Hello! Thank you very very much! The fact that it's possible to use Photoshop and transfer images back and forth is immediately clear. I also saw messages about using SD as a plugin. I haven't tried it. But working in real time is fantastic! I'm almost an artist now! 🤓
Thank you for this video, I am still using CS6, and I have a problem with the photoshop connection node, is there any solution for it. Your channel is really great, appreciate a lot.
thanks a lot for this great tutorial, i exactely folowed your Steps but get everytime the errormessage " Image Resize >> NoneType object has no attribute shape
Yeah a video on Krita has been on the list for a while now, but I need time to test it out since I mainly use software that is standard for production environments. I will do a tutorial on this on Krita in the future, though.
I'm not sure there a way to tunnel the connection between PS as a server and comfy when using it on a cloud service. Unfortunately that's more of a sysadmin question, and I'm not qualified enough to give a proper answer
It depends on what you want to do with a custom image. This workflow pulls the currently active image from photoshop (if you have more than one open, it’s the one you’re currently working on). If you’d like, for example, to use an IPAdapter to influence the overall style of the generated image, you could do that with a custom image, or if you would like to start from another image to encode it into a latent and do a hybrid img2img process that would be possible as well. It all depends on which end result you’d like to use the custom image for. Let me know and I’ll see how I could help you with that.
@@risunobushi_ai Thank you for the reply, yes It does pull the image from photoshop but its not the same image anymore, background, face, is different, I wanted to have the same image, is there a reason why is that happening on my end?. thank you.
The generated image is always going to be at least a bit different from the original one in PS. If you wish for the generated image to be closer to the original image, you can try lowering the denoise value in the Ksampler. The lower the denoise is, the less changes are made to the original image. If you’d like to change one aspect of the original image, like the background only for example, that needs a way more complex workflow than this, which is more intended for quick compositing in PS and having SD take care of merging everything together in a way that makes sense to the eye. In the video, for example, I went through different examples at different denoising values to show how much the image changes depending on the denoise value. I hope this helps, otherwise let me know and I might do a more in-depth video about it!
@@risunobushi_ai Ahh isee, it would be a more complex node😅. Yes, thats what I met like change some aspect of the image like for example adding, watch, or glasses it would be cool if you have a video of that as well thank you so much for your time and quick response I really appreaciate it. Keep up the good work I am your subscriber now! Ive learned alot from your videos.
Thanks! Unfortunately AE is a bit beyond my expertise, being a photographer I only direct videos, not actually edit them, but I’ll look into ways to implement SD in AE if possible a bit further down the road. Right now I’ll focus mainly on photography, illustration, architecture, art / creative direction, and 3d, which are at least somehow related to my area of expertise (or I have friends whose area of expertise is among those and I can ask for what they’d like to see and can give me real life feedback on what I can brew)
@@risunobushi_ai I am gonna try the way you show here. I do digital painting. I think I can do something cool with your way. Thank you so much. It is great inspiration. I am following you!
LAN or WAN? I’m not a sysadmin by any stretch of the definition, but If it’s LAN I don’t think there’s any issues, it should be recognizable on the same network without any open port I’m guessing. If it’s WAN, I’m sure there must be a way to forward the address on which the server is communicating.
@@risunobushi_ai Just a LAN setup when its at home and a VLAN when its outside, ill try to figure it out. Too bad the connection cant just offload the actual photoshop link to comfy via the browser xD
It’s actually an interesting experiment to set up, so let me know if you find out how to make it work before I do, in the meantime I’ll try to do the same since I have a similar setup already in place
You're right, I'm sorry! I updated the description with a wetransfer link to the json file, and I'll update it later with a more definitive link so it doesn't expire after one year.
I'm not sure I've asked you before, but what OS are you on? Was it working for you before? Thanks for the heads up, I'll look into it. I've actually talked to the dev over the weekend, so I might get in touch with them as well.
What's your setup? OS, python version, is comfyui updated, ecc? Some people have reported issues with the popup, but it appears to be happening only to a few
@@risunobushi_ai I'm using windows, python version 3.12. I just installed comfy ui for the first time yesterday. I'm not sure if that's the latest version or not. I would assume so.
Do you have all the dependencies installed? Like pillow, torch, torchvision, open-cv, and numpy. To check, go to comfyUI -> custom_nodes -> comfyui-popup_preview -> window, open in terminal, and run “pip install -r requirements.txt
@@risunobushi_ai "pip install -r requirements.txt '"pip install -r requirements.txt' is not recognized as an internal or external command, operable program or batch file
Can you tell me how to download the LORA LCM model from Huggingface? Apologies for the noob question. I tried downloading the pytorch_lora_weights.safetensors from the 'Files and Versions' tab, but that throws an error in Comfy. "Prompt outputs failed validation. LoraLoader: - Required input is missing: clip" Thank you kindly for the guide!
From that error it looks like you might not be linking the CLIP output from the Load Checkpoint node to the CLIP input in the Load LoRA node? I'm asking because if you downloaded that safetensor you should have the right file
@@risunobushi_ai thank you kindly for the reply. Yep, you're correct. Needed to connect clip. I initially loaded your workflow without further configuring, but comfy is so far out of my comfort zone, it's not even funny. I think I have it working now. Still have some tinkering to do. Really appreciate your videos. Fellow photographer here, trying to figure out how best to use these tools in my workflows
Hello！ yesterday I studied the video carefully, but the real-time preview screen (preview pop-up) was not available, Later, when I painted in PS, I could also generate a shape close to PS. I will try to use the light source on the pictures generated by this workflow Change (IC light, your latest video), the picture quality will become a little blurry, not as high-definition as before
I later shared the node about PhotoshoptocomfyUI in your introduction below to find a way to preview pop-up in real time, but I need to download the official version of Adobe creative cloud desktop. The PS I use is a pirated version. Is there any The method is that the pop-up preview interface appears.
Going in order: - I have only tested the preview popup node with the official release, so I wouldn't know; - if the results are blurry, your relight mask and your source image are different in size
@@risunobushi_ai I installed the official Creative cloud and installed the built-in ComfyUI for adobe photoshop plug-in, but the Pop up preview still did not pop up. I found in my comfyUI startup program that Installing requirements... 'pip' is not recognized as an internal or external command, operable program or batch file. How do I need to install it so that the pop-up window will appear? Thank you so much:)
@@risunobushi_ai Thank you for your guidance. I have now downloaded popup-preview and the pip problem has been solved. However, I found a very strange problem. That is, the color of the product I put in PS is yellow. The real-time preview The interface is blue, and the export from comfyUI is also blue. I don’t know how this problem occurred, but being able to experience the whole process makes me, a beginner, extremely happy. By the way, if I need to add a specific product to a model, this product cannot produce any changes, but needs to be properly integrated with the model. One way I know is to use Krita. Is there a better way to achieve it?
Every model goes into the corresponding folder inside of your comfyUI -> model folder. So, for example, the LCM model is a checkpoint and goes into the checkpoint folder, a LoRA goes into the lora folder, if you had say a controlnet model it'd go inside the controlnet folder, etc. Once they're inside these folders, you just need to click "refresh" on the comfyUI sidebar and comfyUI will recognize that there's new models available.
Sorry, I didn’t see your answer. That’s weird, and it’s the second time I get this feedback from a windows user, where it should run without issues. I now the dev just updated his nodes, so I’m going to take a look at that next week. In the meantime, there’s a bit of a troubleshoot guide in the description of my other video about it, check if it works for you: th-cam.com/video/7G12SaLtEiA/w-d-xo.htmlsi=7ZHTgzB6Xg0SaTfP
Hi, I tried to install it on my Google Colab, I got this error, any idea how? File "/content/drive/MyDrive/ComfyUI/custom_nodes/comfyui-photoshop/__init__.py", line 15, in from photoshop import PhotoshopConnection ModuleNotFoundError: No module named 'photoshop'
I’m not familiar with Collab, so I’m not sure I can be of much help there. Do you git pull the repo inside the custom_nodes folder like you’d do on any other OS?
If you tried to install from comfyUI manager, the issue might be in the manager not finding the node in its repos. Try installing it directly from the custom nodes folder by opening the folder inside the terminal and using git pull “url from the repo”. The repo is in the video description.
@@thisthisthis9444 You can find the infos in the description, but you basically want to head over to the github page for the node: github.com/NimaNzrii/comfyui-photoshop/issues go to your comfyui - custom_nodes folder, right click, open in terminal, write "git clone" and the link to the node found on the green button in the github page, and then you'd need to downgrade the node (since it's been changed a lot because of an upgrade) by: going in the comfyui - custom_nodes - comfyui-photoshop folder, right click, open in terminal, write "git checkout --force 403d4a9af1f947c95367cd40ff8ad6ae65e5df41" this will downgrade the node. Actually I thought I updated this description as well, turns out I updated another video but not this one. it's updated now.
No worries, let’s see if I can help troubleshoot it. Did you git pull from the repo inside the custom-nodes folder? Did you restart comfyUI by any chance after you installed it? ComfyUI doesn’t see new nodes unless it restarts.
Would you mind deleting the node from the custom node folder and try to install it via the manager? It’s under install custom nodes -> search for “photoshop” -> #384, by NimaNzrii. That might help.
@@risunobushi_ai i tried this, it shows that the node is installed but still doesnt come up in the workflow. i cant search for the node nor can i find it in the add node menu.
I , i am a new user, and when i try my first generation i got an error like this Error occurred when executing KSampler: mat1 and mat2 shapes cannot be multiplied (154x2048 and 768x320) 😂
Oh you mean the Photoshop to comfyUI node? If you can’t install it via ComfyUI manager, you need to go to your comfyUI folder -> custom_nodes -> open in terminal -> type “git clone press Enter. What OS are you running? macOS gives some issues sometimes, while it works without issues on Windows and Linux.
I have windows system when i paste custom node and my termination saying install unsuccessfully you want to remove photoshop? Say yes or no.... Not installing please help
great need more video regarding fashion with AI , If do you have any personal contact please share i can suggest you some of content which we are really needed as designer
That photoshop node only works on Windows as it's trying to install its dependencies with "python.exe" 😅 Also the photoshop python package it's using requires 'winreg' which surely won't run on mac or linux.
Oh so that’s the source of all the macOS issues and why most of the users weren’t having any troubles with it! Great to know, I’ll make an update about it so that Mac users know about it. Also yeah, support and docs are non-existent with this node, but it’s the most straightforward I’ve seen that does what’s needed, such a shame it’s not polished. I might try to fork it and see if it’s possible to make it work on both Win / macOS.
Hey there! I’ll try to replicate the issue tomorrow on my MacBook, I’ll see if it’s native to the OS or it’s because of something else. Could you paste the error code?
@@risunobushi_ai 0.0 seconds (IMPORT FAILED): /Users/fabiogarcia/pinokio/api/comfyui11.git/app/custom_nodes/comfyui-photoshop I tried disabling all nodes to see if there was some conflict, but no.
@@fabiotgarcia2 thanks for the error code. I’ve tried but I can’t replicate the issue. It looks like a few users randomly encounter this problem, but I’m not sure why since the majority doesn’t. It could be a matter of having different dependencies installed. The main issue with troubleshooting SD errors is that everyone is running slightly different hardware and software, so it gets really hard to pinpoint where the issue lies if it’s not one of the usual errors.
Hello! Thank you very very much! The fact that it's possible to use Photoshop and transfer images back and forth is immediately clear. I also saw messages about using SD as a plugin. I haven't tried it. But working in real time is fantastic! I'm almost an artist now! 🤓
Thank you homie. Love your work brother.
awesome bro , keep up !!!!
Great work.. love it so much. keep up the good work.. creative for professionals
I cant tell how exited I am, I heart rate is faster noww
Thank you for this video, I am still using CS6, and I have a problem with the photoshop connection node, is there any solution for it. Your channel is really great, appreciate a lot.
thanks a lot for this great tutorial, i exactely folowed your Steps but get everytime the errormessage " Image Resize >> NoneType object has no attribute shape
Nice. Krita can do this with the Krita AI diffusion plug-in, and it’s open source! It also uses comfyui for the job.
Yeah a video on Krita has been on the list for a while now, but I need time to test it out since I mainly use software that is standard for production environments. I will do a tutorial on this on Krita in the future, though.
can this work if I'm running SD/ComfyUI on runpod or any other cloud GPU service? Or do I need to run SD/ComfyUI locally
I'm not sure there a way to tunnel the connection between PS as a server and comfy when using it on a cloud service. Unfortunately that's more of a sysadmin question, and I'm not qualified enough to give a proper answer
Would it un on Mac?
This is Amazing! thank you! Question is there a way to add a a custom image? what node should I use? and where to connect thank you :)
It depends on what you want to do with a custom image. This workflow pulls the currently active image from photoshop (if you have more than one open, it’s the one you’re currently working on). If you’d like, for example, to use an IPAdapter to influence the overall style of the generated image, you could do that with a custom image, or if you would like to start from another image to encode it into a latent and do a hybrid img2img process that would be possible as well. It all depends on which end result you’d like to use the custom image for. Let me know and I’ll see how I could help you with that.
@@risunobushi_ai Thank you for the reply, yes It does pull the image from photoshop but its not the same image anymore, background, face, is different, I wanted to have the same image, is there a reason why is that happening on my end?. thank you.
The generated image is always going to be at least a bit different from the original one in PS. If you wish for the generated image to be closer to the original image, you can try lowering the denoise value in the Ksampler. The lower the denoise is, the less changes are made to the original image.
If you’d like to change one aspect of the original image, like the background only for example, that needs a way more complex workflow than this, which is more intended for quick compositing in PS and having SD take care of merging everything together in a way that makes sense to the eye.
In the video, for example, I went through different examples at different denoising values to show how much the image changes depending on the denoise value. I hope this helps, otherwise let me know and I might do a more in-depth video about it!
@@risunobushi_ai Ahh isee, it would be a more complex node😅. Yes, thats what I met like change some aspect of the image like for example adding, watch, or glasses
it would be cool if you have a video of that as well thank you so much for your time and quick response I really appreaciate it. Keep up the good work I am your subscriber now! Ive learned alot from your videos.
duuuuude! your work is amazing!! do you also do After effect + SD?
Thanks! Unfortunately AE is a bit beyond my expertise, being a photographer I only direct videos, not actually edit them, but I’ll look into ways to implement SD in AE if possible a bit further down the road. Right now I’ll focus mainly on photography, illustration, architecture, art / creative direction, and 3d, which are at least somehow related to my area of expertise (or I have friends whose area of expertise is among those and I can ask for what they’d like to see and can give me real life feedback on what I can brew)
@@risunobushi_ai I am gonna try the way you show here. I do digital painting. I think I can do something cool with your way. Thank you so much. It is great inspiration. I am following you!
Great video! is there any way to connect comfyui to a lan photoshop client rather then having both run on the same computer ?
LAN or WAN? I’m not a sysadmin by any stretch of the definition, but If it’s LAN I don’t think there’s any issues, it should be recognizable on the same network without any open port I’m guessing. If it’s WAN, I’m sure there must be a way to forward the address on which the server is communicating.
@@risunobushi_ai Just a LAN setup when its at home and a VLAN when its outside, ill try to figure it out.
Too bad the connection cant just offload the actual photoshop link to comfy via the browser xD
It’s actually an interesting experiment to set up, so let me know if you find out how to make it work before I do, in the meantime I’ll try to do the same since I have a similar setup already in place
hello!and where is the link to the workflow?
You're right, I'm sorry! I updated the description with a wetransfer link to the json file, and I'll update it later with a more definitive link so it doesn't expire after one year.
The preview pop-up is not working. Perhaps because the photoshop plugin got update. I don't know.
I'm not sure I've asked you before, but what OS are you on? Was it working for you before?
Thanks for the heads up, I'll look into it. I've actually talked to the dev over the weekend, so I might get in touch with them as well.
Thanks for this tutorial. I followed your directions, but I get an all black screen in the pop up and preview image. Any idea why?
What's your setup? OS, python version, is comfyui updated, ecc? Some people have reported issues with the popup, but it appears to be happening only to a few
@@risunobushi_ai I'm using windows, python version 3.12. I just installed comfy ui for the first time yesterday. I'm not sure if that's the latest version or not. I would assume so.
Do you have all the dependencies installed? Like pillow, torch, torchvision, open-cv, and numpy. To check, go to comfyUI -> custom_nodes -> comfyui-popup_preview -> window, open in terminal, and run “pip install -r requirements.txt
By run, I mean type “pip install -r requirements.txt” without the “” in the terminal and press enter
@@risunobushi_ai "pip install -r requirements.txt
'"pip install -r requirements.txt' is not recognized as an internal or external command,
operable program or batch file
Can you tell me how to download the LORA LCM model from Huggingface? Apologies for the noob question. I tried downloading the pytorch_lora_weights.safetensors from the 'Files and Versions' tab, but that throws an error in Comfy. "Prompt outputs failed validation. LoraLoader: - Required input is missing: clip" Thank you kindly for the guide!
From that error it looks like you might not be linking the CLIP output from the Load Checkpoint node to the CLIP input in the Load LoRA node? I'm asking because if you downloaded that safetensor you should have the right file
@@risunobushi_ai thank you kindly for the reply. Yep, you're correct. Needed to connect clip. I initially loaded your workflow without further configuring, but comfy is so far out of my comfort zone, it's not even funny. I think I have it working now. Still have some tinkering to do. Really appreciate your videos. Fellow photographer here, trying to figure out how best to use these tools in my workflows
Hello！ yesterday I studied the video carefully, but the real-time preview screen (preview pop-up) was not available, Later, when I painted in PS, I could also generate a shape close to PS. I will try to use the light source on the pictures generated by this workflow Change (IC light, your latest video), the picture quality will become a little blurry, not as high-definition as before
I later shared the node about PhotoshoptocomfyUI in your introduction below to find a way to preview pop-up in real time, but I need to download the official version of Adobe creative cloud desktop. The PS I use is a pirated version. Is there any The method is that the pop-up preview interface appears.
Going in order:
- I have only tested the preview popup node with the official release, so I wouldn't know;
- if the results are blurry, your relight mask and your source image are different in size
@@risunobushi_ai I installed the official Creative cloud and installed the built-in ComfyUI for adobe photoshop plug-in, but the Pop up preview still did not pop up. I found in my comfyUI startup program that Installing requirements...
'pip' is not recognized as an internal or external command, operable program or batch file. How do I need to install it so that the pop-up window will appear? Thank you so much:)
@@yangchen-zd9zl pip usually comes preinstalled with the latest python releases. Try installing pip manually
@@risunobushi_ai Thank you for your guidance. I have now downloaded popup-preview and the pip problem has been solved. However, I found a very strange problem. That is, the color of the product I put in PS is yellow. The real-time preview The interface is blue, and the export from comfyUI is also blue. I don’t know how this problem occurred, but being able to experience the whole process makes me, a beginner, extremely happy. By the way, if I need to add a specific product to a model, this product cannot produce any changes, but needs to be properly integrated with the model. One way I know is to use Krita. Is there a better way to achieve it?
where to put these lcm / lora / files / how to install them?
Every model goes into the corresponding folder inside of your comfyUI -> model folder. So, for example, the LCM model is a checkpoint and goes into the checkpoint folder, a LoRA goes into the lora folder, if you had say a controlnet model it'd go inside the controlnet folder, etc. Once they're inside these folders, you just need to click "refresh" on the comfyUI sidebar and comfyUI will recognize that there's new models available.
@@risunobushi_ai thanks a lot!
my popup isn't popping up... after the render, there is no popup, no error, nothing. tried on firefox and chrome
What’s your OS?
@@risunobushi_ai windows 10
@@risunobushi_aiwindows 10. Firefox latest
Sorry, I didn’t see your answer. That’s weird, and it’s the second time I get this feedback from a windows user, where it should run without issues. I now the dev just updated his nodes, so I’m going to take a look at that next week. In the meantime, there’s a bit of a troubleshoot guide in the description of my other video about it, check if it works for you: th-cam.com/video/7G12SaLtEiA/w-d-xo.htmlsi=7ZHTgzB6Xg0SaTfP
Hi, I tried to install it on my Google Colab, I got this error, any idea how?
File "/content/drive/MyDrive/ComfyUI/custom_nodes/comfyui-photoshop/__init__.py", line 15, in
from photoshop import PhotoshopConnection
ModuleNotFoundError: No module named 'photoshop'
I’m not familiar with Collab, so I’m not sure I can be of much help there. Do you git pull the repo inside the custom_nodes folder like you’d do on any other OS?
@@risunobushi_ai I have successfully install all the notes except the Photoshop to ComfyUI. There were no issues with the remaining notes.
If you tried to install from comfyUI manager, the issue might be in the manager not finding the node in its repos. Try installing it directly from the custom nodes folder by opening the folder inside the terminal and using git pull “url from the repo”. The repo is in the video description.
@@risunobushi_ai do you mind to write the instruction on how to do this?
@@thisthisthis9444 You can find the infos in the description, but you basically want to head over to the github page for the node: github.com/NimaNzrii/comfyui-photoshop/issues
go to your comfyui - custom_nodes folder, right click, open in terminal, write "git clone" and the link to the node found on the green button in the github page, and then you'd need to downgrade the node (since it's been changed a lot because of an upgrade) by:
going in the comfyui - custom_nodes - comfyui-photoshop folder, right click, open in terminal, write "git checkout --force 403d4a9af1f947c95367cd40ff8ad6ae65e5df41"
this will downgrade the node.
Actually I thought I updated this description as well, turns out I updated another video but not this one. it's updated now.
can someone help ? Im missing MiDaS Depth Map node
Hey there! you can find it here: github.com/Fannovel16/comfyui_controlnet_aux
i manually isntalled the comfyui-photoshop node but it still does not show in my workflow? any idea whats wrong?
No worries, let’s see if I can help troubleshoot it. Did you git pull from the repo inside the custom-nodes folder? Did you restart comfyUI by any chance after you installed it? ComfyUI doesn’t see new nodes unless it restarts.
@@risunobushi_ai i did :(
Would you mind deleting the node from the custom node folder and try to install it via the manager? It’s under install custom nodes -> search for “photoshop” -> #384, by NimaNzrii. That might help.
@@risunobushi_ai i tried this, it shows that the node is installed but still doesnt come up in the workflow. i cant search for the node nor can i find it in the add node menu.
That’s really weird, I’m sorry it’s taking so long to solve. Have you tried importing my json file? Do you see the node if you do?
wow
I , i am a new user, and when i try my first generation i got an error like this
Error occurred when executing KSampler:
mat1 and mat2 shapes cannot be multiplied (154x2048 and 768x320)
😂
Hey there! Would you mind sharing your workflow in .json form? I can take a look at it later and check what’s wrong
hi could you please help me to install photoshop its not installing could you please make a video ??
I’m sorry, but I can’t make a video on installing photoshop. I’m sure there’s plenty of them on TH-cam already though!
@@risunobushi_ai could you suggest me any video where explain how to install Photoshop?
@@risunobushi_ai i tried to install photoshop in comfy ui but its not installing
Oh you mean the Photoshop to comfyUI node?
If you can’t install it via ComfyUI manager, you need to go to your comfyUI folder -> custom_nodes -> open in terminal -> type “git clone press Enter.
What OS are you running? macOS gives some issues sometimes, while it works without issues on Windows and Linux.
I have windows system when i paste custom node and my termination saying install unsuccessfully you want to remove photoshop? Say yes or no.... Not installing please help
Nice, try to use integrations with software easy to "find". XD. Always supporting the poor :)
great need more video regarding fashion with AI , If do you have any personal contact please share i can suggest you some of content which we are really needed as designer
You can send me an email at andrea@andreabaioni.com
That photoshop node only works on Windows as it's trying to install its dependencies with "python.exe" 😅
Also the photoshop python package it's using requires 'winreg' which surely won't run on mac or linux.
Oh so that’s the source of all the macOS issues and why most of the users weren’t having any troubles with it! Great to know, I’ll make an update about it so that Mac users know about it.
Also yeah, support and docs are non-existent with this node, but it’s the most straightforward I’ve seen that does what’s needed, such a shame it’s not polished.
I might try to fork it and see if it’s possible to make it work on both Win / macOS.
@@risunobushi_ai looking forward for this!!! Thanks!
Hi Andrea! Thanks so much for sharing the useful tutorial. But for me the PhotoshopToComfyUI it´s getting an error. I'm on Macbook M2 Pro Max.
Hey there! I’ll try to replicate the issue tomorrow on my MacBook, I’ll see if it’s native to the OS or it’s because of something else. Could you paste the error code?
@@risunobushi_ai 0.0 seconds (IMPORT FAILED): /Users/fabiogarcia/pinokio/api/comfyui11.git/app/custom_nodes/comfyui-photoshop
I tried disabling all nodes to see if there was some conflict, but no.
@@risunobushi_ai Hi Andrea. The error I said is that it´s failed to import. Did you have a chance to test it in your MacBook?
@@fabiotgarcia2 thanks for the error code. I’ve tried but I can’t replicate the issue. It looks like a few users randomly encounter this problem, but I’m not sure why since the majority doesn’t. It could be a matter of having different dependencies installed. The main issue with troubleshooting SD errors is that everyone is running slightly different hardware and software, so it gets really hard to pinpoint where the issue lies if it’s not one of the usual errors.
@@risunobushi_ai Maybe it's incompatibility between ComfyUI version. I am using ComfyUI 1.1