Is SUPER FLUX the Secret to Insane Details?
Added 2024-10-19 21:14:23 +0000 UTCDownload my Workflow below
Comments
In Google search, Search for: comfyui (name of file) Usually you can find it on Github, or Huggingface, and download it manually. If you need help, Utilize the free ChatGPT. Just search for it in your browser.
Skies The Limit
2025-01-20 23:25:38 +0000 UTCThanks, that helped me. I found the clip in ComfyUI-Manager->Model Manager. But the upscaler wasn't there. It seems to be important to take the File with pth-ending. It's a bit annoying that there is so little help. I installed ComfyUI from a docker image with some preinstalled custom nodes. It's extremely confusing as beginner to understand why every node-package brings e.g. it's own Seed-node when it's quiet the same like in other packages.
ChrisKM
2025-01-20 22:39:33 +0000 UTCUse ChatGPT to help you.
Skies The Limit
2024-12-29 22:19:38 +0000 UTCI am a bit concerned when a person accepts a monthly fee for access to his page but does not answer any questions. If I am incorrect, please forgive me. To get the first part running, I had to download: https://huggingface.co/openai/clip-vit-large-patch14/blob/main/model.safetensors and renamed the file to: clip-vit-large-patch14.safetensors Copied it to the folder: clip The second thing I needed: https://huggingface.co/Phips/4xRealWebPhoto_v4_dat2/tree/main Copied that to the folder: upscale_models And it worked Perfect! And I do mean it's excellent!
Skies The Limit
2024-12-29 21:50:53 +0000 UTCin the jason file
xulu bai
2024-10-26 03:44:06 +0000 UTCUpscaler marked as not safe
Ryan Martin
2024-10-24 07:36:17 +0000 UTCCan you explain the upscale+image resize node settings? As the height/width does not seem to match the input nor output. I imagine the upscale model task doubles the initial width/height, so unsure what the resize node does as the width/height are both lower as is the rescale_factor of 0.5.
Mikael Svenson
2024-10-22 05:58:54 +0000 UTCGot really beautiful results with it, really helpful to be able to see a preview before continuing to gen beyond 10 steps, saves a lot of time!
Marco Gotthardt
2024-10-21 21:49:31 +0000 UTCWhat is the image resize node being used? Not able to figure out where this custom node should be coming from.
Tobrun Van Nuland
2024-10-21 18:26:19 +0000 UTCI got all to work, however Flux Loras seem to have no effect on the workflow at all as i try to load my character loras. Do i need to change anything for that to happen ?
suljia sahan
2024-10-21 16:54:41 +0000 UTCIt worked! Thanks, Andrew!
Dima Mitlenko
2024-10-21 14:21:14 +0000 UTCI love this workflow, but is anyone else having issues making the LORAS effect the image in any way? When I use other workflows they seem alot more effective, I cannot see they make any change in this workflow...
Dreambe86
2024-10-21 11:08:35 +0000 UTCThanks, I couldn't find it, but I should have known it was renamed.
Gyuri Ayeon
2024-10-21 03:25:28 +0000 UTCcould you share full comfyui file? all the xtensions loaded etc?
MightyHorus
2024-10-21 02:03:04 +0000 UTCSame question
Artemio Gonzalez Moreno
2024-10-21 01:04:15 +0000 UTCIt has something to do with the GGUF model. When I swapped it out for the Load diffusion Model node and the flux1-dev.safetensor it worked.
Andrew Pattison
2024-10-21 00:48:43 +0000 UTCSame here. Have you found a solution?
Andrew Pattison
2024-10-20 22:57:54 +0000 UTCGot en TypeError: cast_to() got an unexpected keyword argument 'copy' on KSampler
Dima Mitlenko
2024-10-20 20:44:35 +0000 UTCThxs, just got it but getting errors, do u have links to lorah, and all models used in workflow
Alex Shaw
2024-10-20 15:43:15 +0000 UTChttps://huggingface.co/openai/clip-vit-large-patch14/tree/main It's named model.safetensors, so rename it to clip-vit-large-patch14.safetensors once downloaded.
Nick Pateman
2024-10-20 13:06:22 +0000 UTCok, got it working after this the was-node-suite-comfyui was throwing errors on startup, --- WAS Node Suite: Created default conf file at G:\ComfyUI\custom_nodes\was-node-suite-comfyui\was_suite_config.json. Traceback (most recent call last): File "G:\ComfyUI\nodes.py", line 2001, in load_custom_node module_spec.loader.exec_module(module) File "", line 883, in exec_module File "", line 241, in _call_with_frames_removed File "G:\ComfyUI\custom_nodes\was-node-suite-comfyui\__init__.py", line 1, in from .WAS_Node_Suite import NODE_CLASS_MAPPINGS File "G:\ComfyUI\custom_nodes\was-node-suite-comfyui\WAS_Node_Suite.py", line 14511, in build_info = ' '.join(cv2.getBuildInformation().split()) AttributeError: module 'cv2' has no attribute 'getBuildInformation' Cannot import G:\ComfyUI\custom_nodes\was-node-suite-comfyui module for custom nodes: module 'cv2' has no attribute 'getBuildInformation' --- the solution for me was to reinstall opencv pip uninstall opencv-python pip install opencv-python I then had to downgrade numpy using the following command, pip install numpy==1.26.3 workflow seems to be working now, thanks! This is going to be awesome.
Nick Pateman
2024-10-20 12:50:54 +0000 UTCThis might be a dumb question, but you talk about adding a lora in the video, and there's a node, but it is sorta hanging out there and not connected to anything. I see a TON of noodles coming out of the unet loader and going to lots of different model inputs... do I need to shoot a noodle out of the lora "model" output to each of these, and also the clip ones? I'm just trying to figure out what I need to attach and where in order to get the LoRa working.
tdog
2024-10-20 12:42:53 +0000 UTCJust trying to fix it by removing all extensions now throwing an error and starting again. Seemed to have something to do with the resize node.
Nick Pateman
2024-10-20 12:40:41 +0000 UTCUsing this workflow, and installing the missing extensions using the manager has broken my comfyui which now throws a whole bunch of errors on startup.
Nick Pateman
2024-10-20 12:33:04 +0000 UTCI think its the clip-vit-large-patch14. Where can I get the copy you are using?
Pieter van Wyk
2024-10-20 08:16:18 +0000 UTCIm getting just black images. Help?
Pieter van Wyk
2024-10-20 08:07:24 +0000 UTCHi I get "Warning: Ran out of memory when regular VAE encoding, retrying with tiled VAE encoding." and seem to be stuck there - this is on the 3rd ksampler starting at 20 and ending at 30. Im using a 3090. Edit: it appears to progress but its taking 30 minutes for 30 steps
John Tam
2024-10-20 00:02:07 +0000 UTC
