New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Version 2 - Development #26
Comments
@onduboy @UrielCh @ChrisAcrobat @zaqxs123456 @Sparkenstein I've uploaded a Windows build for the new version (instructions above). It has an option for NSFW, to work around incorrect NSFW flagging (#23). It is very very much a beta version, because it is under development. So if you get a chance to try it out, please let me know if it crashes or fails to install. Thanks! :) |
Tried after a reboot with: $Env:PYTORCH_CUDA_ALLOC_CONF = "max_split_size_mb:4096"
.\stable-diffusion-ui.cmd > logs2.txt 2>&1 no effect |
|
Full stacktrace
Related issue: CUDA out of memory |
@UrielCh Thanks for trying this out. Yeah it definitely looks like the error you linked to. Like I mentioned in #23, I'm looking at basujindal's fork of Stable Diffusion, with the intention of using that fork in v2 instead of the stock Stable Diffusion (used currently). That might help reduce the VRAM usage. |
Hi,
But it correctly opens the webpage. I have latest Windows 11, Powershell 7.3.0. |
Hi, i get following Error when i try to Install it
The Website Opens normally Edit: Sadly i dont seem to be able to run this on my GPU? Error: CUDA out of memory. Tried to allocate 1024.00 MiB (GPU 0; 8.00 GiB total capacity; 6.14 GiB already allocated; 0 bytes free; 7.23 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 |
How do you delete the folder and rerun the installer? I got the ZIP file and then unpacked its content. :) |
@pegak You can delete the "stable-diffusion" folder (it'll be next to the "installer" folder and @pegak @oneandonlyjason Sorry about this issue, I'll modify the installer code to re-check the dependencies at the end. It seems like it didn't download some dependencies, not sure why. But I'll add a check at the end to fix any missing dependencies. Thanks for testing this out, and letting me know! :) |
I dont know how much this is in the Scope of This Project (Maybe when the Fork instead of the Original Code is used?) but can you implement other Sampling Methods that you can maybe Select in the Panel? https://github.com/crowsonkb/k-diffusion Other Projects like NightcafeStudio use KLMS instead of PLMS as Default Sampler and when i compare the Outputs from both Samplings i seem to get a lot Better Outputs with it most of the Time KLMS: PLMS: Both Images use the Same seed and same Prompt |
@oneandonlyjason Thanks, that looks interesting! Do you have a link handy for running it locally? The repo only shows an example for training, but I haven't looked closer. Thanks! |
I did not setup it locally yet and havent really found out how because i dont know enough about Python. But i found this Google AI-Notebook. Maybe the Code in this Helps? https://colab.research.google.com/github/pharmapsychotic/ai-notebooks/blob/main/pharmapsychotic_Stable_Diffusion.ipynb#scrollTo=bcHsbr3hblrk |
@pegak @oneandonlyjason - I've updated the v2 version with a new build. This checks for missing dependencies at the end and tries to fix them. Hopefully that should resolve the problem. You can download it from https://drive.google.com/file/d/1MY5gzsQHV_KREbYs3gw33QL4gGIlQRqj/view?usp=sharing and then double-click Please let me know if there are any problems after running it. Thanks! PS: There's a new Discord server for support and development discussions: https://discord.com/invite/u9yhsFmEkB . Please join in for faster discussion and feedback on v2. |
Hi! You requested to get error reports. D:\App\StableDiffusion\stable-diffusion-ui>installer\Scripts\activate.bat (installer) D:\App\StableDiffusion\stable-diffusion-ui\installer\etc\conda\activate.d>cd D:\App\StableDiffusion\stable-diffusion-ui\installer..\scripts (installer) D:\App\StableDiffusion\stable-diffusion-ui\scripts>on_env_start.bat "Stable Diffusion UI" "Ready to rock!" started in D:\App\StableDiffusion\stable-diffusion-ui\stable-diffusion |
Hi @mkuitune sorry about that, and thanks for reporting. Can you please open in And then save and double-click the Please paste the output here. It'll try to fix your installation, and show any errors while fixing. Thanks! |
This is great. Once I have used the installer, what it the best way to get the latest changes? |
Hi @rc1 - a new version is available at https://drive.google.com/file/d/1MY5gzsQHV_KREbYs3gw33QL4gGIlQRqj/view?usp=sharing This version will receive automatic updates in the future. It'll auto-update each time you start it. The updates are very small, so it won't add more than a second or two to the start. Please let me know if there are any problems. Thanks PS: Port and host isn't configurable right now, is it conflicting with something you're running? |
Thanks @cmdr2. So I should delete and reinstall. That's no problem. I'd like to change the host to 0.0.0.0 to make the UI accessible from my other computer. I wonder even if there was somewhere I could ninja edit the script to change it but I am new to windows scripting, so nothing jumped out as obvious (I didn't spend too long) |
Hi @rc1 the new update already sets the host to 0.0.0.0, so I think you should be set. Yes, please download the version I linked previously, and it should auto-download any updates in the future. Please let me know if there are any problems, thanks! PS: There's a new Discord server for support and development discussions: https://discord.com/invite/u9yhsFmEkB . Please join in for faster discussion and feedback on v2. |
Update: a Linux installer for v2 is now available in beta. Instructions for installing it are at: https://github.com/cmdr2/stable-diffusion-ui/blob/v2/README.md#installation |
I'm getting
However it ran fine on docker. Is there a way I can modify memory allocation? |
Hi @mflux - Can you please check if you're using an initial image that's larger than your output image? For e.g. please use 512x512 for your initial image, if that's your desired output image. Larger initial images can cause out of memory errors. I'm working on catching this in the UI. Another suggestion is to disable "Turbo Mode" in the "Advanced Settings", since it'll reduce VRAM usage by 1GB. Please let me know if this works out (or doesn't). Thanks! |
Update: the v2 branch has been integrated into main, so v1 (the docker approach) is officially dead, and v2 is the main version of the software now. Thanks everyone for testing this (here and on Discord), I really appreciate it! :) An experimental support for Mac is pending, I'm continuing to look into it on the side. But closing this issue for now. |
A development version of v2 is available for Windows 10/11 and Linux. Experimental support for Mac will be added soon.
The instructions for installing are at: https://github.com/cmdr2/stable-diffusion-ui/blob/v2/README.md#installation
It is not a binary, and the source code used for building this is open at https://github.com/cmdr2/stable-diffusion-ui/tree/v2
What is this?
This version is a 1-click installer. You don't need WSL or Docker or Python or anything beyond a working NVIDIA GPU with an updated driver. You don't need to use the command-line at all.
It'll download the necessary files from the original Stable Diffusion git repository, and set it up. It'll then start the browser-based interface like before.
An NSFW option is present in the interface, for those people who are unable to run their prompts without hitting the NSFW filter incorrectly.
Is it stable?
It has run successfully for a number of users, but I would love to know if it works on more computers. Please let me know if it works or fails in this thread, it'll be really helpful! Thanks :)
PS: There's a new Discord server for support and development discussions: https://discord.com/invite/u9yhsFmEkB . Please join in for faster discussion and feedback on v2.
The text was updated successfully, but these errors were encountered: