How would i go about adding a "face fix" of sorts? that automatically scans for faces and improves them?

After searching around a bit i landed on something called "ComfyUI-Impact-Pack" because that apparently has the best "face fix" stuff out there for ComfyUI, but i'm completely lost on trying to tie that to the end of your workflow (in part because their examples are just as crap as every other ComfyUI example.. it's referring nodes i don't even know where to get?...)

Any chance you can elaborate on how YOU would automatically fix faces with ComfyUI (+ anything) and your workflow? and maybe show a workflow that includes this?

Thanks!

I'm super interested in a FaceDetailer node integrated into this amazing workflow! I have put out a bounty for someone to extend the workflow to include one that includes lora loading with separate text prompts (hopefully).

I've put out a 500 buzz bounty for this and I would love if @searge was able to get the reward and integrate it into the official workflow!

Thanks again for all the awesome work!

Error occurred when executing SeargeFreeU: SeargeFreeU.get() missing 1 required positional argument: 'freeu_version' File "C:\ComfyUI_windows_portable\ComfyUI\execution.py", line 153, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) File "C:\ComfyUI_windows_portable\ComfyUI\execution.py", line 83, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) File "C:\ComfyUI_windows_portable\ComfyUI\execution.py", line 76, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))

ERROR:root:!!! Exception during processing !!!

ERROR:root:Traceback (most recent call last):

File "S:\Comfy\ComfyUI\execution.py", line 153, in recursive_execute

output_data, output_ui = get_output_data(obj, input_data_all)

File "S:\Comfy\ComfyUI\execution.py", line 83, in get_output_data

return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)

File "S:\Comfy\ComfyUI\execution.py", line 76, in map_node_over_list

results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))

File "S:\Comfy\ComfyUI\custom_nodes\SeargeSDXL\modules\magic_box.py", line 269, in process

(data, stage_result) = self.run_stage(stage, data, stage_input)

File "S:\Comfy\ComfyUI\custom_nodes\SeargeSDXL\modules\magic_box.py", line 248, in run_stage

(data, stage_result) = stage_processor.process(data, stage_input)

File "S:\Comfy\ComfyUI\custom_nodes\SeargeSDXL\modules\stage_vae_decode_sampled.py", line 77, in process

image = NodeWrapper.vae_decoder.decode(vae_model, latent)[0]

File "S:\Comfy\ComfyUI\nodes.py", line 267, in decode

return (vae.decode(samples["samples"]), )

AttributeError: 'NoneType' object has no attribute 'decode'

Hey - gettin all these errors:

Error occurred when executing SeargeMagicBox: module 'comfy.sample' has no attribute 'broadcast_cond' File "/Users/aibrain/ComfyUI/execution.py", line 153, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/aibrain/ComfyUI/execution.py", line 83, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/aibrain/ComfyUI/execution.py", line 76, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i))) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/aibrain/ComfyUI/custom_nodes/SeargeSDXL/modules/magic_box.py", line 269, in process (data, stage_result) = self.run_stage(stage, data, stage_input) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/aibrain/ComfyUI/custom_nodes/SeargeSDXL/modules/magic_box.py", line 248, in run_stage (data, stage_result) = stage_processor.process(data, stage_input) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/aibrain/ComfyUI/custom_nodes/SeargeSDXL/modules/stage_sampling.py", line 126, in process latent = sampler(base_model, base_positive, base_negative, latent, seed, steps, cfg, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/aibrain/ComfyUI/custom_nodes/SeargeSDXL/modules/node_wrapper.py", line 99, in sdxl_sampler result = sdxl_ksampler(base_model, refiner_model, noise_seed, base_steps, refiner_steps, cfg, sampler_name, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/aibrain/ComfyUI/custom_nodes/SeargeSDXL/modules/custom_sdxl_ksampler.py", line 356, in sdxl_ksampler samples = sdxl_sample(base_model, refiner_model, noise, base_steps, refiner_steps, cfg, sampler_name, scheduler, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/aibrain/ComfyUI/custom_nodes/SeargeSDXL/modules/custom_sdxl_ksampler.py", line 184, in sdxl_sample pos_base_copy = comfy.sample.broadcast_cond(base_positive, noise.shape[0], device)

Previous version was working fine and dandy - cheers

Getting the following error using this in Manjaro - Arch Linux - Error occurred when executing SeargeMagicBox:

module 'comfy.sample' has no attribute 'broadcast_cond' File

"/home/zono50/Applications/ComfyUI/execution.py", line 153, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) File "/home/zono50/Applications/ComfyUI/execution.py", line 83, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) File "/home/zono50/Applications/ComfyUI/execution.py", line 76, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i))) File "/home/zono50/Applications/ComfyUI/custom_nodes/SeargeSDXL/modules/magic_box.py", line 269, in process (data, stage_result) = self.run_stage(stage, data, stage_input) File "/home/zono50/Applications/ComfyUI/custom_nodes/SeargeSDXL/modules/magic_box.py", line 248, in run_stage (data, stage_result) = stage_processor.process(data, stage_input) File "/home/zono50/Applications/ComfyUI/custom_nodes/SeargeSDXL/modules/stage_sampling.py", line 126, in process latent = sampler(base_model, base_positive, base_negative, latent, seed, steps, cfg, File "/home/zono50/Applications/ComfyUI/custom_nodes/SeargeSDXL/modules/node_wrapper.py", line 94, in sdxl_sampler result = sdxl_ksampler(base_model, None, noise_seed, base_steps, 0, cfg, sampler_name, File "/home/zono50/Applications/ComfyUI/custom_nodes/SeargeSDXL/modules/custom_sdxl_ksampler.py", line 356, in sdxl_ksampler samples = sdxl_sample(base_model, refiner_model, noise, base_steps, refiner_steps, cfg, sampler_name, scheduler, File "/home/zono50/Applications/ComfyUI/custom_nodes/SeargeSDXL/modules/custom_sdxl_ksampler.py", line 184, in sdxl_sample pos_base_copy = comfy.sample.broadcast_cond(base_positive, noise.shape[0], device)

The error that most of you reported is most likely related to an improper file size of the SD_XL models downloaded by the installer. In my case, they were much smaller than expected. I downloaded the models separately and it now works just fine.

Also, the same error occurs if you try to load an SD_1.5 model - it doesn't work with it (it does not support 1.5 by design)!

I am a huge fan of you bro. Your works SeargeXL and MCP are equivalently awesome. Pure applauses for you.

I run gloal Gen AI business and want to suggest a collaborative research opportunities.

Our goal is to make 'Midjourney for human portraits' and thought you might be interested.

Would you contact me for the detail?

- mail: [email protected]

- discord: eric_sdxl

First of all huge thanks for making this - upped my image quality by 200% :)

Is it possible to tell your default save node where to save the images? as in the folder? I noticed the "convert save folder to input" but cant figure out how to tell him the path.

NVM

Putting this into a .bat file worked just fine.
"python main.py --output-directory xxxx"

This is incredible. Thank you so much!

First of all, thanks for all the hard work, this is great!
Just one thing. I like the idea of saving params to a separate text file because it makes it easier to organize and analyze generations. But I don't like how it clutters the image folder. Any chance you could add an optional setting to set a log folder? Like relative to where the image is saved. If it was just like "./params" and then it would create a "params" folder and put the text files there, that would be perfect.

i get this error message cant fix this can someone help please?

MagicBox: MPS backend out of memory (MPS allocated: 16.34 GB, other allocations: 1.38 GB, max allowed: 18.13 GB). Tried to allocate 620.16 MB on private pool. Use PYTORCH_MPS_HIGH_WATERMARK_RATIO=0.0 to disable upper limit for memory allocations (may cause system failure). File "/Users/artem/ComfyUI/execution.py", line 153, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/execution.py", line 83, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/execution.py", line 76, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i))) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/custom_nodes/SeargeSDXL/modules/magic_box.py", line 269, in process (data, stage_result) = self.run_stage(stage, data, stage_input) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/custom_nodes/SeargeSDXL/modules/magic_box.py", line 248, in run_stage (data, stage_result) = stage_processor.process(data, stage_input) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/custom_nodes/SeargeSDXL/modules/stage_sampling.py", line 126, in process latent = sampler(base_model, base_positive, base_negative, latent, seed, steps, cfg, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/custom_nodes/SeargeSDXL/modules/node_wrapper.py", line 99, in sdxl_sampler result = sdxl_ksampler(base_model, refiner_model, noise_seed, base_steps, refiner_steps, cfg, sampler_name, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/custom_nodes/SeargeSDXL/modules/custom_sdxl_ksampler.py", line 356, in sdxl_ksampler samples = sdxl_sample(base_model, refiner_model, noise, base_steps, refiner_steps, cfg, sampler_name, scheduler, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/custom_nodes/SeargeSDXL/modules/custom_sdxl_ksampler.py", line 190, in sdxl_sample base_samples = base_sampler.sample(noise, pos_base_copy, neg_base_copy, cfg=cfg, latent_image=latent_image, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/samplers.py", line 691, 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 "/Users/artem/ComfyUI/comfy/samplers.py", line 597, in sample samples = sampler.sample(model_wrap, sigmas, extra_args, callback, noise, latent_image, denoise_mask, disable_pbar) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/samplers.py", line 557, in sample samples = getattr(k_diffusion_sampling, "sample_{}".format(sampler_name))(model_k, noise, sigmas, extra_args=extra_args, callback=k_callback, disable=disable_pbar, **extra_options) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/utils/_contextlib.py", line 115, in decorate_context return func(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/k_diffusion/sampling.py", line 580, in sample_dpmpp_2m denoised = model(x, sigmas[i] s_in, *extra_args) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1519, in wrappedcall_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1528, in callimpl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/samplers.py", line 274, in forward out = self.inner_model(x, sigma, cond=cond, uncond=uncond, cond_scale=cond_scale, model_options=model_options, seed=seed) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1519, in wrappedcall_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1528, in callimpl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/samplers.py", line 264, in forward return self.apply_model(*args, kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/samplers.py", line 261, in apply_model out = sampling_function(self.inner_model.apply_model, x, timestep, uncond, cond, cond_scale, model_options=model_options, seed=seed) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/samplers.py", line 249, in sampling_function cond, uncond = calc_cond_uncond_batch(model_function, cond, uncond, x, timestep, max_total_area, model_options) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/samplers.py", line 224, in calc_cond_uncond_batch output = model_options['model_function_wrapper'](model_function, {"input": input_x, "timestep": timestep_, "c": c, "cond_or_uncond": cond_or_uncond}).chunk(batch_chunks) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/custom_nodes/SeargeSDXL/modules/custom_sdxl_ksampler.py", line 64, in unet_function return func(input_x, timestep, c) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/model_base.py", line 68, in apply_model model_output = self.diffusion_model(xc, t, context=context, control=control, transformer_options=transformer_options, **extra_conds).float() ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1519, in wrappedcall_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1528, in callimpl return forward_call(*args, kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/custom_nodes/SeargeSDXL/modules/custom_sdxl_ksampler.py", line 70, in new_unet_forward x0 = old_unet_forward(self, x, timesteps, context, y, control, transformer_options, kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/ldm/modules/diffusionmodules/openaimodel.py", line 625, in forward h = forward_timestep_embed(module, h, emb, context, transformer_options) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/ldm/modules/diffusionmodules/openaimodel.py", line 56, in forward_timestep_embed x = layer(x, context, transformer_options) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1519, in wrappedcall_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1528, in callimpl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/ldm/modules/attention.py", line 560, in forward x = block(x, context=context[i], transformer_options=transformer_options) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1519, in wrappedcall_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1528, in callimpl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/ldm/modules/attention.py", line 390, in forward return checkpoint(self._forward, (x, context, transformer_options), self.parameters(), self.checkpoint) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/ldm/modules/diffusionmodules/util.py", line 123, in checkpoint return func(*inputs) ^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/ldm/modules/attention.py", line 455, in forward n = self.attn1(n, context=contextattn1, value=value_attn1) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1519, in wrappedcall_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1528, in callimpl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/ldm/modules/attention.py", line 366, in forward out = optimized_attention(q, k, v, self.heads) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/ldm/modules/attention.py", line 177, in attention_sub_quad hidden_states = efficient_dot_product_attention( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/ldm/modules/sub_quadratic_attention.py", line 236, in efficient_dot_product_attention return compute_query_chunk_attn( ^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/artem/ComfyUI/comfy/ldm/modules/sub_quadratic_attention.py", line 160, in getattention_scores_no_kv_chunking attn_probs = attn_scores.softmax(dim=-1) ^^^^^^^^^^^^^^^^^^^^^^^^^^^

This was the best all-round winner ComfyUI workflow so for until the last Comfy update.

Unfortunately, the workflow is broken now.It gives me the below error:

AttributeError: module 'comfy.sample' has no attribute 'convert_cond'

🤔 Do you know is there a way we can overcome this error?

Getting the following error:

Registered sys.path: ['/content/drive/MyDrive/ComfyUI/custom_nodes/comfyui_controlnet_aux-main/src/controlnet_aux', '/content/drive/MyDrive/ComfyUI/custom_nodes/comfyui_controlnet_aux-main/src/custom_midas_repo', '/content/drive/MyDrive/ComfyUI/custom_nodes/comfyui_controlnet_aux-main/src/custom_detectron2', '/content/drive/MyDrive/ComfyUI/custom_nodes/comfyui_controlnet_aux-main/src/custom_oneformer', '/content/drive/MyDrive/ComfyUI/custom_nodes/comfyui_controlnet_aux-main/src/custom_mmpkg', '/content/drive/MyDrive/ComfyUI/custom_nodes/comfyui_controlnet_aux-main/src/custom_pycocotools', '/content/drive/MyDrive/ComfyUI/custom_nodes/comfyui_controlnet_aux-main/src/__init__.py', '/content/drive/MyDrive/ComfyUI/custom_nodes/comfyui_controlnet_aux-main/src', '/content/drive/MyDrive/ComfyUI/custom_nodes/ComfyUI_Comfyroll_CustomNodes/comfy', '/content/drive/MyDrive/ComfyUI/custom_nodes/ComfyUI_Comfyroll_CustomNodes/comfy', '/content/drive/MyDrive/ComfyUI/custom_nodes/was-node-suite-comfyui-main/comfy', '/content/drive/MyDrive/ComfyUI/comfy', '/content/drive/MyDrive/ComfyUI', '/env/python', '/usr/lib/python310.zip', '/usr/lib/python3.10', '/usr/lib/python3.10/lib-dynload', '/usr/local/lib/python3.10/dist-packages', '/usr/lib/python3/dist-packages', '/tmp/tmpr9fuw9wl', '../ComfyUI'] /content/drive/MyDrive/ComfyUI/custom_nodes/comfyui_controlnet_aux-main/node_wrappers/dwpose.py:26: UserWarning: DWPose: Onnxruntime not found or doesn't come with acceleration providers, switch to OpenCV with CPU device. DWPose might run very slowly warnings.warn("DWPose: Onnxruntime not found or doesn't come with acceleration providers, switch to OpenCV with CPU device. DWPose might run very slowly") Traceback (most recent call last): File "/content/drive/MyDrive/ComfyUI/nodes.py", line 1735, in load_custom_node module_spec.loader.exec_module(module) File "<frozen importlib._bootstrap_external>", line 883, in exec_module File "<frozen importlib._bootstrap>", line 241, in callwith_frames_removed File "/content/drive/MyDrive/ComfyUI/custom_nodes/SeargeSDXL/__init__.py", line 33, in <module> from .modules._legacy import LEGACY_CLASS_MAPPINGS, LEGACY_DISPLAY_NAME_MAPPINGS File "/content/drive/MyDrive/ComfyUI/custom_nodes/SeargeSDXL/modules/__init__.py", line 29, in <module> from .ui import Defs File "/content/drive/MyDrive/ComfyUI/custom_nodes/SeargeSDXL/modules/ui.py", line 37, in <module> from .custom_sdxl_ksampler import CfgMethods File "/content/drive/MyDrive/ComfyUI/custom_nodes/SeargeSDXL/modules/custom_sdxl_ksampler.py", line 38, in <module> from comfy.model_management import get_torch_device, batch_area_memory, load_models_gpu ImportError: cannot import name 'batch_area_memory' from 'comfy.model_management' (/content/drive/MyDrive/ComfyUI/comfy/model_management.py) Cannot import /content/drive/MyDrive/ComfyUI/custom_nodes/SeargeSDXL module for custom nodes: cannot import name 'batch_area_memory' from 'comfy.model_management' (/content/drive/MyDrive/ComfyUI/comfy/model_management.py)

A bit confused on how to use the Controlnet nodes and what sort of images and effects it produces. (for example, for posing the character) Could you please expand on it in the documentation?

Hi and thanks for sharing and keeping this updated. Seems overwhelming, but it is a very flexible tool. And works fine!
Now the question: I try to put all models in an external folder, with no windows symlink. I failed to know how to do that with annotators folder. I guess is something with folder paths in custom_nodes\SeargeSDXL\init.py, but don't know how to. Don't know python.
Anybody did this? Please tell me how.

Thank you more for the work you have done! I only use your add-on v.4.3.2.

Everything works well, except for one thing. When enlarging a photo and setting in the block "High Resolution" - "hires_mode:normal" - hires_scale: 2x" an error occurs. With the same settings but "hires_scale: 1.5x" everything works fine. Is it possible to fix this?

I attach the log:

Can someone send link to the text to workflow png? For me example from github doesn't work

dont see support for openpose controlnet. Unless i can use custom.. any chance you can suggest how to add it to the latest?

it would be amazing to be allowed to change the workflow, but it is somehow locked. I'd like to add openpose CN, FaceDetailer, and IP Adapter Plus and IPA mask. But cant move or edit anything. What am I missing?

The workflow is amazing, Any chance of latest updates and new integrations?

Good day to all of you. If I may ask, where/how to configure ClipSkip on this one?

Have a blessed day.

Hi;

Might be a stupid question, but do I need to download specific "detail processor"? My only option is "none".

I would like to have a simple reactor face upscaler (not to upscale the whole image just the details that has been replaced in the face.)

I know you are quite busy but I want to make this for a Christmas card of someone that passed away is it possible in any way to pay you to make a very very simple bare bones faceswap upscaler? I could use your software or Topaz to upscale the whole image another time. I do have a few days before Christmas so I hope it can be done somehow.

t3media @ gmail

I love this UI for generation, I also like to hook all the control net boxes up to the image to image box and do frame upscaling for video frames

Somebody please make a video tutorial for V4.3, I'm too dumb to understand

Quite simply, exactly the graph and nodes I was looking for to remaster / polish my own paintings. Spent several days trying to find a workflow for that purpose until I stumbled accross Searge the goat! You sir are the man, ily!

I can't seem to get the upscaler to work on any of the .jsons. Seems to stop after the hi-res detailer. I've got 4xUltraSharp and Siax 200k loaded up, boxes are set to <True> but nothing. No errors or anything.

How use custom control net model?
It's sad that the author doesn't reply anywhere, and the wiki on github is not updated with new information

seems like this workflow is out of date now, the manager is saying that there are a lot of conflicts and node name changes

is there any way to look at a progress bar on the active nodes like other workflows?

thank you! this is the best workflow ever.

Can u add instantID

it's goddd thanks brother just i can't

use depth and other why?

Well, it was the greatest workflow I ever needed, the results were always more than satisfying, but its not working anymore, it broke my Comfyui, and I cannot find a better workflow, Please update it!

Also if anyone know how a good workflow or as good as this one, please let me know.

Edit:

He seems to have updated the workflow so it seems to works fine.

How can I use the sd1.5 model as a refiner for the basic sdxl model? Like Fooocus

It has so many extras - knobs, dials, bells, whistles - it's hard to know what each extra bit can do? But it does produce some crisp and clean images! I added some LoRAs, PAG Advanced, and Face Detailer - and it knocks each image "out-of-the-park!"

Is this dead? I havent seen updates in a long while.

can i use it in photoshop-plus

failed to fetch. Any idea how to fix?

I just installed this and comfy broke.

Thank you so much for all your hard work, i want to add openpose but idk how? if anybody can help that would be appreciated

how do you make it continue after generating the first preview? It just stops and I find no trigger to keep it going to upscale etc... I have all required, recommended and optional files downloaded to the right folders

Does anyone know how i can add more slots for LORAs to this? The Lora node that is used here only has limited slots and using another one after the first node doesn't seem to work out all that well.

I think inpaintinig is having issues, saying movedim is not available. Any thoughts?

Does anyone know if this is compatible with pony? Sorry if this is a silly question.

Doesn't work with latest Comfyui versions , nodes no longer work