IDM VTON on comfyUI for Realistic Clothing on Digital Models + Workflow
ฝัง
- เผยแพร่เมื่อ 14 พ.ค. 2024
- 🌟 Visite for Latest AI Digital Models Workflows: aiconomist.gumroad.com/
IDM VTON for Realistic Clothing on Digital Models + ComfyUI Workflow
Discover the secrets of enhancing digital character outfits using IDM VTON on ComfyUI. This tutorial will teach you the techniques to bring your digital fashion creations to life.
Part 1: • IDM VTON for Realistic...
🌟 Checkpoint Model: civitai.com/models/112902/dre...
🌟 Workflow + 16 Garment + 45 AI Character: aiconomist.gumroad.com/l/idm-...
🌟 IDM VTON ComfyUI: github.com/TemryL/ComfyUI-IDM...
ComfyUI Custom Nodes:
🌟 ComfyUI Impact Pack : github.com/ltdrdata/ComfyUI-I...
🌟 ComfyUI's ControlNet Auxiliary Preprocessors : github.com/Fannovel16/comfyui...
🌟 ComfyUI_IPAdapter_plus : github.com/cubiq/ComfyUI_IPAd...
🌟 segment anything : github.com/storyicon/comfyui_...
🌟 ComfyUI Nodes for Inference : github.com/LykosAI/ComfyUI-In...
How to Install ComfyUI on Windows: • ComfyUI on Windows: EA...
𝕏 Follow me : X.com/Aiconomist1
💼 For Business: aiconomistbusiness@gmail.com - วิทยาศาสตร์และเทคโนโลยี
Hey everyone! 📢 If you encounter any errors or issues, please send me screenshots via the email provided in the video description. This way, I can assist you better and more efficiently. It's challenging to troubleshoot through text alone, so your cooperation is greatly appreciated! 🙏
This is great. thank you. I talked to a few fashion design houses and it is essential for them where the model is wearing the exact version of their outfit, something that is getting closet but not in the nature of diffusion models.
Thanks for such a pure & Useful Tutorial. 👍💯
Does anybody have the same issue with memory? How to fix it?
Error occurred when executing PipelineLoader:
Allocation on device 0 would exceed allowed memory. (out of memory)
Currently allocated : 15.56 GiB
Requested : 12.50 MiB
Device limit : 15.70 GiB
Free (according to CUDA): 1.81 MiB
PyTorch limit (set by user-supplied memory fraction)
: 17179869184.00 GiB
RuntimeError:
"CUDA Setup failed despite GPU being available." Something about CUDA libs needing to be added to PATH. Windows 11.
how to solve this problem?
Error occurred when executing GroundingDinoSAMSegment (segment anything):
Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument mat2 in method wrapper_CUDA_mm)
Hello, I bought the workfow but, eventhoguh I have installed hd-nodes-comfyui it doesn't work.
Hi great tutorial. The problem is that on Kaggle need more Ram. I see that the first Generation download about 40GB and the image generation is stopped
I’m curious to know why does the model images need to be a specific width and height?
Hey, your videos are so helpful. Can you make a video about how to run this workflow in runpod.
This workflow is crashing on the image receiver. Because it checks before runtime not during. I have to start the workflow again and then it continues. Am i doing something wrong?
24GB VRAM its too much
Can you demonstrate how to change the background of a portrait photo naturally? I’m currently doing simple compositing, but the shadows and lighting look very unnatural.
I get this error :
Error occurred when executing DensePosePreprocessor
Is this methode same gd as the methode of Hugingface or this is better. Here I see more control in backend..but if it does better Job!
Hey, great workflow, but i can't get it to run through.
I am getting errors at the VTON pipeline node (which is looking different, it has ckpt_path field with path to: models/idm_vton).
Error occurred when executing PipelineLoader:
Error no file named scheduler_config.json found in directory models/idm_vton.
Am i missing something?
hello, have you find a solution? i'm getting the same problem
Update your node from the comfyui manager
I got the same problem today. It was OK yesterday at my 2nd try. The 1st try ended up with an error message with float32 but I changed to float16, it worked and I saved the workflow to do today, but the workflow doesn't work today
Btw, updating nodes doesn't work
Great tutorial. The only problem for me is the performance. Even with a 4080 16GB vram it is taking forever to render.
How long does it take to run whole pipeline on 3090? After initial run, when everything is loaded
40 - 44 sec for IDM VTON + 20 sec for the refining part. i guess 1min
@@Aiconomist I have been getting missing files error as comfy is not loading the necessary models and files into this custom nodes' model folder. Do you know the fix?
Hey , as you did say in comments of your last video that you are going to launch a course on " AI influencer " I'm just curious when you are going to launch it ?
Thanks for asking about the AI influencer course. working on it while juggling my full-time job, but I'm making it beginner friendly. Updates coming soon on X .com or on TH-cam community.
@@Aiconomistthanks for the update. I am also interested in this course
Why did you remove CannyEdge from previous workflow?
you're absolutely right i overlooked including canny controlnet in the final video recording. i believe its addition would have significantly enhanced the quality of the result.
@@AiconomistThat would be great to try!! Could you please update the download workflow json file with Cnet?
WHERE DO YOU GET THE JSON FILE OF THE WORKFLOW
?
Great tutorial, somehow for me it takes ages. The IDM-VTON Interface needs 6 Minutes for a single Image. I am on a 4080 usually rendering 8 Images in SDXL 1024 x 1024 in 45 seconds.
yes do u find any solution for that
@@alchemy4327 I use Idapter V2 instead.
do one with a1111