Post "1 Click Installer for Automatic1111 SD Web UI, SDXL, ControlNet, All ControlNet Models, TensorRT (RTX Accelerator) and RTX 5000 SUPPORT on Windows" by "Furkan Gözükara" from patreon

Furkan Gözükara

trying to compile xformers to make it work :D

Furkan Gözükara

nice one let me fix that too :D

Nedo Braun

i tried myself your 1-click SUPIR to get it working with the 5090, but as i expected i got an error somewhere ;) but no hurry Doc!

Furkan Gözükara

Forge more up to date let me also update that

Furkan Gözükara

You are welcome. Let me know what you need :)

Nedo Braun

Next app working with the 5000er series! thank you Doc.!

Rabadan

Which is better today, Automatic1111 or webui-forge?

Furkan Gözükara

no it wont work. you need to have python git cuda installed. follow this video please : https://youtu.be/DrhUHnYfwC0

Jimmy V

Hi, just found this page after purchasing a laptop with a 4050 in it to help with SD. My question is, I'm running SD on Stability Matrix, will this 1-click installer work for this type of install or should I just do a fresh install of SD only?

Furkan Gözükara

thank you so much. yes now the scripts auto recognize and handle yaml files

jonathan sanchez

Are the .yaml files no longer necessary? are they obsolete? Your ControlNet model download script does not download them. thank you for your work, honestly it's one of the best you can find on the internet.

Furkan Gözükara

for dreambooth you need to have a specific version here : https://www.patreon.com/posts/auto-installer-84773926

Furkan Gözükara

thank you so much welcome

KwotheEragon

:-) I was fist happy but now i am not happy anymore. I found out that this new installer doenst work with dreambooth. It would have save me a lot of time, if there had been a notice about this.

KwotheEragon

Thx so much, i just found your amazing youtube video. I am so hapapy there is more, more on patreon. All your work and tutorilas that is masterpiece of AI. I think there is nothing commpared to your youtobe and your patreon. Thx

Furkan Gözükara

also please send me screenshot of your test so i can try on my computer to verify : monstermmorpg@gmail.com or from discord

Furkan Gözükara

if you get error during installation that means your C++ tools is not properly installed : https://youtu.be/-NjNy7afOQ0 also make sure that default python is 3.10.11 i installed several times yesterday perfectly working no errors

C. Jonas

Hi Furkan, when selecting Controlnet -> IP-Adapter -> IP-Adapter-Auto Preprocessor and ip-adapter-faceid-plusv2_sd15 [6e14fc1a] Model, i get the following red error message in the console. When installing a1111 from scratch it also gives an error when installing onnxruntime and all the dependencies, stating something about cuda 12 and incompatibilty with onnxruntime? I guess this leads the error below? Can you check? 2024-05-20 11:29:59,858 - ControlNet - INFO - Preview Resolution = 512 2024-05-20 11:29:59,858 - ControlNet - INFO - ip-adapter-auto => ip-adapter_face_id_plus 2024-05-20 11:29:59.9806034 [E:onnxruntime:Default, provider_bridge_ort.cc:1548 onnxruntime::TryGetProviderInfo_CUDA] D:\a\_work\1\s\onnxruntime\core\session\provider_bridge_ort.cc:1209 onnxruntime::ProviderLibrary::Get [ONNXRuntimeError] : 1 : FAIL : LoadLibrary failed with error 126 "" when trying to load "g:\installer_auto1111\stable-diffusion-webui\venv\lib\site-packages\onnxruntime\capi\onnxruntime_providers_cuda.dll" Applied providers: ['CPUExecutionProvider'], with options: {'CPUExecutionProvider': {}} find model: G:\installer_auto1111\stable-diffusion-webui\extensions\sd-webui-controlnet\annotator\downloads\insightface\models\buffalo_l\1k3d68.onnx landmark_3d_68 ['None', 3, 192, 192] 0.0 1.0 2024-05-20 11:30:00.5415761 [E:onnxruntime:Default, provider_bridge_ort.cc:1548 onnxruntime::TryGetProviderInfo_CUDA] D:\a\_work\1\s\onnxruntime\core\session\provider_bridge_ort.cc:1209 onnxruntime::ProviderLibrary::Get [ONNXRuntimeError] : 1 : FAIL : LoadLibrary failed with error 126 "" when trying to load "g:\installer_auto1111\stable-diffusion-webui\venv\lib\site-packages\onnxruntime\capi\onnxruntime_providers_cuda.dll" Applied providers: ['CPUExecutionProvider'], with options: {'CPUExecutionProvider': {}} find model: G:\installer_auto1111\stable-diffusion-webui\extensions\sd-webui-controlnet\annotator\downloads\insightface\models\buffalo_l\2d106det.onnx landmark_2d_106 ['None', 3, 192, 192] 0.0 1.0 2024-05-20 11:30:00.6250399 [E:onnxruntime:Default, provider_bridge_ort.cc:1548 onnxruntime::TryGetProviderInfo_CUDA] D:\a\_work\1\s\onnxruntime\core\session\provider_bridge_ort.cc:1209 onnxruntime::ProviderLibrary::Get [ONNXRuntimeError] : 1 : FAIL : LoadLibrary failed with error 126 "" when trying to load "g:\installer_auto1111\stable-diffusion-webui\venv\lib\site-packages\onnxruntime\capi\onnxruntime_providers_cuda.dll" Applied providers: ['CPUExecutionProvider'], with options: {'CPUExecutionProvider': {}} find model: G:\installer_auto1111\stable-diffusion-webui\extensions\sd-webui-controlnet\annotator\downloads\insightface\models\buffalo_l\det_10g.onnx detection [1, 3, '?', '?'] 127.5 128.0 2024-05-20 11:30:00.7516377 [E:onnxruntime:Default, provider_bridge_ort.cc:1548 onnxruntime::TryGetProviderInfo_CUDA] D:\a\_work\1\s\onnxruntime\core\session\provider_bridge_ort.cc:1209 onnxruntime::ProviderLibrary::Get [ONNXRuntimeError] : 1 : FAIL : LoadLibrary failed with error 126 "" when trying to load "g:\installer_auto1111\stable-diffusion-webui\venv\lib\site-packages\onnxruntime\capi\onnxruntime_providers_cuda.dll" Applied providers: ['CPUExecutionProvider'], with options: {'CPUExecutionProvider': {}} find model: G:\installer_auto1111\stable-diffusion-webui\extensions\sd-webui-controlnet\annotator\downloads\insightface\models\buffalo_l\genderage.onnx genderage ['None', 3, 96, 96] 0.0 1.0 2024-05-20 11:30:00.9349707 [E:onnxruntime:Default, provider_bridge_ort.cc:1548 onnxruntime::TryGetProviderInfo_CUDA] D:\a\_work\1\s\onnxruntime\core\session\provider_bridge_ort.cc:1209 onnxruntime::ProviderLibrary::Get [ONNXRuntimeError] : 1 : FAIL : LoadLibrary failed with error 126 "" when trying to load "g:\installer_auto1111\stable-diffusion-webui\venv\lib\site-packages\onnxruntime\capi\onnxruntime_providers_cuda.dll"

・F・リュウゼリアス

Latest version as of March 14, 2024 auto installer Open the installer_auto1111_v6 ZIP file sequentially install_automatic1111_Web_UI ↓ install_controlNet_and_download_models ↓ install_TensorRT If you do ↓ WARNING: Skipping nvidia-cudnn-cu11 as it is not installed. A message appears saying that CUDA11 is not installed. Is it okay to install CUDA? Since it says +cu121, is CUDA12.1 good? Is it okay to download and install CUDA11.6 normally? Skip it for now update_torch_xformers_install_adetailer ↓ download_TensorRT_optional ↓ Here too, it was written on WEB/YouTube that it says 0.1.2.3.4.5. When I run 0. Download all models 1. Download Realistic Vision 5.1 Tensor files 2. Download SD 1.5 Base Tensor files 3. Download Rev Animated Base 1.2.2 Tensor files 4. Download SDXL Base 1.0 Tensor files There were only 5 items. For now, I pressed 0 and converted everything, but if If you run this also on the SDXL model downloaded later in Civitai Will it be added to the WEB-UI TensorRT profile? Currently, even if I start the WEB-UI without being able to run this install_TensorRT batch file, the TensorRT tab will appear this time. In the webui-user batch file Although only --xformers is written, Do I need to include --no-half-vae, --medvram, and --lowvram? YouTube said that if I don't include them, a bug will occur, but is it okay to not include them? --opt-sdp-no-mem-attention f Can I include it? lastly Is stable-diffusion-webui-forge applicable? Thank you in advance.

Furkan Gözükara

when did you exactly got this error? without using TensorRT? it was giving error but still working last time i tried unless you compile TensorRT on that specific model. please give more context

Franco Acosta Diaz

Hello I Have this Error: A tensor with all NaNs was produced in VAE. Web UI will now convert VAE into 32-bit float and retry. To disable this behavior, disable the 'Automatically revert VAE to 32-bit floats' setting. To always start with 32-bit VAE, use --no-half-vae commandline flag. ========================================================================================== *** Error running process_batch: /workspace/stable-diffusion-webui/extensions/Stable-Diffusion-WebUI-TensorRT/scripts/trt.py Traceback (most recent call last): File "/workspace/stable-diffusion-webui/modules/scripts.py", line 808, in process_batch script.process_batch(p, *script_args, **kwargs) File "/workspace/stable-diffusion-webui/extensions/Stable-Diffusion-WebUI-TensorRT/scripts/trt.py", line 302, in process_batch if self.idx != sd_unet.current_unet.profile_idx: AttributeError: 'NoneType' object has no attribute 'profile_idx' --- 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████| 20/20 *** Error running process_batch: /workspace/stable-diffusion-webui/extensions/Stable-Diffusion-WebUI-TensorRT/scripts/trt.py<03:36, 3.51it/s] Traceback (most recent call last): File "/workspace/stable-diffusion-webui/modules/scripts.py", line 808, in process_batch script.process_batch(p, *script_args, **kwargs) File "/workspace/stable-diffusion-webui/extensions/Stable-Diffusion-WebUI-TensorRT/scripts/trt.py", line 302, in process_batch if self.idx != sd_unet.current_unet.profile_idx: AttributeError: 'NoneType' object has no attribute 'profile_idx' --- 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████| 20/20 [00:05<00:00, 3.79it/s] *** Error running process_batch: /workspace/stable-diffusion-webui/extensions/Stable-Diffusion-WebUI-TensorRT/scripts/trt.py<03:05, 3.98it/s] Traceback (most recent call last): File "/workspace/stable-diffusion-webui/modules/scripts.py", line 808, in process_batch script.process_batch(p, *script_args, **kwargs) File "/workspace/stable-diffusion-webui/extensions/Stable-Diffusion-WebUI-TensorRT/scripts/trt.py", line 302, in process_batch if self.idx != sd_unet.current_unet.profile_idx: AttributeError: 'NoneType' object has no attribute 'profile_idx' --- 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████| 20/20 [00:05<00:00, 3.82it/s] Reusing loaded model model/francoV1-step00001856.safetensors [549bc545a3] to load model/francoV1-step00003712.safetensors:54<03:11, 3.75it/s] Calculating sha256 for /workspace/stable-diffusion-webui/models/Stable-diffusion/model/francoV1-step00003712.safetensors: d81dc5a6b805f24f533d97d787dd9e2ab105cea38731b275126c3b71ccb04b4f Loading weights [d81dc5a6b8] from /workspace/stable-diffusion-webui/models/Stable-diffusion/model/francoV1-step00003712.safetensors Applying attention optimization: xformers... done. Weights loaded in 297.5s (send model to cpu: 3.4s, calculate hash: 8.8s, load weights from disk: 0.2s, apply weights to model: 252.2s, apply half(): 15.3s, apply dtype to VAE: 15.3s, move model to device: 2.3s). *** Error running process_batch: /workspace/stable-diffusion-webui/extensions/Stable-Diffusion-WebUI-TensorRT/scripts/trt.py Traceback (most recent call last): File "/workspace/stable-diffusion-webui/modules/scripts.py", line 808, in process_batch script.process_batch(p, *script_args, **kwargs) File "/workspace/stable-diffusion-webui/extensions/Stable-Diffusion-WebUI-TensorRT/scripts/trt.py", line 302, in process_batch if self.idx != sd_unet.current_unet.profile_idx: AttributeError: 'NoneType' object has no attribute 'profile_idx'

Furkan Gözükara

thanks a lot

Furkan Gözükara

ye not working yet sadly. there is a development for it though

C. Jonas

Another question: Can i use tensorrt with controlnet? I read that many controlnet models or preproccessors do not work with tensorrt installed.

C. Jonas

Thanks for the update for windows and thanks for your hard work! I really learned so much with your tutorials and programs / scripts!

Chris

Okay thanks, will try!

Furkan Gözükara

try to install another disk. looks like you have permission issues for some reason on your computer. stderr: fatal: detected dubious ownership in repository at

C. Jonas

Hi Mate, the installation of install_automatic1111_Web_UI.bat gives the following error: installing stuff stuff stuff then: Collecting mpmath>=0.19 Using cached https://download.pytorch.org/whl/mpmath-1.3.0-py3-none-any.whl (536 kB) Installing collected packages: mpmath, urllib3, typing-extensions, sympy, pillow, numpy, networkx, MarkupSafe, idna, fsspec, filelock, charset-normalizer, certifi, requests, jinja2, torch, torchvision Successfully installed MarkupSafe-2.1.5 certifi-2024.2.2 charset-normalizer-3.3.2 filelock-3.13.1 fsspec-2024.2.0 idna-3.6 jinja2-3.1.3 mpmath-1.3.0 networkx-3.2.1 numpy-1.26.4 pillow-10.2.0 requests-2.31.0 sympy-1.12 torch-2.1.2+cu121 torchvision-0.16.2+cu121 typing-extensions-4.10.0 urllib3-2.2.1 [notice] A new release of pip available: 22.3.1 -> 24.0 [notice] To update, run: g:\Auto1111_secourses\stable-diffusion-webui\venv\Scripts\python.exe -m pip install --upgrade pip Installing clip Installing open_clip Installing xformers Cloning assets into G:\Auto1111_secourses\stable-diffusion-webui\repositories\stable-diffusion-webui-assets... Cloning into 'G:\Auto1111_secourses\stable-diffusion-webui\repositories\stable-diffusion-webui-assets'... remote: Enumerating objects: 20, done. remote: Counting objects: 100% (20/20), done. remote: Compressing objects: 100% (18/18), done. remote: Total 20 (delta 0), reused 20 (delta 0), pack-reused 0 Receiving objects: 100% (20/20), 132.70 KiB | 310.00 KiB/s, done. Traceback (most recent call last): File "g:\Auto1111_secourses\stable-diffusion-webui\launch.py", line 48, in main() File "g:\Auto1111_secourses\stable-diffusion-webui\launch.py", line 39, in main prepare_environment() File "g:\Auto1111_secourses\stable-diffusion-webui\modules\launch_utils.py", line 410, in prepare_environment git_clone(assets_repo, repo_dir('stable-diffusion-webui-assets'), "assets", assets_commit_hash) File "g:\Auto1111_secourses\stable-diffusion-webui\modules\launch_utils.py", line 197, in git_clone run(f'"{git}" -C "{dir}" checkout {commithash}', None, "Couldn't checkout {name}'s hash: {commithash}") File "g:\Auto1111_secourses\stable-diffusion-webui\modules\launch_utils.py", line 115, in run raise RuntimeError("\n".join(error_bits)) RuntimeError: Couldn't checkout {name}'s hash: {commithash}. Command: "git" -C "G:\Auto1111_secourses\stable-diffusion-webui\repositories\stable-diffusion-webui-assets" checkout 6f7db241d2f8ba7457bac5ca9753331f0c266917 Error code: 128 stderr: fatal: detected dubious ownership in repository at 'G:/Auto1111_secourses/stable-diffusion-webui/repositories/stable-diffusion-webui-assets' 'G:/Auto1111_secourses/stable-diffusion-webui/repositories/stable-diffusion-webui-assets' is on a file system that does not record ownership To add an exception for this directory, call: git config --global --add safe.directory G:/Auto1111_secourses/stable-diffusion-webui/repositories/stable-diffusion-webui-assets Press any key . . .

Furkan Gözükara

good idea checking right now stay tuned

Pallavi Chauhan

Have you made an installer for DRAG GAN gradio app?

Furkan Gözükara

I hope it works good. thanks for mention

Bartosz Polecki

Yeah it dropped like 2 days ago but im glad i made you check it out ;) , XA has potential to make good old sd1.5 cnet TILE/Resample work with sdxl and i love tile so much. Anyway goodnight or day :)

Furkan Gözükara

Hello I just checked and no extensions are made for it yet. We need firs either Automatic1111 extension or Forge extension. I started following both

Bartosz Polecki

Hi Mate, could you make easy installation and tutorial for X-adapter ? Im to stupid to make it run :D anyway love your stuff, regards

Furkan Gözükara

ye try fresh install and check it out

Hassan Alhassan

ill reinstall it again while going through the video, it was working on the previous version fine. python 3.10.11

Furkan Gözükara

hello. what is your python version and how did you install? i just made a video to show how to install properly : https://youtu.be/-NjNy7afOQ0

Hassan Alhassan

am getting stuck here, Building wheels for collected packages: tensorrt Building wheel for tensorrt (pyproject.toml) ... /

Furkan Gözükara

thank you so much. i just did update and added those commands shared in that thread. not anymore DLL error when starting :)

Tapiocapioca

Take your time. Always thank you about your support.

Furkan Gözükara

ah that dll error. actually i am also getting that but let me check if i can auto fix it. checking thread again

Tapiocapioca

My install is fresh, I think something happen after you made the guide. The error I had is about some dll. Anyway you are amazing with your job.

Furkan Gözükara

hello. when i do a fresh installation it is working. sadly can't make it work for existing installations for everyone

Tapiocapioca

Hello, this is not working anymore, I fixed it with this https://github.com/NVIDIA/Stable-Diffusion-WebUI-TensorRT/issues/27 I think you can fix the script really easy :)

Furkan Gözükara

unfortunately you cant use --medvram or --lowvram. they load some part onto CPU and TensorRT doesnt support it

Vasiliy Bulanov

Hi! Got this error, is it ever occur to you? RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument mat1 in method wrapper_CUDA_addmm) Trying to fix it that way: def transfer_tensors_to_gpu(): if torch.cuda.is_available(): tensor_on_cpu = torch.tensor([1, 2, 3], dtype=torch.float32) tensor_on_gpu = tensor_on_cpu.cuda() looks probably a bit naive.:) Anyway I would be very grateful for any tip.

Michael Harleman

The fix didn't solve the problem. It only created another one. I don't remember what happened, but SD requires its own install of cudnn. Otherwise, it detects your GPU and Cuda:0 as two separate processors, and doesn't know which one to use. It is designed to use the one installed with Auto1111. I reinstalled from the beginning and am now running Tensor to rebuild the seriously long wait for a model it needs to export. The problem seems to stem from the step the fix takes where it uninstalls Cuda 11 thinking once cudnn is installed you don't need it anymore. It gets rid of the warnings, but causes another issue.

Furkan Gözükara

also do a fresh install of automatic1111. fixes a lot in many cases. i will update video link in the post

Furkan Gözükara

please use dev branch atm. there have been updates

Michael Harleman

https://github.com/NVIDIA/Stable-Diffusion-WebUI-TensorRT/issues/27#issuecomment-1767570566

Michael Harleman

I didn't use the dev branch because the video said that TensorRT failed to generate the model because you needed to be in the master branch. I'm really lost. It has been in processing mode all day. I have just been letting it run. As long as it shows it is processing, I don't want to mess it up. I'm not sure what to do to fix the issue now. I don't want to wait another 10 hours or however long to generate the model again. I'm afraid to use the fix now.

Furkan Gözükara

hello again. i just tested scripts. worked fine. what was the error you were getting with scripts and how you solved? I tested with dev branch of automatic1111

Michael Harleman

TensorRT seems to be working fine. I'm not sure why I get the errors, but I spoke to you on Discord and you told me to just ignore the errors. The fix was working however when I tried it, so you might want to look at it.

Furkan Gözükara

you have error during generating tensorRT or when using can you clarify that? if you have error when using try to switch to the dev branch of automatic1111

Michael Harleman

I got annoying startup errors having something to do with cudnn three times before the interface would start. I found this post that describes a fix which seemed to work: https://github.com/NVIDIA/Stable-Diffusion-WebUI-TensorRT/issues/27#issuecomment-1767570566. I used the fix and it seemed to get rid of the errors, but the model won't generate. It gets stuck on the last step and never generates the ONNX model file. I ended up restarting the process from the beginning of the tutorial. I'll see if the automatic installer works if I don't make any mistakes. If I may offer a suggestion, I think you clearly know what you are doing, but your students have never seen any of the things you are demonstrating done before, so they can't absorb it nearly as fast as you can cram it into an hour and a half video presentation. I have calculated that for every 30 minutes of your tutorials it will take me 8 to 10 hours to understand what I need to do and fix all the errors I manage to create. Not everyone that takes your classes is fit for Harvard or MIT.

Furkan Gözükara

wow interesting

Hassan Alhassan

for some reason, i used IP-Adapter-SDXL contol net and it worked without rebuilding

Furkan Gözükara

i don't have such issue. are you trying controlnet? --medvram or --lowvram? are you using auto installer etc?

Hassan Alhassan

the issue am facing is every time i restart A1111 i need to rebuild the model again, am not sure why, otherwise it will give me the error below, i did switch to Dev then Git pull then back to master still this is showing RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument mat1 in method wrapper_CUDA_addmm)

Furkan Gözükara

not yet but they are working i see. e.g. : https://github.com/NVIDIA/Stable-Diffusion-WebUI-TensorRT/issues/116

Samuel Mallevaes

Hello, did you get an answer about the Lora for SDXL issue ?

Furkan Gözükara

hello. a tutorial very soon hopefully with auto installer

Joel Maynard

hello, can i install TensorRT on runpod in automatic 1111

Furkan Gözükara

sadly not yet. it is mistake of the developers. i hope they will fix it soon.

Ran Bensimon

do we know if there's a fix for the CUDA errors in the startup?

Furkan Gözükara

yes switch to dev branch of automatic1111 sd web ui. also remove --medvram or --lowvram

Ran Bensimon

I'm getting this error: return F.linear(input, self.weight, self.bias) RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument mat1 in method wrapper_CUDA_addmm) Did I miss something? :)

Furkan Gözükara

put installer into a new folder. it will make a new one. i think it works with all RTX give it a try

Ran Bensimon

Thanks! it will affect Quadro RTX as well? And also, auto installer will overwrite existing installation? or make a new one?

Furkan Gözükara

added auto switch to the script and published a new video watch it

Furkan Gözükara

dev version will not overwrite older version but may not work very well with all extensions you have

Furkan Gözükara

yep huge improvement

Furkan Gözükara

yes please give it a try on sd 1.5 and let me know

Rafael Villalba

Does this mean that my RTX 3050 4GB could work? Heheh. Thank you anyway if that's not the case.

jaehyun AI

Wow... thank you. I'm thrilled. Another productivity upgrade today.

Rabadan

I'm using the default version. Is it possible to switch to the DEV version and overwrite the default content in some way? I would not like to lose the actual installation of the A1111 that has already been carried out.

Rabadan

Sorry. How?

Furkan Gözükara

yes switch to dev version of automatic1111 then do a git pull. it works

Rabadan

RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument mat1 in method wrapper_CUDA_addmm) Any ideia?

Furkan Gözükara

ah ye AMD different. it doesnt support xFormers

Furkan Gözükara

please delete python 3.10.6 and install python 3.10.11 as shown in this video : https://youtu.be/B5U7LJOvH6g

Mr. GT

This is the warning message I am getting: Launching Web UI with arguments: WARNING:xformers:WARNING[XFORMERS]: xFormers can't load C++/CUDA extensions. xFormers was built for: PyTorch 2.0.1+cu118 with CUDA 1108 (you have 2.0.0+cpu) Python 3.10.11 (you have 3.10.6) Please reinstall xformers (see https://github.com/facebookresearch/xformers#installing-xformers) Memory-efficient attention, SwiGLU, sparse and more won't be available. Set XFORMERS_MORE_DETAILS=1 for more details No module 'xformers'. Proceeding without it.

Mr. GT

I encountered the exact problem and found this and installed it. My question is: are all the settings in stable diffusion tool the same as yours here or shall I leave them as defaults?

Furkan Gözükara

hi Joe. please add --lowvram to your command line arguments in webui-user.bat file

John S

Hi SEC, I have this error: "OutOfMemoryError: CUDA out of memory. Tried to allocate 20.00 MiB (GPU 0; 4.00 GiB total capacity; 2.12 GiB already allocated; 18.50 MiB free; 2.14 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF" could you help with this??

Furkan Gözükara

thank you so much

Phillip Berry

The fact that i can run all of these on my external SSD is wild.. You the man! worth the subscribe for sure!

Furkan Gözükara

yes it will terminate CMD once installation has been completed . if you run again it deletes and start from 0. but i can modify script. how you would like it to behave? by the way after it is completed you run Automatic1111 Web UI instance from inside the folder automatic_web_ui with webui-user.bat

Bendef SLatt

I am using python 3.10/12150.1013 and I run the bat and it downloads the models completely, and exits terminal automatically. When I navigate back to the execution location, the models folder is filled and seems finished. 12 or so gigs size-wise. When I try to run auto_1111_sdxl.bat again it immediately removes all data from the installer script and starts over again. This is driving me up the wall with the quality of my internet connection. What gives?

Furkan Gözükara

it depends on each extension. but you can tell automatic1111 to look where for each model. it has starting parameters : https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Command-Line-Arguments-and-Settings

Ran Bensimon

Is there a way to place all models in the same folder and refer the GUI to them? for example, I have one installation for img2img/txt2img and another installation for Deforum, so I don't want to have duplicates of models/lora/controlnet folders. is there a way to place them all in one place and tell the installers where to find them?

Furkan Gözükara

thank you so much for your support

Rafał Ryniak

Dziękuję bardzo jutro nad tym popracuje

Furkan Gözükara

ah you are using AMD gpu. for AMD gpus i never worked but there is this fork can you test it? https://github.com/lshqqytiger/stable-diffusion-webui-directml

Rafał Ryniak

"Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check" Czy to znaczy że na mojej karcie Radeon r280 SDXL nie ruszy?

Furkan Gözükara

Please start ui from stable diffusion folder. Enter inside folder and find webui-user.bat file.

Jon

I was just closing out my windows, is there a way to keep the UI running? Thank you

Furkan Gözükara

Yes I can update scripts. But why are you doing that? You only need to run the bat file 1 time. Hopefully I can update script and let you know tomorrow. I am going to sleep now

Jon

Once I close the .bat cmd prompt and then reopen it, it redownloads the .safetensors files every time. Is there a way to prevent this?

Furkan Gözükara

most possibly you don't have python installed. please install python 3.10.9 and make sure that it is set as default. here my tutorial : https://youtu.be/B5U7LJOvH6g - please let me know what do you see when you open a cmd and type python. sorry for late reply

Reke

Hi I just tried to run the bat file but it closes almost immediately after allowing to "run anyway". Any thoughts? thanks

Furkan Gözükara

thank you so much for your support

EnjoyerOfAll

Thank you for this and your videos.

Furkan Gözükara

i am so sorry you were right. just fixed and uploaded auto_1111_sdxl.bat. please download again

JoeMarcus

3.10.11

Furkan Gözükara

it should not happen. what is your installed python version? i am gonna test again now. i am using python 3.10.9

JoeMarcus

Downloading sd_xl_base_1.0.safetensors: 100% Downloading sd_xl_refiner_1.0.safetensors: --->> as soon as it hits 100% on this part the CMD window closes and thats it...