{"id":38730,"date":"2025-07-06T09:52:16","date_gmt":"2025-07-06T01:52:16","guid":{"rendered":"https:\/\/www.1ai.net\/?p=38730"},"modified":"2025-07-02T12:57:50","modified_gmt":"2025-07-02T04:57:50","slug":"comfyui-ai%e7%bb%98%e7%94%bb%e5%a4%aa%e6%85%a2%e6%80%8e%e4%b9%88%e5%8a%9e%ef%bc%9fgudacudnnpytorch%e5%ae%89%e8%a3%85%ef%bc%8c%e8%ae%a9%e4%bd%a0%e7%9a%84comfyui%e8%b5%b7%e9%a3%9e","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/38730.html","title":{"rendered":"What to do when comfyui AI painting is too slow?GUDA+cuDNN+pytorch installation to make your comfyUI take off!"},"content":{"rendered":"

Some people's computers are clearly well configured and have good graphics cards, but when they doAI Painting<\/a>When it's slow, it's probably because you don't have CUDA and cuDNN installed.<\/p>\n

My computer just did the system, comfyUI also reinstalled, comfyUI did not install CUDA, out of the map becomes very slow very slow, today I will take you to the CUDA, cuDNN and pytorch installation, so that comfyUI speed up a bit.<\/p>\n

\"43d674f1j00syr9ja0076d000u000irm\"<\/p>\n

Before I installed it, I ran a graph with kontext and it came in at 285 seconds.<\/p>\n

\"be062c4bj00syr9j60003d0009u0010m\"<\/p>\n

I. Installing CUDA<\/p>\n

1, in the installation of the first to determine which version of your computer to install the highest CUDA support<\/p>\n

First open the NVIDIA control panel on your computer<\/p>\n

\"e862d103j00syr9j8002gd000fm007xm\"<\/p>\n

CLICK ON THE NVIDIA CONTROL PANEL FOR SYSTEM INFORMATION TO THE BOTTOM LEFT OF THE PANEL<\/p>\n

\"ba1380a7j00syr9ja002md000qo00jrm\"<\/p>\n

Finding NVCGDA64.DLL means that our computer supports the installation of CUDA<\/p>\n

Remember the version number 12.9.90 at the end, this is the highest version supported by our computer and we can only install CUDA versions less than or equal to 12.9.90.<\/p>\n

2. Determine the version number of CUDA according to python<\/p>\n

Open PyTorch's webpage (https:\/\/pytorch.org) and scroll down to see the python version selection<\/p>\n

Since our CUDA supports up to 12.9.90, we have a choice of 11.8, 12.6 and 12.8 versions of CUDA.<\/p>\n

Here we choose version 12.8.<\/p>\n

\"6544bd29j00syr9j90031d000qq00hjm\"<\/p>\n

3. Download CUDA<\/p>\n

Once you have determined the version of CUDA, go to the CUDA download page and select the correct version to download:<\/p>\n

Website:<\/p>\n

https:\/\/developer.nvidia.com\/cuda-toolkit-archive<\/p>\n

\"6e9b11bej00syr9ja00gdd000u000r4m\"<\/p>\n

3. Installation of CUDA<\/p>\n

Download and install<\/p>\n

Since I already have Visual Studio installed, I didn't check it here.<\/p>\n

Click Next and default all the way to the end of the installation.<\/p>\n

\"12d51b26j00syr9j80031d000gi00ccm\"<\/p>\n

Once installed, check if the installation was successful<\/p>\n

win+r opens the running window, enter cmd to open the command line window<\/p>\n

Enter \"nvcc \u2013version\" to view CUDA version number, currently V12.8.61, indicating that CUDA was installed successfully<\/p>\n

enter \"set cuda\" to see if the environment variable is set<\/p>\n

\"5dca8c98j00syr9j6001dd000kn00aom\"<\/p>\n

You can also check to see if the environment variables have been set by doing the following:<\/p>\n

Right on my computer<\/p>\n

\"25266232j00syr9j7001hd000u000iom\"<\/p>\n

Click on Environment Variables and see that CUDA has been added to the system variables<\/p>\n

\"56075bf9j00syr9ja00ald000u000gwm\"<\/p>\n

II. Installation of cuDNN<\/p>\n

CUDA should be used in conjunction with cuDNN.<\/p>\n

The version number of cuDNN should be equal to or less than the version number of CUDA<\/p>\n

cuDNN URL:<\/p>\n

https:\/\/developer.nvidia.com\/rdp\/cudnn-archive#a-collapse897-120<\/p>\n

Since we installed CUDA version 12.8, the corresponding cuDNN should be installed with version 8.9.7.<\/p>\n

\"9f3205b4j00syr9ja00a1d000u000r4m\"<\/p>\n

After clicking on the link for version 8.9.7, download the first local an zip file.<\/p>\n

You need to log in before you can download it, just register with any email address and log in.<\/p>\n

\"3dd6c16aj00syr9ja0031d000ps00kom\"<\/p>\n

After downloading and unzipping, you get some files<\/p>\n

\"f2f07a27j00syr9j7000jd000gj006am\"<\/p>\n

Copy these files and paste them into the CUDA path you just made, replacing the original files<\/p>\n

The default installation path for CUDA is:<\/p>\n

C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v12.8<\/p>\n

\"a446e398j00syr9j9001sd000lm00eum\"<\/p>\n

After the copy is complete, the cuDNN installation is successful.<\/p>\n

III. Installing PyTorch<\/p>\n

Going back to the PyTorch page just above, notice the two circled parts in the image below, the circled part above means that Python must be 3.9 or later<\/p>\n

If you don't have python on your computer, you can install a<\/p>\n

python:https:\/\/www.python.org\/<\/p>\n

The bottom part is the pytorch installation command, copy it<\/p>\n

\"a47a24a2j00syr9ja004ud000mr00nsm\"<\/p>\n

Bring up the command line tool, paste the command you just copied, and hit enter.<\/p>\n

pip3 install tooltochvision tochaudio-index-url https:\/\/download.pytorch.org\/whl\/cu128<\/p>\n

\"a7eed619j00syr9j9000wd000u000gvm\"<\/p>\n

Wait patiently for the installation to complete<\/p>\n

\"aa2e1dabj00syr9ja0064d000u000gwm\"<\/p>\n

Now that CUDA, cuDNN, and pytorch are installed, it takes a lot less time to run the kontext workflow.<\/p>\n

\"834f6269j00syr9j60007d000fg001gm\"<\/p>\n

Four, other advanced settings of the Autumn Leaf starter<\/p>\n

In fact, Autumn Leaf Launcher has a lot of useful functions integrated into it, and will use it to bring us a lot of convenience.<\/p>\n

Click the following to enter the expert mode: Click the lower left-hand setting > General setting > expert mode<\/p>\n

\"fc773e06j00syr9ja004yd000u000irm\"<\/p>\n

1. Change the default browser<\/p>\n

Google Chrome is recommended here because it is the most stable browser, and if you don't have one in your computer, it is recommended to download one.<\/p>\n

\"36d57a66j00syr9ja004ud000u000irm\"<\/p>\n

2. python path override<\/p>\n

Set the path to python in the installation package as shown below<\/p>\n

\"d0ffc91aj00syr9ja005ud000u000irm\"<\/p>\n

2, check whether to install transformers and huggingface<\/p>\n

Go back to the python folder in akiba comfyui and type cmd in the address bar and enter to bring up the command line.<\/p>\n

\"a0cc6226j00syr9j9002cd000mk00i9m\"<\/p>\n

Enter the command: python.exe -m pip show transformers<\/p>\n

This command is used to check if transformers are installed.<\/p>\n

\"4df2524ej00syr9j9002td000u000gwm\"<\/p>\n

I have this one already installed. If yours is not installed, enter the following command to install it:<\/p>\n

python.exe -m pip install transformers<\/p>\n

Next, check if huggingface is installed by typing the following command<\/p>\n

python.exe -m pip show huggingface_hub<\/p>\n

\"1d5e5288j00syr9ja004md000u000gwm\"<\/p>\n

I have this one already installed. If yours is not installed, enter the following command to install it:<\/p>\n

python.exe -m pip install huggingface_hub<\/p>\n

3. Use the same method to set the git path override<\/p>\n

\"a90e9eaaj00syr9ja005ld000u000irm\"<\/p>\n

4. Setting up a web proxy<\/p>\n

Make sure the port number is the same when setting up the web proxy<\/p>\n

Then turn on the four switches behind the proxy and turn off the mirroring switch<\/p>\n

\"2d2fe245j00syr9ja004xd000u000irm\"<\/p>\n

Final Thoughts<\/p>\n

In fact, the GUDA, PyTorch and xformers installation methods we talked about today can be easily done in the Autumn Leaves Integration Pack.<\/p>\n

Finds \"Advanced Options-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------\u3002<\/p>\n

Of course, you'd better make sure which version your computer supports before installing it as I mentioned above.<\/p>\n

\"f6921890j00syr9ja005kd000u000ixm\"<\/p>\n

That's all for today, I hope you find it helpful.<\/p>","protected":false},"excerpt":{"rendered":"

Someone with a very high computer setup and a good graphic card, but at a very slow pace when you're doing an AI painting, you probably didn't install CUDA and CUDNN. I've just finished my computer, and I've got a new system, and I've got no CUDA, and I've got a slow, slow picture, and I'm gonna need you to install CUDA, CUDNN, and pytorch today, so that we can move the cofyUI. Before it was installed, I ran a map with kontext, which was 285 seconds. I. Installation of CUDA 1 and determination of which version of the computer will be most supported by the CUDA before it is installed<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[144],"tags":[491,1989,6276],"collection":[],"class_list":["post-38730","post","type-post","status-publish","format-standard","hentry","category-baike","tag-ai","tag-comfyui"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/38730","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=38730"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/38730\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=38730"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=38730"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=38730"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=38730"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}