not found: pytorch_lightning                                                             
18:01:52-944173 INFO     Installing package: pytorch_lightning==1.9.4                                                     
18:01:52-945716 DEBUG    Running pip: install --upgrade pytorch_lightning==1.9.4                                          
18:01:59-813525 DEBUG    Package not found: transformers                                                                  
18:01:59-814816 INFO     Installing package: transformers==4.34.1                                                         
18:01:59-816209 DEBUG    Running pip: install --upgrade transformers==4.34.1                                              
18:02:29-636281 DEBUG    Package not found: tomesd                                                                        
18:02:29-637637 INFO     Installing package: tomesd==0.1.3                                                                
18:02:29-639030 DEBUG    Running pip: install --upgrade tomesd==0.1.3                                                     
18:02:31-723445 DEBUG    Package not found: urllib3                                                                       
18:02:31-724499 INFO     Installing package: urllib3==1.26.15                                                             
18:02:31-726243 DEBUG    Running pip: install --upgrade urllib3==1.26.15                                                  
18:02:34-119410 DEBUG    Package not found: Pillow                                                                        
18:02:34-120462 INFO     Installing package: Pillow==9.5.0                                                                
18:02:34-122186 DEBUG    Running pip: install --upgrade Pillow==9.5.0                                                     
18:02:37-869919 DEBUG    Package not found: timm                                                                          
18:02:37-871262 INFO     Installing package: timm==0.9.7                                                                  
18:02:37-872818 DEBUG    Running pip: install --upgrade timm==0.9.7                                                       
18:02:42-318379 DEBUG    Package not found: pydantic                                                                      
18:02:42-319705 INFO     Installing package: pydantic==1.10.13                                                            
18:02:42-321227 DEBUG    Running pip: install --upgrade pydantic==1.10.13                                                 
18:02:45-046095 DEBUG    Package not found: typing-extensions                                                             
18:02:45-047165 INFO     Installing package: typing-extensions==4.8.0                                                     
18:02:45-048709 DEBUG    Running pip: install --upgrade typing-extensions==4.8.0                                          
18:02:46-977135 INFO     Verifying packages                                                                               
18:02:46-978172 DEBUG    Package not found: clip                                                                          
18:02:46-979679 INFO     Installing package: git+https://github.com/openai/CLIP.git                                       
18:02:46-981196 DEBUG    Running pip: install --upgrade git+https://github.com/openai/CLIP.git                            
18:02:50-893230 DEBUG    Package not found: invisible-watermark                                                           
18:02:50-894567 INFO     Installing package:                                                                              
                         git+https://github.com/patrickvonplaten/invisible-watermark.git@remove_onnxruntime_depedency     
18:02:50-896288 DEBUG    Running pip: install --upgrade                                                                   
                         git+https://github.com/patrickvonplaten/invisible-watermark.git@remove_onnxruntime_depedency     
18:02:55-595317 DEBUG    Package not found: pi_heif                                                                       
18:02:55-596575 INFO     Installing package: pi-heif                                                                      
18:02:55-598032 DEBUG    Running pip: install --upgrade pi-heif                                                           
18:02:57-746631 DEBUG    Package not found: tensorflow                                                                    
18:02:57-747681 INFO     Installing package: tensorflow==2.13.0                                                           
18:02:57-749324 DEBUG    Running pip: install --upgrade tensorflow==2.13.0                                                
18:05:22-230452 INFO     Verifying submodules                                                                             
18:05:23-044065 DEBUG    Submodule: extensions-builtin/sd-extension-chainner / main                                       
18:05:23-102352 DEBUG    Submodule: extensions-builtin/sd-extension-system-info / main                                    
18:05:23-128286 DEBUG    Submodule: extensions-builtin/sd-webui-agent-scheduler / main                                    
18:05:23-164819 DEBUG    Submodule: extensions-builtin/sd-webui-controlnet / main                                         
18:05:23-201025 DEBUG    Submodule: extensions-builtin/stable-diffusion-webui-images-browser / main                       
18:05:23-225582 DEBUG    Submodule: extensions-builtin/stable-diffusion-webui-rembg / master                              
18:05:23-248377 DEBUG    Submodule: modules/k-diffusion / master                                                          
18:05:23-277489 DEBUG    Submodule: modules/lora / main                                                                   
18:05:23-318687 DEBUG    Submodule: wiki / master                                                                         
18:05:23-338068 DEBUG    Register paths                                                                                   
18:05:23-471264 DEBUG    Installed packages: 191                                                                          
18:05:23-472913 DEBUG    Extensions all: ['Lora', 'sd-extension-chainner', 'sd-extension-system-info',                    
                         'sd-webui-agent-scheduler', 'sd-webui-controlnet', 'stable-diffusion-webui-images-browser',      
                         'stable-diffusion-webui-rembg']

Everything is great and all, and the image generates flawlessly. there are no problems.
but when i shut down the server, and start the webui again, now it installs the xformers package like so:

Click to view the code

./webui.sh --share --listen --debug --insecure --cors-origins "*"
Create and activate python venv
Launching launch.py...
18:10:59-100586 DEBUG    Logger: file=/home/PACE/sm98741n/automatic/sdnext.log level=10 size=0 mode=create                
18:10:59-103865 INFO     Starting SD.Next                                                                                 
18:10:59-105463 INFO     Python 3.10.13 on Linux                                                                          
18:10:59-131453 INFO     Version: app=sd.next updated=2023-11-11 hash=39085ab3                                            
                         url=https://github.com/vladmandic/automatic/tree/master                                          
18:10:59-376435 INFO     Latest published version: cff5d637bc1be5b9fb5168d6a5af3f6abe13fc24 2023-11-12T22:35:33Z          
18:10:59-379400 INFO     Platform: arch=x86_64 cpu=x86_64 system=Linux release=3.10.0-957.el7.x86_64 python=3.10.13       
18:10:59-380459 DEBUG    Setting environment tuning                                                                       
18:10:59-381993 DEBUG    Torch overrides: cuda=False rocm=False ipex=False diml=False openvino=False                      
18:10:59-383706 DEBUG    Torch allowed: cuda=True rocm=True ipex=True diml=True openvino=True                             
18:10:59-385995 INFO     nVidia CUDA toolkit detected: nvidia-smi present                                                 
18:10:59-388113 DEBUG    Package not found: xformers                                                                      
18:10:59-389459 INFO     Installing package: --pre xformers                                                               
18:10:59-390749 DEBUG    Running pip: install --upgrade --no-deps --pre xformers                                          
18:11:17-165639 DEBUG    Repository update time: Sat Nov 11 11:38:07 2023                                                 
18:11:17-167133 INFO     Startup: standard                                                                                
18:11:17-168613 INFO     Verifying requirements                                                                           
18:11:17-183986 INFO     Verifying packages                                                                               
18:11:17-186079 INFO     Verifying submodules                                                                             
18:11:17-972847 DEBUG    Submodule: extensions-builtin/sd-extension-chainner / main                                       
18:11:18-032388 DEBUG    Submodule: extensions-builtin/sd-extension-system-info / main                                    
18:11:18-059338 DEBUG    Submodule: extensions-builtin/sd-webui-agent-scheduler / main                                    
18:11:18-096591 DEBUG    Submodule: extensions-builtin/sd-webui-controlnet / main                                         
18:11:18-173467 DEBUG    Submodule: extensions-builtin/stable-diffusion-webui-images-browser / main                       
18:11:18-196916 DEBUG    Submodule: extensions-builtin/stable-diffusion-webui-rembg / master                              
18:11:18-221115 DEBUG    Submodule: modules/k-diffusion / master                                                          
18:11:18-250736 DEBUG    Submodule: modules/lora / main                                                                   
18:11:18-293032 DEBUG    Submodule: wiki / master                                                                         
18:11:18-312654 DEBUG    Register paths                                                                                   
18:11:18-429320 DEBUG    Installed packages: 214                                                                          
18:11:18-430824 DEBUG    Extensions all: ['Lora', 'sd-extension-chainner', 'sd-extension-system-info',                    
                         'sd-webui-agent-scheduler', 'sd-webui-controlnet', 'stable-diffusion-webui-images-browser',      
                         'stable-diffusion-webui-rembg']

now some reason, image generation doesn't work. it gives this ouput, regardless of wether cross-attention optimization is set to xformers or not:

Click to view the code

18:26:27-255020 DEBUG    txt2img:                                                                                         
                         id_task=task(bly74a763yl2wrf)|prompt=1man|negative_prompt=|prompt_styles=[]|steps=20|sampler_inde
                         x=4|latent_index=None|full_quality=True|restore_faces=False|tiling=False|n_iter=1|batch_size=1|cf
                         g_scale=6|clip_skip=1|seed=-1.0|subseed=-1.0|subseed_strength=0|seed_resize_from_h=0|seed_resize_
                         from_w=0||height=512|width=512|enable_hr=False|denoising_strength=0.5|hr_scale=2|hr_upscaler=None
                         |hr_force=False|hr_second_pass_steps=20|hr_resize_x=0|hr_resize_y=0|image_cfg_scale=6|diffusers_g
                         uidance_rescale=0.7|refiner_steps=5|refiner_start=0.8|refiner_prompt=|refiner_negative=|override_
                         settings_texts=[]                                                                                
18:26:27-282138 DEBUG    Sampler: sampler="Euler a" config={'scheduler': 'default', 'brownian_noise': False}              
  0%|                                                                                              | 0/20 [00:00<?, ?it/s]
18:26:27-289341 ERROR    Exception: No operator found for `memory_efficient_attention_forward` with inputs:               
                              query       : shape=(2, 4096, 8, 40) (torch.float16)                                        
                              key         : shape=(2, 4096, 8, 40) (torch.float16)                                        
                              value       : shape=(2, 4096, 8, 40) (torch.float16)                                        
                              attn_bias   : <class 'NoneType'>                                                            
                              p           : 0.0                                                                           
                         `decoderF` is not supported because:                                                             
                             xFormers wasn't build with CUDA support                                                      
                             attn_bias type is <class 'NoneType'>                                                         
                             operator wasn't built - see `python -m xformers.info` for more info                          
                         `[email protected]` is not supported because:                                                       
                             xFormers wasn't build with CUDA support                                                      
                             requires device with capability > (8, 0) but your GPU has capability (7, 0) (too old)        
                             operator wasn't built - see `python -m xformers.info` for more info                          
                         `tritonflashattF` is not supported because:                                                      
                             xFormers wasn't build with CUDA support                                                      
                             requires device with capability > (8, 0) but your GPU has capability (7, 0) (too old)        
                             operator wasn't built - see `python -m xformers.info` for more info                          
                             triton is not available                                                                      
                             requires GPU with sm80 minimum compute capacity, e.g., A100/H100/L4                          
                             Only work on pre-MLIR triton for now                                                         
                         `cutlassF` is not supported because:                                                             
                             xFormers wasn't build with CUDA support                                                      
                             operator wasn't built - see `python -m xformers.info` for more info                          
                         `smallkF` is not supported because:                                                              
                             max(query.shape[-1] != value.shape[-1]) > 32                                                 
                             xFormers wasn't build with CUDA support                                                      
                             dtype=torch.float16 (supported: {torch.float32})                                             
                             operator wasn't built - see `python -m xformers.info` for more info                          
                             unsupported embed per head: 40                                                               
18:26:27-294424 ERROR    Arguments: args=('task(bly74a763yl2wrf)', '1man', '', [], 20, 4, None, True, False, False, 1, 1, 
                         6, 6, 0.7, 1, -1.0, -1.0, 0, 0, 0, 512, 512, False, 0.5, 2, 'None', False, 20, 0, 0, 5, 0.8, '', 
                         '', [], 0, False, False, 'positive', 'comma', 0, False, False, '', 0, '', [], 0, '', [], 0, '',  
                         [], False, True, False, False, False, False, 0, None, None, False, None, None, False, None, None,
                         False, 50, UiControlNetUnit(enabled=False, module='none', model='None', weight=1, image=None,    
                         resize_mode='Crop and Resize', low_vram=False, processor_res=-1, threshold_a=-1, threshold_b=-1, 
                         guidance_start=0, guidance_end=1, pixel_perfect=False, control_mode='Balanced',                  
                         save_detected_map=True), UiControlNetUnit(enabled=False, module='none', model='None', weight=1,  
                         image=None, resize_mode='Crop and Resize', low_vram=False, processor_res=-1, threshold_a=-1,     
                         threshold_b=-1, guidance_start=0, guidance_end=1, pixel_perfect=False, control_mode='Balanced',  
                         save_detected_map=True), UiControlNetUnit(enabled=False, module='none', model='None', weight=1,  
                         image=None, resize_mode='Crop and Resize', low_vram=False, processor_res=-1, threshold_a=-1,     
                         threshold_b=-1, guidance_start=0, guidance_end=1, pixel_perfect=False, control_mode='Balanced',  
                         save_detected_map=True)) kwargs={}                                                               
18:26:27-299936 ERROR    gradio call: NotImplementedError                                                                 
╭────────────────────────────────────────── Traceback (most recent call last) ───────────────────────────────────────────╮
│ /home/PACE/sm98741n/automatic/modules/call_queue.py:34 in f                                                            │
│                                                                                                                        │
│   33 │   │   │   try:                                                                                                  │
│ ❱ 34 │   │   │   │   res = func(*args, **kwargs)                                                                       │
│   35 │   │   │   │   progress.record_results(id_task, res)                                                             │
│                                                                                                                        │
│ /home/PACE/sm98741n/automatic/modules/txt2img.py:66 in txt2img                                                         │
│                                                                                                                        │
│   65 │   if processed is None:                                                                                         │
│ ❱ 66 │   │   processed = processing.process_images(p)                                                                  │
│   67 │   p.close()                                                                                                     │
│                                                                                                                        │
│ /home/PACE/sm98741n/automatic/modules/processing.py:738 in process_images                                              │
│                                                                                                                        │
│    737 │   │   │   with context_hypertile_vae(p), context_hypertile_unet(p):                                           │
│ ❱  738 │   │   │   │   res = process_images_inner(p)                                                                   │
│    739 │   finally:                                                                                                    │
│                                                                                                                        │
│ /home/PACE/sm98741n/automatic/extensions-builtin/sd-webui-controlnet/scripts/batch_hijack.py:42 in                     │
│ processing_process_images_hijack                                                                                       │
│                                                                                                                        │
│    41 │   │   │   # we are not in batch mode, fallback to original function                                            │
│ ❱  42 │   │   │   return getattr(processing, '__controlnet_original_process_images_inner')(p,                          │
│    43                                                                                                                  │
│                                                                                                                        │
│ /home/PACE/sm98741n/automatic/modules/processing.py:865 in process_images_inner                                        │
│                                                                                                                        │
│    864 │   │   │   │   with devices.without_autocast() if devices.unet_needs_upcast else device                        │
│ ❱  865 │   │   │   │   │   samples_ddim = p.sample(conditioning=c, unconditional_conditioning=u                        │
│    866 │   │   │   │   x_samples_ddim = [decode_first_stage(p.sd_model, samples_ddim[i:i+1].to(                        │
│                                                                                                                        │
│                                                ... 38 frames hidden ...                                                │
│                                                                                                                        │
│ /home/PACE/sm98741n/automatic/venv/lib/python3.10/site-packages/xformers/ops/fmha/__init__.py:223 in                   │
│ memory_efficient_attention                                                                                             │
│                                                                                                                        │
│   222 │   """                                                                                                          │
│ ❱ 223 │   return _memory_efficient_attention(                                                                          │
│   224 │   │   Inputs(                                                                                                  │
│                                                                                                                        │
│ /home/PACE/sm98741n/automatic/venv/lib/python3.10/site-packages/xformers/ops/fmha/__init__.py:321 in                   │
│ _memory_efficient_attention                                                                                            │
│                                                                                                                        │
│   320 │   if all(x.requires_grad is False for x in [inp.query, inp.key, inp.value]):                                   │
│ ❱ 321 │   │   return _memory_efficient_attention_forward(                                                              │
│   322 │   │   │   inp, op=op[0] if op is not None else None                                                            │
│                                                                                                                        │
│ /home/PACE/sm98741n/automatic/venv/lib/python3.10/site-packages/xformers/ops/fmha/__init__.py:337 in                   │
│ _memory_efficient_attention_forward                                                                                    │
│                                                                                                                        │
│   336 │   if op is None:                                                                                               │
│ ❱ 337 │   │   op = _dispatch_fw(inp, False)                                                                            │
│   338 │   else:                                                                                                        │
│                                                                                                                        │
│ /home/PACE/sm98741n/automatic/venv/lib/python3.10/site-packages/xformers/ops/fmha/dispatch.py:120 in _dispatch_fw      │
│                                                                                                                        │
│   119 │   """                                                                                                          │
│ ❱ 120 │   return _run_priority_list(                                                                                   │
│   121 │   │   "memory_efficient_attention_forward",                                                                    │
│                                                                                                                        │
│ /home/PACE/sm98741n/automatic/venv/lib/python3.10/site-packages/xformers/ops/fmha/dispatch.py:63 in _run_priority_list │
│                                                                                                                        │
│    62 │   │   msg += "\n" + _format_not_supported_reasons(op, not_supported)                                           │
│ ❱  63 │   raise NotImplementedError(msg)                                                                               │
│    64                                                                                                                  │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
NotImplementedError: No operator found for `memory_efficient_attention_forward` with inputs:
     query       : shape=(2, 4096, 8, 40) (torch.float16)
     key         : shape=(2, 4096, 8, 40) (torch.float16)
     value       : shape=(2, 4096, 8, 40) (torch.float16)
     attn_bias   : <class 'NoneType'>
     p           : 0.0
`decoderF` is not supported because:
    xFormers wasn't build with CUDA support
    attn_bias type is <class 'NoneType'>
    operator wasn't built - see `python -m xformers.info` for more info
`[email protected]` is not supported because:
    xFormers wasn't build with CUDA support
    requires device with capability > (8, 0) but your GPU has capability (7, 0) (too old)
    operator wasn't built - see `python -m xformers.info` for more info
`tritonflashattF` is not supported because:
    xFormers wasn't build with CUDA support
    requires device with capability > (8, 0) but your GPU has capability (7, 0) (too old)
    operator wasn't built - see `python -m xformers.info` for more info
    triton is not available
    requires GPU with sm80 minimum compute capacity, e.g., A100/H100/L4
    Only work on pre-MLIR triton for now
`cutlassF` is not supported because:
    xFormers wasn't build with CUDA support
    operator wasn't built - see `python -m xformers.info` for more info
`smallkF` is not supported because:
    max(query.shape[-1] != value.shape[-1]) > 32
    xFormers wasn't build with CUDA support
    dtype=torch.float16 (supported: {torch.float32})
    operator wasn't built - see `python -m xformers.info` for more info
    unsupported embed per head: 40

the only option i see from online and stuff are to delete the venv, and re-install.
but i would then have to keep re-installling.
any clues?

Answered by SamSilmarilData

Nov 13, 2023

alright, i went into the venv and re-installed xformers meself. works now.

View full answer

SamSilmarilData · 2023-11-13T01:06:42Z

SamSilmarilData
Nov 13, 2023
Author

alright, i went into the venv and re-installed xformers meself. works now.

1 reply

giusparsifal May 26, 2024

Hello, As I newbie and I have same error, can you telle me please exactly what you mean "i went into the venv and re-installed xformers meself"?
Thank you

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Installation Issue: xFormers wasn't build with CUDA support #2505

{{title}}

{{editor}}'s edit

{{editor}}'s edit

Replies: 1 comment 1 reply

{{title}}

{{title}}

Select a reply

Installation Issue: xFormers wasn't build with CUDA support #2505

SamSilmarilData Nov 12, 2023

Replies: 1 comment · 1 reply

SamSilmarilData Nov 13, 2023 Author

giusparsifal May 26, 2024

SamSilmarilData
Nov 12, 2023

Replies: 1 comment 1 reply

SamSilmarilData
Nov 13, 2023
Author