How i use GGUF Models?

Missing Nodes: -Anything Everywhere -ImpactGaussianBlurMask -LoadAndResizeImage And can't find them in Manager.

Hello. I have some workflows to be used in ComfyUI. When I select the workflows, I can't get them to work properly due to some missing models, LoRAs, etc. For someone who knows how to do this, it's very simple, but I haven't been able to get it right. I'm looking for someone who can download the models, LoRAs, and other elements used in the workflow, make sure everything works smoothly, and then send the workflow back to me. The models, LoRAs, etc., used in the workflow have already been provided by the workflow owner, but I just couldn't manage to do it myself. As I said, what I need is for the workflow to work correctly when I open it in ComfyUI. I will share examples with those who submit offers.

The llama folder did not appear

我看到好多.. Live Portrait 的node , 都不知道要用那個好了.

amazing, it works. Worth pointing out you need ffmpeg and a environment path set in path to where it lies,. then you can make videos rather than anim gifs. But is there a way of the faces not getting all messed up?

I have no idea what this is or how to use comfyui, but im gonna learn it if this is what it can do. I dot many low quality pics that I need to upscale (upscale that can add details) I even have Topaz, but i dont like the results.

if i dont want any camera movent how i can archieve ?

Hi @CG TOP TIPS, when I start the workflow I get this error: ExpressionEditor DLL load failed while importing aggregations: The specified module was not found. # ComfyUI Error Report ## Error Details - **Node Type:** ExpressionEditor - **Exception Type:** ImportError - **Exception Message:** DLL load failed while importing aggregations: The specified module was not found. Do you know how to fix the error? everything is installed correctly, even the "Expression Editor (PHM)" module is visible in the stream, but when I start it the error comes out. :( Thanks!

Good video friend, I have 2 questions since I am an automatic 1111 user and I wanted to ask you if you can use the ControlNet reference model here?? and 2 would be that it is used for face repair in Comfyui ? since in automatic I use adetailer and I don't know what is used here

How do i use my own image and change the pose? Or i have to have more own trained lora or something?

why you just put the links at info box, i am new on it and trying to digg by myself

This doesn't look easy lol

OMG, 20 minutes for 20 steps at 1024.... and I was sad that i render 1024 in 8 seconds...

hi can you create a great workflow ( using minimum nodes) for the best adobe firefly alternative ? bascially editing image with prompt and mask using FLUX. ? that would be a great video idea too

Always choose .safetensors over .bin, .safetensors are both more speed-efficient and free from security risks compared to .bin files!

Can this be used for adding detail to an image after I have removed an object? I can see a use case there, where I can remove objects and then fill in with SDXL inpainting, but the results are far from good.

cant work..- **Node Type:** ControlNetLoader - **Exception Type:** _pickle.UnpicklingError - **Exception Message:** invalid load key, '\xbc'.

action is not allowed with this security level of configuration

LMAO... time to open fashion shop

My output video only do 3sec. My original video is 10sec. How do I get it to do the whole 10sec?

when i click, the alpha texture is imprinted into the mesh along with the black box surrounding it. How can i prevent this and only have the texture imprinted like you have done?

How aply pose?

4:40 you dont show us how you got out of that drawing screen. its cut off in your video

thx for movie more this ERROR: occurred when executing CheckpointLoaderSimple: ERROR: Could not detect model type of: C:\ComfyUI\ComfyUI_windows_portable\ComfyUI\models\checkpoints\flux1-fp8.safetensors File "C:\ComfyUI\ComfyUI_windows_portable\ComfyUI\execution.py", line 317, in execute output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\ComfyUI\ComfyUI_windows_portable\ComfyUI\execution.py", line 192, in get_output_data return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\ComfyUI\ComfyUI_windows_portable\ComfyUI\execution.py", line 169, in _map_node_over_list process_inputs(input_dict, i) File "C:\ComfyUI\ComfyUI_windows_portable\ComfyUI\execution.py", line 158, in process_inputs results.append(getattr(obj, func)(**inputs)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\ComfyUI\ComfyUI_windows_portable\ComfyUI odes.py", line 539, in load_checkpoint out = comfy.sd.load_checkpoint_guess_config(ckpt_path, output_vae=True, output_clip=True, embedding_directory=folder_paths.get_folder_paths("embeddings")) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\ComfyUI\ComfyUI_windows_portable\ComfyUI\comfy\sd.py", line 527, in load_checkpoint_guess_config raise RuntimeError("ERROR: Could not detect model type of: {}".format(ckpt_path))

interesting

it seems that the upscaler does not only upscale the image 😁😄😄😇

Magic!

reference @stevietee3878 advice below. The workflow had been working but now getting following error.... # ComfyUI Error Report ## Error Details - **Node Type:** ExpressionEditor - **Exception Type:** safetensors_rust.SafetensorError

Where do i get the model?

Hello! Thanks for the video, it's great! I get an error: The request output failed validation. CLILoader: - Value is missing from the list: clip_name: 't5\google_t5-v1_1-xxl_encoderonly-fp8_e4m3fn.safetensors' is missing from [] How to fix this? Thank you!

Why the extra VAE Encode -> VAE Decode step at the end?

can you make workflow using gguf models?

How can my ComfyUI have "Anything everywhere" node ?.

You lost me from the start, my ComfyUI doesn't have a "manager" button haha

Hi there! The model files in Cog folder takes 40GB !!! I miss any mention of that in advance :( ... EDIT : UGH....but 19.6GB are in the '.git' folder .... wtf??? deleted that...sry for spamm

I'm curious, why is it better than 16 or 20 steps directly, instead you do 8+8 steps ? Anyway if that "Unsample" can help remove that annoying grid noise we get after upscaling Flux, I'm all for it

Practically unusable. The results from this (if you can work through the bugs and get it functioning) are worse than img2vid from a year ago. It does follow the pose well, but the generation is blotchy and inconsistent. I used the exact pics/vid and settings in this tutorial and it looks nothing like his results. Unsubbed and thumbs DOWN

You are one of the most prolific creators on youtube!! finding it hard to keep up LOL!! Keep up the good work

Is the "latent upscale by factor (WAS)" suppose to increase image dimensions? I've tried values of 1.5 and 2 expecting the image dimensions to increase by half and by double yet the image dimensions are always the same regardless.

Nice work, anything I can find to reduce the plastic face look from Flux.

thanks!😊

Hi, bro! Need your advice! Is it possible to do the following using ControlNet and Flux? The girl in the generated photos should always have the same face from the original photo 1. The object in her hands will also be created from the original photo 2, for example a bottle of water. The generated photo shows a girl with a bottle of water. In another prompt, the generation again uses the original photo 1 with a girl. Source photo 2 - Coca-Cola bottle. On The generated photo is a girl with a bottle of Coca-cola in her hand. it is necessary to change different situations in the background and different poses of the girl in the generated photos, and at the same time the girl must have the same face.

Nawet wirtualna rzeczywistość nie będzie już rzeczywista :/

Did you try new methods to restore old pictures or colourise black and white images with FLUX?

this could work in stableDifusion FORGE?

Thank you for sharing, but this workflow will alter the subject's appearance. Is it possible to add a mask to exclude the face or head area?

LoadTheMistoFluxControlNet Allocation on device 0 would exceed allowed memory. (out of memory) Currently allocated : 7.26 GiB Requested : 144.00 MiB Device limit : 8.00 GiB Free (according to CUDA): 0 bytes PyTorch limit (set by user-supplied memory fraction) : 17179869184.00 GiB

So I installed the nodes, but using your workflow doesn't give me consistent results. The result is totally different than the controller shape. How to fix that?

That looks amazing, shout out to you! But can you explain to me what's your thought process behind it, how do you know how to combine all the components? Do you recommend any materials that helps in understanding things that happen here?