How to install A1111 on MacOS (step by step guide) #2
Replies: 10 comments 75 replies
|
I downloaded the 0.0.2 a1111-setup script today and tried to run it, but it was unsuccessful: So i tried following the "existing install" instructions above instead. @viking1304 - I wonder if it would be a good idea for a1111-setup.sh to include those commands if an existing install was detected? After doing these, the Web UI loaded up successfully :) |
|
My M1 shows almost 100% usage of GPU during generation. 
|
|
I just released version 0.0.3, which brings some tweaks and minor improvements. |
|
Weird I wonder why my Radeon has low utilization. I'm just thinking... I'd like to remove the PyTorch install I have (the developer build) and go with the one you recommend. How can I uninstall PyTorch (installed using pip as instructed here)? |
|
Actually after some Googling I think I figured it out |
580X 8GB in my iMac (with exactly the same params as my M1, which does not work for you)
|
This is the exact reason why I do not change command line args by default anymore. Git reset hard should restore the original file from A1111, where params are not set at all, and then it will change the command line only if you use the Since I have a feeling that your installation is still a bit messy, please do this:
|
So after uninstalling my dev version of PyTorch, and following the instructions above to create a fresh install of A1111, I now have it running on PyTorch 2.1.2 :) 
I'm getting brown/grey boxes generated on the first run of a fresh install of A1111 though, somewhat as expected, since we already determined last week that the default arguments don't work for my setup (perhaps due to me being on Apple Intel.... what we first discussed on Discord anyway), so I updated them. For this initial run of SD UI, Terminal said:
so I edited my webui-user.sh file and updated line 13 to:
i.e. changing After doing this, now no models will load in the UI for me! :( |
You can run it from anywhere except
I thought this sentence was clear enough:
You had PyTorch installed as a global package, which usually is not a problem, but since it was newer than the one I specified, the development version was used. Now you can choose which version you want to use inside venv that A1111 uses.
Those are the default command line parameters for Mac that the A1111 team added, which, in theory, should work for any standard Mac. They will be used unless you specify something different in
I noticed that, in some cases, models couldn't load and produce errors when I switch from SDXL to 1.5 or vice versa if I do not switch VAE back to Automatic first. In this case, I recommend switching VAE back to Automatic and restarting webui before trying to change the model. |
The same thing just happened to me on iMac with 580x. I solved the problem by adding Try those command lines, please:
|
|
Thanks for the tip here. Unfortunately adding There seems to be an issue with the model loading component of the Web UI now..... no model is being loaded on startup of the Web UI, and I cannot switch the model in the UI any longer. I don't understand why it cannot find I don't think it's the subfolder that's the problem, since this issue was occurring even before I put my models in subfolders, plus it's mentioned it works here. Suspecting it was symlink issue, I deleted the symlink and copied the actual "v1-5-pruned-emaonly.safetensors" model from my other drive to the /models/Stable-diffusion/ folder. After that, I still get the "cannot find model" error: .... but when I try to generate an image, nothing happens. And when I try to switch model, nothing happens. :( Is there a way for "flush" previously loaded models and have A1111 rescan all models again like new? I don't particularly wanna run the a1111-setup.sh script again since it took me a while to set up all my symlinks! :) |
|
No worries - I did a complete reinstall of A1111 and am back up and running! |
|
Hi @viking1304, First off, thanks for your step-by-step guide on getting A1111 up and running on my MACOS. I've found it really useful, but I'm getting an error with Steps 5 and 6 of the common steps. I've followed all the Steps before that for the new install. For Step 5, I'm encountering this error: And for Step 6, I'm running into this issue: I must admit, I've never really used the terminal – probably like five times in my entire life. So, any guidance on what I might be doing wrong would be greatly appreciated. |
|
You need to be inside the stable-diffusion-webui folder for that to work. You probably closed the terminal after you did some of the previous steps, so the terminal sent you to your home folder when you re-opened it. To get to the proper folder, you need a command from the start: But... if you do not feel "at home" with the terminal, you can use my script, which will do everything for you. Just run this and wait a little: |
|
Thank you very much for the quick reply! It has worked this time and I was able to run it! |
|
Thanks a lot for the helpful script. How do I proceed with an update?
Do I run the script again? Will the directories with the stable diffusion models and the generated images retained? |
|
That sounds strange. What does Yes, you can rerun the script, and it should fix everything. Just be sure to use the latest version since something new and strange happens constantly. It will not touch your models, loras or extensions, but it will remove your venv folder, and a1111 will create a new one. The only thing that might need to be changed manually later would be command-line parameters in webui-user.sh, since it will be reset to default. If you want to use my params (that works best for me) instead of the default ones, run the script like this |
|
Thank you for your support Viking!
|
|
It looks like you were one of those who reverted A1111 to v1.7.0 by force. This should fix the detached state and give you the latest version of A1111:
I will update my script to handle detached head cases soon. Do not use torch 2.2.0 or newer (including nighly) if you are on macOS 14.4. Euler A and similar samplers do not work properly in that environment. Everything works as expected with 2.1.2. |
|
awesome script, thanks for making it! installing Forge with it significantly reduced time for image generation. but only for SD 1.5 models, SDXL models loading time is approx minute, moving takes around same time and 512x512px 20 steps image generation takes around 40 min. is there anything possible to get SDXL model working faster or just stick to SD 1.5? |
|
Thanks for reply! I later also discovered that configuration (webui-user.sh) coming with installation was working only with 512x512 images. Anything above, even 768x768 on SD1.5 was causing Mac to freeze. I looked in some threads/forums for info, changing configuration to what it is now in log have solved this issue. 512x512 20 steps images are generated in less than minute, 1024x1024 20 steps is taking around 3-5 minutes. System info: Times for SD/SDXL 512x512 20 step image generations (simple prompt: Photo, Red apple on wooden table, outdoors): Log from terminal (base) UserName@Mac GenAI % cd Forge ################################################################ ################################################################ ################################################################ ################################################################ ################################################################ To create a public link, set |
|
Do not use Try this instead: |
|
Unfortunately with this configuration SDXL model still takes more than 20 min to generate 512x512 image: To load target model SDXLClipModel |
|
Your main problem is RAM. I think you should stick with 1.5, and try lightning since it can generate good images in fewer steps (like 4 or 8). Also, do not use 512x512 with SDXL (except with turbo). Use 768x768, 1024x1024, 1024x768 or 768x1024. For params, try this as suggested in your previous log: and also this |
|
Thanks! No luck with this settings and suggestion too. Will stick to sd1.5 for now |
|
Thanks for the script, installation went without issues using the 0.1.0 script from 3 weeks ago. I have used A1111 some on PC, and decided it was time to try on MacOS. Running on Sonoma 14.4.1 on M3 Max w/36GB RAM, I've found that several of the sampling methods produce only static. To test, I'm just using the base 1.5 pruned. These samplers produce just static: I've tried a few variations of the command line args based on the mention of noise results, but those apparently were just for gpus on intel. Any suggestions on how to proceed? Should I try a different version of PyTorch? Also, does xformers work on arm macs? I'm getting conflicting information as I thought xformers was nvidia only, but some places mention it running on arm too. EDIT: OK, I'm not sure what fixed it, but I thought maybe it would be best to use the recommended Python version, so I backed down from 3.12 to 3.10.9. Maybe it was coincidence, but launching webui.sh, it downloaded a new torch nightly version, and the broken samplers are working. I assume once 2.4 is stable, we can switch from nightly to stable? |
|
Thanks for the follow-up. In the updated script, could it warn of incorrect python version and how to correct it? I'm still trying to wrap my head around what version of python will be used by a1111. Brew keeps installing the latest version, which I guess doesn't matter as it isn't in the shell path. The DMG version of python (I chose 3.10.9) is apparently the one to take precedent. Something else the script may want to address is the python certificates. Without the certs installed, it appears you can't use some (all?) SDXL models. |
|
My script will install the latest version of python3.10 and other prerequisites using brew, so there is no need to install anything manually. Script will even install brew itself if it is not installed. A1111 has a script that detects macOS and forces python3.10. In most cases problem occurs if the default python3 is 3.12 and 3.10 is manually installed and not in the path. You should not have a problem with the wrong python version if you are using my script and do not have any manually installed python versions. I could add the code to check if python3.10 will be detected by a1111 and throw an error if needed. I could also show a list of all detected python installations. But, I think that this detection should happen on a1111 itself. I will think about it and see if I will add a check to my script or make another PR to A1111. Let's see will they accept my current PR for different torch versions for Intel and ARM. I do not want to have two PRs that will change the same file at the same time. Please post the result of this command: |
I think where the confusion set in is that I had installed the most recent 3.12 Python DMG from Python.org for another project and promptly forgot about it. Then when using your script, I didn't realize that when I was downgrading the brew version from 3.12 to 3.10, it wasn't fixing the issues with A1111. I had just made the assumption that the brew version was going to be used until realizing I hand't looked at my path. (I have since replaced the 3.12.2 dmg with the 3.10.9 dmg and all is well.) Just FYI: |
|
You can install multiple Python versions on the same machine without any issues. It is not a downgrade, so other Python versions will not be affected. I currently have this: Friendly advice: remove all python versions installed with dmg, and just install any python version you need using brew. This one will install the latest stable version, so it will be 3.12 at this moment: Then just install 3.10 with: Remove venv and rerun ./webui.sh and it will create a new one for you, or just rerun my script since it will remove venv for you anyway. |
|
Got it, thanks. |
|
Hi, thank you for uploading this tool and instruction. After a fresh install using your tool, I am still getting this error: AttributeError: 'NoneType' object has no attribute 'lowvram' (AUTOMATIC1111/stable-diffusion-webui#15793) What troubleshooting steps do you suggest? |
|
Run this command and post the output:
I am not 100% sure if other macOS versions have the same display format as the last one, but the serial should not be in the first 12 lines of SPHardwareDataType output. Also, post the result of this command: |
|
Hi viking1304, I have a favour to ask: I'm using this in conjuction with an animation software 'Touch Designer' to create audio reactive visuals. Do you have any experience with this software and, if so, would you be willing to run/process/export a project on my behalf? many thanks. |
|
Please post HW info I asked, so I can see what exact Mac model and macOS you have, before I can give you any specific advice. |
|
I'm having an issue where I launch with There's definitely a 'venv' in the directory: ~/stable-diffusion-webui_dev/stable-diffusion-webui/venv But the installation just wanted to use my Homebrew's installation of Python and my system's PyTorch (2.1.2). I'm using pyenv to manage Python for my system (macOS Sonoma 14.4.1, M3 Max)—could that be causing the problem? Any help would be much appreciated. |
|
I have M3 Pro and 14.5, so that is not an issue. You are trying to activate non-existing venv (path is wrong), but do not do that. Please post the result of this command: Also this: And this |
|
@viking1304 Thanks so much for helping... There was a typo in my previous post... My venv was actually here: And I used Sorry for any confusion. Here's my Python installations (per your request):
However, I actually decided to do a fresh install following your guide and I discovered the following: On subsequent launches, after the first launch, if I forget to activate the venv before running So it seems like I need to make sure and always "manually" activate the venv via I thought that maybe adding something to the webui-user.sh file like... ...would activate the venv when running I'm curious though, is there a way to fix the problem I was having without removing the venv, forcing a new venv to be created during the next Could it be a cache issue, where it remembers the last time it ran, and if it ran w/out the venv activated, it'll do the same for the next run? |
|
You have a folder named A proper installation would only have one of those, not both. I advise you to download and run my script, which will do everything for you. Just run Then move your models from the current install to that new one, and remove the complete |
|
Your script is a gem. Thanks again for your help. |
|
First off thank you for the great script I really wished I had found it 2 weeks ago when I first started trying to get SD working on my mac!! I have hit a problem with your script when running against an existing install, it goes through all the stages and gets to the point of deletng venv and then resinstalling all the packages and then bombs out with 'No module named 'packaging'' as below. I have tried completley deleting the previous install and the doing a fresh but it fails at the same point above. Do you have any thoughts why this is happening? |
|
also I just tried on the original install and it produces the orginal error |
|
Can you please try something, since I do not have Intel Mac at home? I can only test this next week at work. Go to That will revert a1111 to version 1.8.0 Remove Please let me know if it work or not. If that also does not work, just do |
|
I tried what you suggested and I got two errors in terminal/ gui as follows:
I will revert back as you suggest. Just want to say a big thank you for working through these problems !!! |
|
I found iMac 2015 at work, and did some tests. I was getting This basically means that you need to use CPU only mode for Macs that are not on MacOS 13 or newer. Since this iMac cannot be updated to MacOS 13 (12.7.5 is the latest one it can run) and only have 2GB of VRAM, I had to do some parameters tweaking: This works, but it is very slow. I might add MacOS version and video memory checks to my script, and add some warnings and suggestions, but also tweaks (if possible). I will not update my install script before the new release of a1111, since they merged my PR that handles torch versions for ARM and Intel Macs. Otherwise I would need to make another release of my script after new A1111 release, so please have a little patience. |
|
Hi Viking, Apologies for taking some time to come back on this. I couldn not get your fix above working on my setup but after finding a little tweak for launch parameteresI have finally got stable diffusion working, its slooooow as you said but it does work. Here are the steps that I have gone through and the tweak that finally got it working.
I dont know why I need that last set of parameters after you got it working on other intel hardware. The two assumptions that I can offer are 1) there is something unique to my build (could be corrupted in some way) or 2) my OS is newer and has slightly differing requirements from what you tested on. Nonetheless without your script and troubleshooting on intel hw I would never have got SD working by myself - again a BIG thank you!!!!! |
dear bro:i had troubles though,i know nothing about it,hope you can help me to solve these problem,appreciate it!
|
|
|
update:i deleted all python from my libary the i ran your setups: |
i ran ./webui.sh then it showed these:
|
|
If you use or that will just display debug information and quit. Nothing more. To do actuall installation, you need to do this (without -i): Do not forget to do things explained in |
|
thank you bro,i am tryna do it more again. |
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
September 14th 2024 version
If you want to install Stable Diffusion webUI on your Mac or fix your current Stable Diffusion installation as fast as possible, use my a1111-setup script that will do everything for you.
Run
sh a1111-setup.shorchmod 755 a1111-setup.sh && ./a1111-setup.sh. Wait a few minutes (depending on your computer and internet speed) until you see WebUI opened in your default browser.Of course, you can still use this step-by-step guide if you want full control.
Important note
Stable Diffusion webUI and Forge will automatically install the best version of PyTorch for you. Currently, it is 2.3.1 for ARM and 2.1.2 for Intel Macs. *as soon as I find some time to properly test Torch 2.4.1 I will create a new pull request
PyTorch dropped support for Intel Macs. The last version of PyTorch that properly works on Intel Macs with MacOS 14.4+ is 2.1.2
Development versions of PyTorch can still be used on ARM Macs, but I do not recommend that.
DO NOT try to install Stable Diffusion versions optimized for AMD cards (DirectML, Zluda...) on iMacs with Radeon, since Mac uses Metal even for Radeon Cards
Step by step guide
The original version of this guide was tested and has been working without a problem on those Mac models:
The latest version was tested on:
New install:
(if you are installing Stable Diffusion for the first time)
If Homebrew is not installed, follow the instructions at https://brew.sh to install it. Keep the terminal window open and follow the instructions under "Next steps" to add Homebrew to your PATH
Open a new terminal window and run
brew install cmake protobuf rust python@3.10 git wgetType
cdto be sure that you are in your home folderClone the web UI repository by running
git clone https://github.com/AUTOMATIC1111/stable-diffusion-webuistable-diffusion-webui/models/Stable-diffusion. If you skip this step, the default model will be downloaded.Temporary problem with the download of the default model (September 2024)
The default model, which was automatically downloaded by the webui, has been deleted from huggingface. You can download it from here or you can download any SD1.5 or SDXL model you like from https://civitai.com/
Put the model you downloaded into
stable-diffusion-webui/models/Stable-diffusion.Existing install:
(if you have already installed Stable Diffusion)
Go to the stable-diffusion-webui folder using
cd stable-diffusion-webui(if your installation is in another folder go to that folder instead)
Use
git reset --hard origin/masterto remove local changes of AUTOMATIC1111 filesUse
git pullto update to the latest version of AUTOMATIC1111Run
venv/bin/pip cache purgeto clean pip’s wheel cacheRemove venv folder with
rm -rf venvor using Finder.Common steps:
Please check if you still have
venvfolder inside your stable diffusion folder. If you have it, please go back to the Existing install part and do everything from that part fistOpen webui-user.sh, find this line
and add the line below to use the recommended command line parameters for Mac
Your webui-user.sh should look like this:
Note: Do not edit
webui.shorwebui-macos-env.shOptional step: Since some models might give you this error
run this command to fix that error in advance
More information here
./webui.sh. A new Python virtual environment will be created and activated using venv and all missing dependencies will be automatically downloaded and installed. Web UI will be launched automatically at the endAdditional notes:
To relaunch the web UI process later, run
./webui.shagain. To update, rungit pullbefore running./webui.shIn some cases, default command line parameters provided by the A1111 team are not the best choice, and in some rare cases, they do not work.
You can try and see which of the command line parameters below works the best for you.
Command line args for Macs implemented by the A1111 team:
My recommendation for Macs with 16 GB or more of RAM:
For Macs with less then 16 GB of RAM:
For special cases when
--no-half-vaeproduce only noise:Those combinations also worked without error on my Macs:
All reactions