Wow, wow, wow... I gotta say, that's a brilliant explanation and showcase of different takes on image generation workflow in ComfyUI. I've followed your tutorials closely and learned much about the different aspects of AI image generation in ComfyUI and using various nodes. Keep up the good work!
I like how you go over each setting in detail, it might be lengthy and boring to some people but damnit I needed all that info like months ago, other creators just pass over them quickly or just don't talk about them at all👍👍👍
The colour blending worked very badly for me, i got no extra colour on the B/W photos whatsoever - just some strange mioire effects. the Canny and depthmap was very enjoyable and informative though.
Your channel has very detailed and informative lessons👍. I would support you financially, but transactions from my country are blocked because we are unsuccessfully trying to destroy the entire civilized world😏. So just a huge thank you for your hard work🙏
Ohh I totally get it what you are saying, I never agree with these blocking behaviours. Thank you so much!! Word of support just made my day. 🫶🏼. If you need anything specific, let me know. We are always looking for ideas from and for the users who watch the channel. 🙂
@@controlaltai I am still learning the basics and can hardly offer anything specific. It's a shame that such professionals (deeply versed in the topic of ai) as you are not popular enough in contrast to channels with all sorts of nonsense. So I just expressed my gratitude for sharing your vast knowledge with us, gave you a Like👍 and wrote a comment to promote your channel on the TH-cam platform.
Great video, thank you. I have a question, in your canny/depth set up with 2 image sources, which image/controlnet, becomes the main image and which becomes the accent. How do you define the accent? In your example it was the canny with “a pirate” defined in the positive prompt. Would there be a time where the depth would act as the accent and the canny the main? Like if you can explain how with this set up, you organize the main and the accent.
Hi, the first image node going through the pass through via imagescaletototalpixels node becomes main and other becomes accent (secondary). For image you can define primary via making it pass through this node. For controlnet is more about the control weight. In the same example where image 1 via cn1 is primary and img2 is secondary if you make cn2 same weight as cn1 you will just still have a very strong outline superimposed over img1. This will be canny. That's why I reduced the weight of cn2 so it can blend very well. The results would be very different if you use a latent image as a source instead of passing the image through the imagescaletototalpixels node.
was trying to work in VR, and its obvious need UI specific for VR. I would imagine like sorceres game, where you see ingrediens and throwing everything in one pot 😅
Thank you, Have not tried it but maybe this helps: cmu-perceptual-computing-lab.github.io/openpose/web/html/doc/md_doc_02_output.html#autotoc_md40 Refer to the json output section, and this: github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/include/openpose/flags.hpp For the keypoint_scale
Thank you for the support! You are welcome. If you need something specific let me know. Taking request from members, will start making specific workflow tutorials.
Can you provide an explanation for converting any image into a drawing for coloring? Any explanation intended for coloring book makers with the addition of inscriptions over the pictures, especially the necessity of the quality of the pictures and the clarity of the lines
Well you can just generate a drawing directly for coloring books. It's not actually required to take an existing image and then convert it to drawing. AI comes handy in two aspects. One: you physically draw something, sketch, and instead of spending hours/days to finish it, you can use the AI to convert your drawings into art. Second: Use the AI to just generate the drawing for making a coloring book, for commercial or personal use. This can be done via Stable Diffusion, Blue Willow or MidJourney.
@Mehdi0montahw the prompt depends on the platform, what platform are you using? For blue willow, I have made a tutorial here: th-cam.com/video/K4fZW6dS9DY/w-d-xo.htmlsi=hpbyhwiw6BAA8qJZ, for stable diffusion check out civit ai: civitai.com/search/models?sortBy=models_v3&query=Coloring%20book
@@controlaltai I only use Stable Diffusion and ComfyUI And focus on them .Thank you. I will try the model you suggested, and if you come up with a way to remove the blackness in the majority of the images produced, please share it with us.
Thanks for the tutorial, but is there a way to use it on mac? I downloaded everything but mac cannot install the .bat file and there are no "preprocessor" nodes in my Comfy UI. Do you have any ideas about this?
I had an error when running the first depth controlnet workflow. When the program wants to run KSampler advanced, it released this error line: 'ModuleList' object has no attribute '1' can you help me fix it?
problem solved. It was because the checkpoint I use is from SD 1.5 . The checkpoint use for this tutorial must be based on SDXL because the controlnet used is also from SDXL.
Hello, please when i tried queue prompting the midas depth map i got this error (Error occurred when executing MiDaS-DepthMapPreprocessor: PytorchStreamReader failed reading zip archive: failed finding central directory) I have tried checking how to resolve this but haven't been able to
@@controlaltai sorry i have tried searching , please where can i download the depth map model and which folder do i put it in...I have tried several ways to search for the midas but havent't been able to find it
@@abiodunshonibare832 Here is the link to the dept model: huggingface.co/stabilityai/control-lora/tree/main/control-LoRAs-rank256 This will go in the following folder: ComfyUI_windows_portable\ComfyUI\models\controlnet
@@controlaltai thank you for your response, I found out what the problem was yesterday. So when loading the Midas depth for the first time it tends to download first, that initial download after I clicked queue prompt wasn’t complete hence why I got that error , so I had to delete the incomplete file and run it again, then it worked
Thanks, but i have this problem when i try to fix problems in "python embeded folder" it says this: Fatal Python error: init_fs_encoding: failed to get the Python codec of the filesystem encoding Python runtime state: core initialized ModuleNotFoundError: No module named 'encodings' Current thread 0x00002900 (most recent call first): I have update comfyui and python dependencies.
Very hard to diagnose this without knowing the environment you are running it on. Simple solution download the latest portable version of com in a new folder and run it from there to see if the problem is with the existing environment. If the new folder comfy works, just port models, output folder, and manually re install each custom node.
*Hello G. Seth when I run the workflow that you show us in the video, this error appears in the terminal, could you please help me resolve it. Appears when it reach the KSampler node (Advanced)* ERROR:root:!!! Exception during processing !!! ERROR:root:Traceback (most recent call last): File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\execution.py", line 152, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\execution.py", line 82, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\execution.py", line 75, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i))) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI odes.py", line 1402, in sample return common_ksampler(model, noise_seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise, disable_noise=disable_noise, start_step=start_at_step, last_step=end_at_step, force_full_denoise=force_full_denoise) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI odes.py", line 1338, in common_ksampler samples = comfy.sample.sample(model, noise, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Impact-Pack\modules\impact\sample_error_enhancer.py", line 9, in informative_sample return original_sample(*args, **kwargs) # This code helps interpret error messages that occur within exceptions but does not have any impact on other operations. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\sample.py", line 100, in sample samples = sampler.sample(noise, positive_copy, negative_copy, cfg=cfg, latent_image=latent_image, start_step=start_step, last_step=last_step, force_full_denoise=force_full_denoise, denoise_mask=noise_mask, sigmas=sigmas, callback=callback, disable_pbar=disable_pbar, seed=seed) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 703, in sample return sample(self.model, noise, positive, negative, cfg, self.device, sampler, sigmas, self.model_options, latent_image=latent_image, denoise_mask=denoise_mask, callback=callback, disable_pbar=disable_pbar, seed=seed) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 601, in sample pre_run_control(model, negative + positive) File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 452, in pre_run_control x['control'].pre_run(model, percent_to_timestep_function) File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\controlnet.py", line 296, in pre_run comfy.utils.set_attr_param(self.control_model, k, self.control_weights[k].to(dtype).to(comfy.model_management.get_torch_device())) File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\utils.py", line 301, in set_attr_param return set_attr(obj, attr, torch.nn.Parameter(value, requires_grad=False)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\utils.py", line 295, in set_attr obj = getattr(obj, name) ^^^^^^^^^^^^^^^^^^ File "C:\Users iver\AI\ComfyUI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\torch n\modules\module.py", line 1688, in __getattr__ raise AttributeError(f"'{type(self).__name__}' object has no attribute '{name}'") AttributeError: 'ModuleList' object has no attribute '1' Prompt executed in 12.69 seconds
Brilliant and no fuss work. Keep it up bro. With this quality of tutorial, you will outgrow any major channel...
Wow, wow, wow... I gotta say, that's a brilliant explanation and showcase of different takes on image generation workflow in ComfyUI. I've followed your tutorials closely and learned much about the different aspects of AI image generation in ComfyUI and using various nodes. Keep up the good work!
Thank you!!
I like how you go over each setting in detail, it might be lengthy and boring to some people but damnit I needed all that info like months ago, other creators just pass over them quickly or just don't talk about them at all👍👍👍
Ahh thank you!!
The colour blending worked very badly for me, i got no extra colour on the B/W photos whatsoever - just some strange mioire effects. the Canny and depthmap was very enjoyable and informative though.
Your channel has very detailed and informative lessons👍. I would support you financially, but transactions from my country are blocked because we are unsuccessfully trying to destroy the entire civilized world😏. So just a huge thank you for your hard work🙏
Ohh I totally get it what you are saying, I never agree with these blocking behaviours. Thank you so much!! Word of support just made my day. 🫶🏼. If you need anything specific, let me know. We are always looking for ideas from and for the users who watch the channel. 🙂
@@controlaltai I am still learning the basics and can hardly offer anything specific. It's a shame that such professionals (deeply versed in the topic of ai) as you are not popular enough in contrast to channels with all sorts of nonsense. So I just expressed my gratitude for sharing your vast knowledge with us, gave you a Like👍 and wrote a comment to promote your channel on the TH-cam platform.
Would also love to see more on SEGS in controlnet....
Will put that in the todo list. Thanks!
We require a professional episode on converting images to lineart while completely removing the black and gray parts
I will try and see if I am able to do so.
@@controlaltai Thank you for your response and interest in your followers’ requests
Great video, thank you. I have a question, in your canny/depth set up with 2 image sources, which image/controlnet, becomes the main image and which becomes the accent.
How do you define the accent?
In your example it was the canny with “a pirate” defined in the positive prompt. Would there be a time where the depth would act as the accent and the canny the main?
Like if you can explain how with this set up, you organize the main and the accent.
Hi, the first image node going through the pass through via imagescaletototalpixels node becomes main and other becomes accent (secondary).
For image you can define primary via making it pass through this node.
For controlnet is more about the control weight.
In the same example where image 1 via cn1 is primary and img2 is secondary if you make cn2 same weight as cn1 you will just still have a very strong outline superimposed over img1. This will be canny. That's why I reduced the weight of cn2 so it can blend very well.
The results would be very different if you use a latent image as a source instead of passing the image through the imagescaletototalpixels node.
was trying to work in VR, and its obvious need UI specific for VR.
I would imagine like sorceres game, where you see ingrediens and throwing everything in one pot 😅
Great tutorial. Any ideas on how to load JSON files for openpose? The Apply ControlNet node obviously does not accept JSON files, only images.
Thank you, Have not tried it but maybe this helps: cmu-perceptual-computing-lab.github.io/openpose/web/html/doc/md_doc_02_output.html#autotoc_md40
Refer to the json output section, and this:
github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/include/openpose/flags.hpp
For the keypoint_scale
It was a very helpful tutorial, thank you for dedicating the time and explanation. You help us a lot!
Thank you for the support! You are welcome. If you need something specific let me know. Taking request from members, will start making specific workflow tutorials.
Impressive!
Can you provide an explanation for converting any image into a drawing for coloring? Any explanation intended for coloring book makers with the addition of inscriptions over the pictures, especially the necessity of the quality of the pictures and the clarity of the lines
Well you can just generate a drawing directly for coloring books. It's not actually required to take an existing image and then convert it to drawing. AI comes handy in two aspects.
One: you physically draw something, sketch, and instead of spending hours/days to finish it, you can use the AI to convert your drawings into art.
Second: Use the AI to just generate the drawing for making a coloring book, for commercial or personal use. This can be done via Stable Diffusion, Blue Willow or MidJourney.
@@controlaltai I did that, but I want clear drawing lines and a solution to the problem of black shadows inside the drawings. Any advice?
@Mehdi0montahw the prompt depends on the platform, what platform are you using? For blue willow, I have made a tutorial here: th-cam.com/video/K4fZW6dS9DY/w-d-xo.htmlsi=hpbyhwiw6BAA8qJZ, for stable diffusion check out civit ai: civitai.com/search/models?sortBy=models_v3&query=Coloring%20book
@@controlaltai I only use Stable Diffusion and ComfyUI And focus on them .Thank you. I will try the model you suggested, and if you come up with a way to remove the blackness in the majority of the images produced, please share it with us.
@@Mehdi0montahw Sure will do. 👍
Excuse me, do you have any tutorial on how to make a batch of multiple depth or softedge images per image?
You just connect the load image node with the required ore processors and the save nodes.
Thanks for the tutorial, but is there a way to use it on mac? I downloaded everything but mac cannot install the .bat file and there are no "preprocessor" nodes in my Comfy UI. Do you have any ideas about this?
Welcome and Sorry, I don’t have any idea about mac. Checkout out the GitHub repository for mac support.
I can't find the colorcorrect node, and I've also searched in the manager, how can I get this node?
Check for comfy ui post processing nodes. Its there. Refer here 22:53
@@controlaltaiOH！！！！ Thanks a million for adding sunshine to my day!
I had an error when running the first depth controlnet workflow. When the program wants to run KSampler advanced, it released this error line:
'ModuleList' object has no attribute '1'
can you help me fix it?
problem solved. It was because the checkpoint I use is from SD 1.5 . The checkpoint use for this tutorial must be based on SDXL because the controlnet used is also from SDXL.
Hi, yes. I am using control lora from stability ai for the ControlNet, they are all SDXL base. Checkpoint is also sdxl base.
Hello, please when i tried queue prompting the midas depth map i got this error (Error occurred when executing MiDaS-DepthMapPreprocessor:
PytorchStreamReader failed reading zip archive: failed finding central directory) I have tried checking how to resolve this but haven't been able to
No idea what this error is, have you downloaded the depth map model correctly and placed it in the ControlNet folder?
@@controlaltai sorry i have tried searching , please where can i download the depth map model and which folder do i put it in...I have tried several ways to search for the midas but havent't been able to find it
@@abiodunshonibare832 Here is the link to the dept model: huggingface.co/stabilityai/control-lora/tree/main/control-LoRAs-rank256 This will go in the following folder: ComfyUI_windows_portable\ComfyUI\models\controlnet
@@controlaltai thank you for your response, I found out what the problem was yesterday. So when loading the Midas depth for the first time it tends to download first, that initial download after I clicked queue prompt wasn’t complete hence why I got that error , so I had to delete the incomplete file and run it again, then it worked
Thanks, but i have this problem when i try to fix problems in "python embeded folder" it says this:
Fatal Python error: init_fs_encoding: failed to get the Python codec of the filesystem encoding
Python runtime state: core initialized
ModuleNotFoundError: No module named 'encodings'
Current thread 0x00002900 (most recent call first):
I have update comfyui and python dependencies.
Very hard to diagnose this without knowing the environment you are running it on. Simple solution download the latest portable version of com in a new folder and run it from there to see if the problem is with the existing environment. If the new folder comfy works, just port models, output folder, and manually re install each custom node.
@@controlaltai Ok, thanks, i'm going to try, but where is tor what is he latest version?
Whatever version is there on GitHub, download that. For python I recommend 3.11 version only.
@@controlaltai Thanks, it´s possible to mave consistent characters in Comfy UI? Because i want to make a comic with consistent characters.
Yes, consistent characters are very much possible in comfyui. You just need to get a workflow correct.
*Hello G. Seth
when I run the workflow that you show us in the video, this error appears in the terminal, could you please help me resolve it.
Appears when it reach the KSampler node (Advanced)*
ERROR:root:!!! Exception during processing !!!
ERROR:root:Traceback (most recent call last):
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\execution.py", line 152, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\execution.py", line 82, in get_output_data
return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\execution.py", line 75, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI
odes.py", line 1402, in sample
return common_ksampler(model, noise_seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise, disable_noise=disable_noise, start_step=start_at_step, last_step=end_at_step, force_full_denoise=force_full_denoise)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI
odes.py", line 1338, in common_ksampler
samples = comfy.sample.sample(model, noise, steps, cfg, sampler_name, scheduler, positive, negative, latent_image,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Impact-Pack\modules\impact\sample_error_enhancer.py", line 9, in informative_sample
return original_sample(*args, **kwargs) # This code helps interpret error messages that occur within exceptions but does not have any impact on other operations.
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\sample.py", line 100, in sample
samples = sampler.sample(noise, positive_copy, negative_copy, cfg=cfg, latent_image=latent_image, start_step=start_step, last_step=last_step, force_full_denoise=force_full_denoise, denoise_mask=noise_mask, sigmas=sigmas, callback=callback, disable_pbar=disable_pbar, seed=seed)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 703, in sample
return sample(self.model, noise, positive, negative, cfg, self.device, sampler, sigmas, self.model_options, latent_image=latent_image, denoise_mask=denoise_mask, callback=callback, disable_pbar=disable_pbar, seed=seed)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 601, in sample
pre_run_control(model, negative + positive)
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\samplers.py", line 452, in pre_run_control
x['control'].pre_run(model, percent_to_timestep_function)
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\controlnet.py", line 296, in pre_run
comfy.utils.set_attr_param(self.control_model, k, self.control_weights[k].to(dtype).to(comfy.model_management.get_torch_device()))
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\utils.py", line 301, in set_attr_param
return set_attr(obj, attr, torch.nn.Parameter(value, requires_grad=False))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\utils.py", line 295, in set_attr
obj = getattr(obj, name)
^^^^^^^^^^^^^^^^^^
File "C:\Users
iver\AI\ComfyUI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\torch
n\modules\module.py", line 1688, in __getattr__
raise AttributeError(f"'{type(self).__name__}' object has no attribute '{name}'")
AttributeError: 'ModuleList' object has no attribute '1'
Prompt executed in 12.69 seconds
Hard to diagnose like this without checking the workflow. Can you check the controlnet model and checkpoint, both should be SDXl.
Could you share the configuration with an image?
Hi, all workflows are shared for channel members via community post. I hope you understand, Thank You! 🙏
d