1-Click INSTALL ULTIMATE FLUX & ComfyUI WebUI!

I have git and have the ComfyUI-Manager folder exists int the directory, I did all above steps anyway, no manager button in comfy ui

Long YuHe

Translation please...what does this mean? SamplerCustomAdvanced expected scalar type Half but found BFloat16

James Hanggi

yes

Aitrepreneur

Hi! is still available for runpod?

Ricardo Campos

it works fine you just need to have a runpod account already or create one. Runpod.io is a website it's not "for macOS" or any specific OS

Aitrepreneur

the comfyUI template wont load on run pod when I click the link. Is this for macOS or no?

Roy Henry

the installers are not for Mac, it's only for Windows. The sh file is for runpod running on my comfyui template

Aitrepreneur

the .sh file just says process complete and nothing else

Roy Henry

I'm on Mac it wants me to open on whiskey the .bat auto installer? I thought I was supposed to open in terminal

Roy Henry

it's an installer made for the people who already have comfyui previously installed and don't want to reinstall it. It automatically downloads all the missing models

Aitrepreneur

you mean git for windows? I added it to the latest installer just in case but usually it's better if people have it installed correctly already though

Aitrepreneur

What is MODELS-ULTIMATE_FLUX_INSTALL.bat for?

Aaron1611

Thank you, it worked! You should had the Github install and this to your tutorial because it doesn't work on a fresh windows install. I'm pretty sure most newbies doesn't have github installed and don't know they need to update all of this. Anyway, thank again for your swift replies and explanation! ;)

Einziger

Always the same issue whenever you load a brand new workflow 1. Update comfyui 2. Click on "Install Missing Custom Nodes" and click on the first small checkbox to select all the nodes and then click install. 3. It will then ask you to restart, so click on that restart button, wait for the webui to reboot 4. Click the refresh button (F5) 5. If the message is still there but all nodes are installed, just clear the workflow and reload the json file inside So it should go like this: Load the Ultimate Workflow json file ---> Update Comfyui --> Install Missing custom nodes ---> Restart webui ---> Refresh with F5 ---> IF the message is still there and there are no more missing custom nodes to install ---> Load some random workflow json file----> Then load the Ultimate Workflow json file again and the message should disapear also make sure that your comfyui is correctly updated you can run the update_comfyui_and_python_dependencies.bat file present in the ComfyUI_windows_portable\update folder as well

Aitrepreneur

I followed your steps but when comfy ui restarts many nodes are red and I have this message: When loading the graph, the following node types were not found: FluxGuidance (In group node 'workflow/positive prompt guidance') workflow/positive prompt guidanceRemove from workflow workflow/Negative PromptRemove from workflow LoadFluxControlNet (In group node 'workflow/Flux Controlnet') workflow/Flux ControlnetRemove from workflow ApplyFluxControlNet (In group node 'workflow/Flux Controlnet') XlabsSampler (In group node 'workflow/XLabs Sampler') workflow/XLabs SamplerRemove from workflow ModelSamplingFlux (In group node 'workflow/Other Nodes') workflow/Other NodesRemove from workflow DualCLIPLoaderGGUF UnetLoaderGGUF Nodes that have failed to load will show as red on the graph.

Einziger

probably because you don't have git for windows installed already so it couldn't have clone the manager github, it's an easy fix, just dl git for windows and install it: https://github.com/git-for-windows/git/releases/download/v2.47.0.windows.1/Git-2.47.0-64-bit.exe then go to the ComfyUI_windows_portable\ComfyUI\custom_nodes folder click on the folder path then type cmd press enter and in the command prompt window that just appeared you're gonna type: git clone https://github.com/ltdrdata/ComfyUI-Manager.git then press enter now if you reopen comfyui the manager will be here

Aitrepreneur

Uncensored doesn't mean that it can generate hardcore nsfw images, just that nudes are possible

Aitrepreneur

Bro, I followed all the steps. I launched the Ultimate Json, I had a lot of errors then I searched for the "manager button", there is no "manager button"... Did I missed something ?

Einziger

Video says this is uncensored, its not, is there something Im missing or is the title in the video wrong?

Dani

probably because you either chose option 1 or 2 during the install so either the Q4 or Q5 model were downloaded instead and since the workflow by default selects the Q8 all you have to do is just click on the model loader name and chose the apropriate model

Aitrepreneur

Prompt outputs failed validation DualCLIPLoaderGGUF: - Value not in list: clip_name2: 't5-v1_1-xxl-encoder-Q8_0.gguf' not in ['ViT-L-14-TEXT-detail-improved-hiT-GmP-TE-only-HF.safetensors', 't5-v1_1-xxl-encoder-Q3_K_L.gguf', 't5xxl_fp16.safetensors']

Psi

thank you. Those files were not in the right folder so I put the bat file in the comfyui folder and ran it. odd thing was that it placed the gguf files in the comfyui_windows_portable folder. Not sure why so I moved them to the unet folder and restarted and now I can choose them in comfyui

Chris Christopher

check the models/unet folder and see if you see a .gguf model in there, if it is just select the model on the right side. If not then that means that there was an issue during the installation for some reason, you can use the MODELS-ULTIMATE_FLUX_INSTALL.bat and put that inside the ComfyUI_windows_portable\ComfyUI folder before running it, it will download all the missing files

Aitrepreneur

By need I meant that I found that I could get the models appearing in the f8 side. On the gguf when I click a model it disappears. It looks like only the f8 models downloaded for me so I can't use the gguf side

Chris Christopher

what do you mean need? You don't need to use the F8, gguf are the recommended models and they are already downloaded during install automatically

Aitrepreneur

I just found that I need to use the F8 side of the workflow. Where do I get the GGUF models?

Chris Christopher

flux models are supposed to be in the models/unet folder

Aitrepreneur

I just used the one click install and it loaded, it seems, fine. The flux models were there to choose from. When I loaded the workflow and updated all the nodes, now if I click on the node where the flux model should be it shows no models loaded. I've even checked the comfyUI folder and can't find the flux models anywhere. Where did they go?

Chris Christopher

Got it. My bad, I was just following along on your youtube video and it was not mentioned in there. Thanks for your help.

DrummerMan419

read the post the link is written in the beginning

Aitrepreneur

Where is the ultimate_flux json file that we are supposed to load in step 4? I can't find it

DrummerMan419

The ComfyUI-GGUF node seems to be causing some errors. I went through the install instructs several times, after importing the Json and updating ComfyUI and installing the missing nodes, I get this when it relaunches. Each time I restart ComfyUI it shows 'ComfyUI-GGUF' still needs to be updated, but it doesn't seem to update after selecting it to update and then restarting/refreshing browser. FETCH DATA from: G:\aiArtGeneration\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\extension-node-map.json [DONE] [] Update ComfyUI Cmd('git') failed due to: exit code(1) cmdline: git pull -v -- origin stderr: 'error: The following untracked working tree files would be overwritten by merge:' CUSTOM NODE PULL: Fail Install: pip packages Line is comment...skipping FETCH DATA from: G:\aiArtGeneration\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\extension-node-map.json [DONE] [] FETCH DATA from: G:\aiArtGeneration\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\.cache\1742899825_extension-node-map.json [DONE] FETCH DATA from: G:\aiArtGeneration\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\.cache\1514988643_custom-node-list.json [DONE] FETCH DATA from: G:\aiArtGeneration\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\.cache\746607195_github-stats.json [DONE] FETCH DATA from: G:\aiArtGeneration\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\.cache\1742899825_extension-node-map.json [DONE] FETCH DATA from: G:\aiArtGeneration\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\.cache\1514988643_custom-node-list.json [DONE] FETCH DATA from: G:\aiArtGeneration\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\.cache\746607195_github-stats.json [DONE] FETCH DATA from: G:\aiArtGeneration\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\.cache\1742899825_extension-node-map.json [DONE] Install custom node 'ComfyUI-GGUF' Try fixing: ['https://github.com/city96/ComfyUI-GGUF'] Install: pip packages Attempt to fixing '['https://github.com/city96/ComfyUI-GGUF']' is done. After restarting ComfyUI, please refresh the browser. FETCH DATA from: G:\aiArtGeneration\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\extension-node-map.json [DONE]

Gravnos

Also thank you for all the work you have put in to the latest 1-click. It is working very well, and with some limitations, I am now creating very decent homebrew UE5-like renders in not a lot of time.

RT Solvalou

I was watching a few videos on youtube, and it seems that there are free paths to creating your own LORAs if you have 12gb VRAM or more using Fluxgym? Is this true? Is there a way to run fluxgym through ComfyUI or integrate the LORA into the dev model early in the rendering process? Sorry for so many questions, but I am definitely excited to begin creating my own LORAs.

RT Solvalou

Health is most important! Take the stress out and get well soon! All the best.

Violent Pixelation

for which purpose? what error do you have?

Aitrepreneur

what do you mean static?

Aitrepreneur

This portable version 1ClickAuto installer will not allow installation of Insightface that I can find. Have tried both through manager and gitbash manually. Constant conflicts with building wheel.

Shaun Newton

Install all went smoothly, thank you. Its only generating static for me though.

Adam

and this when I try to load the queue: `newbyteorder` was removed from the ndarray class in NumPy 2.0. Use `arr.view(arr.dtype.newbyteorder(order))` instead.

Michael O'Neill

Okay, I figured it out. Erased the folder and reinstalled, things are looking better now, but I'm still getting two errors: Text Find and Replace (In group node 'workflow>Florence') workflow>Florence

Michael O'Neill

well...that's just not possible, if you followed everything step by step, the nodes wouldn't uninstall themselves so I'm not sure where you got it wrong, best thing would be to just record a video of you doing everything step by step from the beginning once you relaunch the webui, watch also the video as well step by step

Aitrepreneur

When loading the graph, the following node types were not found: FluxGuidance (In group node 'workflow/positive prompt guidance') workflow/positive prompt guidanceRemove from workflow workflow/Negative PromptRemove from workflow ModelSamplingFlux (In group node 'workflow/Other Nodes') workflow/Other NodesRemove from workflow workflow/Positive PromptRemove from workflow workflowsampler and seed upscaler workflowExtra nodes upscaler workflowFlorence workflowFlux Controlnet workflowImage gen sampler workflowOther Nodes workflowParam workflowsampler img workflowOther Nodes 2 workflowNegative Prompt workflowSampler 3 workflowNodes 3 DualCLIPLoaderGGUF UnetLoaderGGUF workflowXLabs Sampler workflowpositive prompt guidance Nodes that have failed to load will show as red on the graph.

Michael O'Neill

seems to be getting worse...

Michael O'Neill

if comfyui was updated and every missing nodes were installed correctly and there are no missing nodes in the missing nodes tab, then sometimes you just need to clear the workflow, load another one and RE-load the ultimate all in one workflow file for the ui to get updated correctly

Aitrepreneur

That helped, but I'm still getting this error... Warning: Missing Node Types When loading the graph, the following node types were not found: Text Find and Replace (In group node 'workflow>Florence') workflow>Florence workflowFlorence workflowNodes 3 No selected item Nodes that have failed to load will show as red on the graph.

Michael O'Neill

Read the patreon post: 4. The webui will launch, load the ULTIMATE_FLUX_ALL-IN-ONE.json file, you will see a lot of errors, then click on the manager button and click "Update comfyui". Then click on "Install Missing Custom Nodes" and click on the first small checkbox to select all the nodes and then click install. It will then ask you to restart, so click on that restart button, wait for the webui to reboot, click refresh once it's done and there you go😎

Aitrepreneur

yeah, though so. Have you installed the zluda comfyui version? If so there might be some nodes that are not compatible with AMD but I can't be sure of that

Aitrepreneur

I'm getting some error messages, can you help? When loading the graph, the following node types were not found: Text Find and Replace (In group node 'workflow/Florence') workflow/FlorenceRemove from workflow FluxGuidance (In group node 'workflow/positive prompt guidance') workflow/positive prompt guidanceRemove from workflow workflow/Negative PromptRemove from workflow ModelSamplingFlux (In group node 'workflow/Other Nodes') workflow/Other NodesRemove from workflow DualCLIPLoaderGGUF UnetLoaderGGUF Nodes that have failed to load will show as red on the graph.

Michael O'Neill

AMD

Otniel Ocampo

what's your GPU?

Aitrepreneur

The Upscaler Gave me this message but I have no idea what it mean!! got prompt Requested to load FluxClipModel_ Loading 1 new model loaded completely 0.0 5062.70263671875 True using sdpa for attention No flash_attn import to remove Florence2LanguageForConditionalGeneration has generative capabilities, as `prepare_inputs_for_generation` is explicitly overwritten. However, it doesn't directly inherit from `GenerationMixin`. From �vv4.50👈 onwards, `PreTrainedModel` will NOT inherit from `GenerationMixin`, and this model will lose the ability to call `generate` and other related functions. - If you're using `trust_remote_code=True`, you can get rid of this warning by loading the model with an auto class. See https://huggingface.co/docs/transformers/en/model_doc/auto#auto-classes - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception). - If you are not the owner of the model architecture class, please contact the model code owner to update it. Compiling in progress. Please wait... Exception Code: 0xC0000094 0x00007FFF044C639F, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_cuda.dll(0x00007FFF03020000) + 0x14A639F byte(s), ?split_with_sizes_copy_out_cuda@native@at@@YAXAEBVTensor@2@V?$ArrayRef@_J@c10@@_JV?$ArrayRef@VTensor@at@@@5@@Z() + 0x5FDDF byte(s) 0x00007FFF0448A965, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_cuda.dll(0x00007FFF03020000) + 0x146A965 byte(s), ?split_with_sizes_copy_out_cuda@native@at@@YAXAEBVTensor@2@V?$ArrayRef@_J@c10@@_JV?$ArrayRef@VTensor@at@@@5@@Z() + 0x243A5 byte(s) 0x00007FFF044B39C7, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_cuda.dll(0x00007FFF03020000) + 0x14939C7 byte(s), ?split_with_sizes_copy_out_cuda@native@at@@YAXAEBVTensor@2@V?$ArrayRef@_J@c10@@_JV?$ArrayRef@VTensor@at@@@5@@Z() + 0x4D407 byte(s) 0x00007FFF044BFBDB, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_cuda.dll(0x00007FFF03020000) + 0x149FBDB byte(s), ?split_with_sizes_copy_out_cuda@native@at@@YAXAEBVTensor@2@V?$ArrayRef@_J@c10@@_JV?$ArrayRef@VTensor@at@@@5@@Z() + 0x5961B byte(s) 0x00007FFF044C64E1, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_cuda.dll(0x00007FFF03020000) + 0x14A64E1 byte(s), ?split_with_sizes_copy_out_cuda@native@at@@YAXAEBVTensor@2@V?$ArrayRef@_J@c10@@_JV?$ArrayRef@VTensor@at@@@5@@Z() + 0x5FF21 byte(s) 0x00007FFF04DF1379, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_cuda.dll(0x00007FFF03020000) + 0x1DD1379 byte(s), ?impl@structured_topk_out_cuda@native@at@@QEAAXAEBVTensor@3@_J1_N200@Z() + 0x279 byte(s) 0x00007FFF04CB2345, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_cuda.dll(0x00007FFF03020000) + 0x1C92345 byte(s), ?where_outf@cuda@at@@YAAEAVTensor@2@AEBV32@00AEAV32@@Z() + 0x7B195 byte(s) 0x00007FFF04B770BD, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_cuda.dll(0x00007FFF03020000) + 0x1B570BD byte(s), ?bucketize_outf@cuda@at@@YAAEAVTensor@2@AEBV32@0_N1AEAV32@@Z() + 0x6358D byte(s) 0x00007FFEFBEE070A, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_cpu.dll(0x00007FFEFB2E0000) + 0xC0070A byte(s), ?call@topk@_ops@at@@SA?AV?$tuple@VTensor@at@@V12@@std@@AEBVTensor@3@VSymInt@c10@@_J_N3@Z() + 0x29A byte(s) 0x00007FFEFB3C27A7, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_cpu.dll(0x00007FFEFB2E0000) + 0xE27A7 byte(s), ?topk_symint@Tensor@at@@QEBA?AV?$tuple@VTensor@at@@V12@@std@@VSymInt@c10@@_J_N2@Z() + 0x57 byte(s) 0x00007FF866535D16, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_python.dll(0x00007FF8664E0000) + 0x55D16 byte(s), ??C?$THPPointer@U_frame@@@@QEAAPEAU_frame@@XZ() + 0x13FD6 byte(s) 0x00007FF8666B104D, C:\Utilities\ComfyUI-FLUX-Zluda\venv\lib\site-packages\torch\lib\torch_python.dll(0x00007FF8664E0000) + 0x1D104D byte(s), ?release@?$THPPointer@UPyCodeObject@@@@QEAAPEAUPyCodeObject@@XZ() + 0xADB0D byte(s) 0x00007FF8B64B9EEA, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x19EEA byte(s), PyObject_IsTrue() + 0xD2 byte(s) 0x00007FF8B6519892, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x79892 byte(s), _PyEval_EvalFrameDefault() + 0x3FF2 byte(s) 0x00007FF8B65149D7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x749D7 byte(s), _PyFunction_Vectorcall() + 0x87 byte(s) 0x00007FF8B64CA8AF, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x2A8AF byte(s), _PyObject_FastCallDictTstate() + 0x6B byte(s) 0x00007FF8B65D81F4, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x1381F4 byte(s), _PyObject_Call_Prepend() + 0x7C byte(s) 0x00007FF8B65D8150, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x138150 byte(s), _PyBytesWriter_Resize() + 0x10EC byte(s) 0x00007FF8B651A598, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x7A598 byte(s), _PyEval_EvalFrameDefault() + 0x4CF8 byte(s) 0x00007FF8B65149D7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x749D7 byte(s), _PyFunction_Vectorcall() + 0x87 byte(s) 0x00007FF8B64CA8AF, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x2A8AF byte(s), _PyObject_FastCallDictTstate() + 0x6B byte(s) 0x00007FF8B65D81F4, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x1381F4 byte(s), _PyObject_Call_Prepend() + 0x7C byte(s) 0x00007FF8B65D8150, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x138150 byte(s), _PyBytesWriter_Resize() + 0x10EC byte(s) 0x00007FF8B651A598, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x7A598 byte(s), _PyEval_EvalFrameDefault() + 0x4CF8 byte(s) 0x00007FF8B6513615, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x73615 byte(s), _PyObject_GC_Malloc() + 0x1055 byte(s) 0x00007FF8B64FC00C, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5C00C byte(s), PyVectorcall_Call() + 0xB8 byte(s) 0x00007FF8B64FBE87, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5BE87 byte(s), _PyObject_Call() + 0x143 byte(s) 0x00007FF8B651B5E7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x7B5E7 byte(s), _PyEval_EvalFrameDefault() + 0x5D47 byte(s) 0x00007FF8B65149D7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x749D7 byte(s), _PyFunction_Vectorcall() + 0x87 byte(s) 0x00007FF8B64FC00C, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5C00C byte(s), PyVectorcall_Call() + 0xB8 byte(s) 0x00007FF8B64FBE87, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5BE87 byte(s), _PyObject_Call() + 0x143 byte(s) 0x00007FF8B651B5E7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x7B5E7 byte(s), _PyEval_EvalFrameDefault() + 0x5D47 byte(s) 0x00007FF8B6513615, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x73615 byte(s), _PyObject_GC_Malloc() + 0x1055 byte(s) 0x00007FF8B64FC00C, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5C00C byte(s), PyVectorcall_Call() + 0xB8 byte(s) 0x00007FF8B64FBE87, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5BE87 byte(s), _PyObject_Call() + 0x143 byte(s) 0x00007FF8B651B5E7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x7B5E7 byte(s), _PyEval_EvalFrameDefault() + 0x5D47 byte(s) 0x00007FF8B6518620, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x78620 byte(s), _PyEval_EvalFrameDefault() + 0x2D80 byte(s) 0x00007FF8B6513615, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x73615 byte(s), _PyObject_GC_Malloc() + 0x1055 byte(s) 0x00007FF8B64FC00C, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5C00C byte(s), PyVectorcall_Call() + 0xB8 byte(s) 0x00007FF8B64FBE87, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5BE87 byte(s), _PyObject_Call() + 0x143 byte(s) 0x00007FF8B651B5E7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x7B5E7 byte(s), _PyEval_EvalFrameDefault() + 0x5D47 byte(s) 0x00007FF8B65149D7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x749D7 byte(s), _PyFunction_Vectorcall() + 0x87 byte(s) 0x00007FF8B6516033, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x76033 byte(s), _PyEval_EvalFrameDefault() + 0x793 byte(s) 0x00007FF8B65149D7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x749D7 byte(s), _PyFunction_Vectorcall() + 0x87 byte(s) 0x00007FF8B651BACD, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x7BACD byte(s), _PyEval_EvalFrameDefault() + 0x622D byte(s) 0x00007FF8B65149D7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x749D7 byte(s), _PyFunction_Vectorcall() + 0x87 byte(s) 0x00007FF8B651BACD, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x7BACD byte(s), _PyEval_EvalFrameDefault() + 0x622D byte(s) 0x00007FF8B65149D7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x749D7 byte(s), _PyFunction_Vectorcall() + 0x87 byte(s) 0x00007FF8B6516033, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x76033 byte(s), _PyEval_EvalFrameDefault() + 0x793 byte(s) 0x00007FF8B6516A94, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x76A94 byte(s), _PyEval_EvalFrameDefault() + 0x11F4 byte(s) 0x00007FF8B65149D7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x749D7 byte(s), _PyFunction_Vectorcall() + 0x87 byte(s) 0x00007FF8B64FBFB0, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5BFB0 byte(s), PyVectorcall_Call() + 0x5C byte(s) 0x00007FF8B64FBE87, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5BE87 byte(s), _PyObject_Call() + 0x143 byte(s) 0x00007FF8B651B5E7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x7B5E7 byte(s), _PyEval_EvalFrameDefault() + 0x5D47 byte(s) 0x00007FF8B6516A94, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x76A94 byte(s), _PyEval_EvalFrameDefault() + 0x11F4 byte(s) 0x00007FF8B6516A94, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x76A94 byte(s), _PyEval_EvalFrameDefault() + 0x11F4 byte(s) 0x00007FF8B65149D7, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x749D7 byte(s), _PyFunction_Vectorcall() + 0x87 byte(s) 0x00007FF8B6513769, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x73769 byte(s), _PyObject_GC_Malloc() + 0x11A9 byte(s) 0x00007FF8B64FBFB0, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5BFB0 byte(s), PyVectorcall_Call() + 0x5C byte(s) 0x00007FF8B64FBD93, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0x5BD93 byte(s), _PyObject_Call() + 0x4F byte(s) 0x00007FF8B6588962, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0xE8962 byte(s), _PyRuntimeState_Fini() + 0x5F2 byte(s) 0x00007FF8B65888DE, C:\Users\ETNWHVAC\AppData\Local\Programs\Python\Python310\python310.dll(0x00007FF8B64A0000) + 0xE88DE byte(s), _PyRuntimeState_Fini() + 0x56E byte(s) 0x00007FF945D11BB2, C:\Windows\System32\ucrtbase.dll(0x00007FF945CF0000) + 0x21BB2 byte(s), _configthreadlocale() + 0x92 byte(s) 0x00007FF9468C7374, C:\Windows\System32\KERNEL32.DLL(0x00007FF9468B0000) + 0x17374 byte(s), BaseThreadInitThunk() + 0x14 byte(s) 0x00007FF94831CC91, C:\Windows\SYSTEM32\ntdll.dll(0x00007FF9482D0000) + 0x4CC91 byte(s), RtlUserThreadStart() + 0x21 byte(s) Press any key to continue . . .

Otniel Ocampo

what's your GPU?

Aitrepreneur

I'll add it soon in the V2 workflow

Aitrepreneur

I'll be back very soon, got a few health issues these last few weeks

Aitrepreneur

Waiting impatiently for new content from K… :D

James Stewart

How or where would you add a face swap workflow like ReActor?

Lagartija

please help. i got this after restarting wht the run_nvidia_gpu.bat C:\ComfyUI_windows_portable>.\python_embeded\python.exe -s ComfyUI\main.py --windows-standalone-build [START] Security scan [DONE] Security scan ## ComfyUI-Manager: installing dependencies done. ** ComfyUI startup time: 2024-10-16 23:38:29.116276 ** Platform: Windows ** Python version: 3.11.8 (tags/v3.11.8:db85d51, Feb 6 2024, 22:03:32) [MSC v.1937 64 bit (AMD64)] ** Python executable: C:\ComfyUI_windows_portable\python_embeded\python.exe ** ComfyUI Path: C:\ComfyUI_windows_portable\ComfyUI ** Log path: C:\ComfyUI_windows_portable\comfyui.log Prestartup times for custom nodes: 0.0 seconds: C:\ComfyUI_windows_portable\ComfyUI\custom_nodes\rgthree-comfy 1.4 seconds: C:\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager Traceback (most recent call last): File "C:\ComfyUI_windows_portable\ComfyUI\main.py", line 90, in import execution File "C:\ComfyUI_windows_portable\ComfyUI\execution.py", line 13, in import nodes File "C:\ComfyUI_windows_portable\ComfyUI\nodes.py", line 21, in import comfy.diffusers_load File "C:\ComfyUI_windows_portable\ComfyUI\comfy\diffusers_load.py", line 3, in import comfy.sd File "C:\ComfyUI_windows_portable\ComfyUI\comfy\sd.py", line 5, in from comfy import model_management File "C:\ComfyUI_windows_portable\ComfyUI\comfy\model_management.py", line 143, in total_vram = get_total_memory(get_torch_device()) / (1024 * 1024) ^^^^^^^^^^^^^^^^^^ File "C:\ComfyUI_windows_portable\ComfyUI\comfy\model_management.py", line 112, in get_torch_device return torch.device(torch.cuda.current_device()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\ComfyUI_windows_portable\python_embeded\Lib\site-packages\torch\cuda\__init__.py", line 878, in current_device _lazy_init() File "C:\ComfyUI_windows_portable\python_embeded\Lib\site-packages\torch\cuda\__init__.py", line 305, in _lazy_init raise AssertionError("Torch not compiled with CUDA enabled") AssertionError: Torch not compiled with CUDA enabled C:\ComfyUI_windows_portable>pause

Simon

I noticed that the 'batch' crashes, for example, at a number like 30. Is there any way to generate more images randomly? Additionally, I was shocked at how difficult it is (at least for me) to extract the prompt from a previously generated PNG image. How can I easily extract the image description from a PNG, even outside of Comfy

Tomasz

Is it avoidable that the model gets loaded again for each generation? That wastes a lot of time.

Simon

Hoping to see your take on using Pose controlnet for character creation (similar to https://www.youtube.com/watch?v=MbQv8zoNEfY&t=131s) It's looking promising but unable to match his results. Your one click installers are a lifesaver for stuff like this since it ensures all the models are the same.

Tri3Ax

Also appears that ComfyUI_NAI-mod has been renamed/replaced by ComfyUI-Universal-Styler

Seanehty

Well - after all the troubleshooting steps in the video AND clearing and reloading the workspace. It worked

McGenie

Not sure if I reset - but I did reload it. I have started a fresh install on a different pod. And then I’ll try your method as well.

McGenie

Did you try clearing the workflow then reloading it from the json?

Seanehty

How did you figure it out? I also don't have the manager button

Sebastion Udelhofen

How did you get it to work? Did you install the checkpoint manually? For me, it gives an error instead of generating images when I click "queue"

Zen Crypto

Good evening, how do I create the same character in different places and with different poses of the same character?

Demetrius Ricarte

hi, soem nodes didnt install ComfyUI-GGUF fix failed: This action is not allowed with this security level configuration. x-flux-comfyui fix failed: This action is not allowed with this security level configuration. ah I updated everything and it solved the issue.

Mohamed Owda

I wanted to try the running this in Runpod. But after installing - updating Comfy/all - installing the missing nodes I get this message: Missing node types: workflowFlorence workflowFlux Controlnet workflownodes 3 worksflow sampler --- Any ideas as to why. Works fine in my local installation.

McGenie

This was the answer to the issue, thanks I appreciate your Big Brain!! haha

Otniel Ocampo

Solution found: After the 2nd restart, 'clear' the workflow that is present in ConfyUI and load the workflow fresh from the ULTIMATE_FLUX_ALL-IN-ONE.json. You can find the ULTIMATE_FLUX_ALL-IN-ONE.json as an attachment to this post: https://www.patreon.com/posts/112499790 The Florence error can be removed by creating the directory in the error by hand. The ComfyUI_NAI-mod error can be resolved by changing the 'directory' name in the corresponding py file. See this reddit post: https://www.reddit.com/r/comfyui/comments/1cqpjkf/comment/ligc0ie/. There appears to be a code error in the version that is being used in this instance. Simple text edit, do not be afraid to edit the py file. ;-)

Seanehty

So I see that this has been reported. Nodes are not being loaded: florence sampler controlnet Etc. I did follow the video instructions and on first run of the installer: Update confiyUI and restart Install missing, (there were 12) and restart I see the following errors in the console window: The system cannot find the path specified for the 'Florence2ModelLoader': 'D:\\ai\\Aitrepreneur\\ComfyUI_windows_portable\\ComfyUI\\models\\LLM Which is correct this path does not exist. This path also seems to be missing: custom_nodes/ComfyUI_NAI-mod/CSV Have I missed something fundamental? Thanks

Seanehty

I am a complete beginner and I run the one-klick installer - It worked perfectly - Thanks!!! Would it be possible for you to show how to make a Lora training for Flux?

Patrik Patrik

Thank you for the wonderful workflow for Comfy, you're doing an amazing job! If I may ask for one more thing, could you possibly add checkpoints to this workflow as well as some automation tools (multiple prompts at once, though I might be able to handle that myself)? I'm only wondering about one thing—sometimes when I use two LORA models (one of them being my own trained model), everything works very well and quickly (RTX 3090, i5-12600k, 64GB RAM). However, often after generating a few images or right from the start (even when everything worked fine before), the images take around 40 seconds or even more per iteration to generate. I don't know what causes this (I'm not overloading the GPU with things like YouTube videos), but it's a bit frustrating. Some LORA models from Civitai, even when used alone, also take a long time to generate images.

Tomasz

I fixed the error by updating the security level in the config to "weak". leaving that here for others who might encounter the same problem

VXLM

same error, exact same workflowFlorence workflowFlux Controlnet workflowNodes 3 workflowXLabs Sampler

VXLM

yep, I'm an idiot....figured this out

Jeff Rheinhart

Good evening, first timer here. I loaded the 1-click install, it installed everything except for Git (Win 10 x64). Got Git from it's site, installed it. Still missing the Manager button in the WebUI. I'm idiot I'll admit, and probably missed something....help, please?

Jeff Rheinhart

yeah because for downloading directly from civitai you need a free api key and use the wget command. For this you need to create an account on civitai, then go to settings, scroll down and create a new api key, save it somewhere. Then use the following command: wget "URL?token=YOURAPINUMBER" -O "nameofthemodel.extension" --content-disposition where URL Is in this case: https://civitai.com/api/download/models/932183 YOURAPINUMBER is your api and nameofthemodel.extension in this case is FadooStyleV2_r1.safetensors so basically you get something like this: wget "https://civitai.com/api/download/models/932183?token=YOURAPINUMBER" -O "FadooStyleV2_r1.safetensors" --content-disposition again just replace YOURAPINUMBER with your Api

Aitrepreneur

Command works but as you can see from this image, runpod isnt downloading the full lora from civtai and its causing me errors when i queue for generation https://imgur.com/a/ZsY73vP

Seeker Comics

uploading the checkpoint manually would take too much time, hence why you should dl it with the curl command: like for the the Q8 model for example: curl -L -o "flux1-dev-Q8_0.gguf" "https://huggingface.co/city96/FLUX.1-dev-gguf/resolve/main/flux1-dev-Q8_0.gguf?download=true" replace flux1-dev-Q8_0.gguf with the name of the model you want and https://huggingface.co/city96/FLUX.1-dev-gguf/resolve/main/flux1-dev-Q8_0.gguf?download=true with the model url

Aitrepreneur

I even rebooted! Anyways super easy. Love it

Kevin O'Malley

i tried putting the checkpoint in the folder but RunPod only put a portion of the checkpoint into the folder every time. And draging and dropping loras is not instantaneous like it used to; its very slow.

Seeker Comics

maybe all you needed was just to refresh the webui, everything was correclty installed but it displayed it as non installed

Aitrepreneur

yes, it's another webui

Aitrepreneur

Flux does not work on a1111. For the rest put the checkpoint in the models/unet folder and then select it in the model loader after refreshing

Aitrepreneur

so weird. After updating installing, reinstalling...etc.... I decided to load a workflow from xlabs from github which worked and ran it, then opened the ULTIMATE workflow and everything loaded. Ive been using comfy for over a year, this has never happened. anyway. works now.

Kevin O'Malley

I have this issue as well. I couldn't find these nodes online either. I reinstalled comfyui on a new runpod pod and went through the steps. This is the only error I see.

Vaivij

If i have used your installer for AUTO1111, can i still go ahead also use the Comfy-Flux auto install?

Soda Sullivan

I can confirm that I had this issue when I put ULTIMATE-FLUX_AUTO_INSTALL-RUNPOD.sh in 'workspace' directory insted of 'comfyui' directory as mentioned in the video. The issue went away when I moved the downloaded models to comfyui directory instead. So, if you are doing this again from scratch, make sure you copy over the file into 'comfyui' directory and not the 'workspace' directory.

Vaivij

Please, I beg you, make a one-click installation for Flux A1111 in RunPod! The current setup with ComfyUI has so many problems that even learning how to troubleshoot them is a steep learning curve. I'm experiencing LoRA issues with your workflow, as well as checkpoint problems. For instance, what if I want to use Flux FP8 alongside a different checkpoint for Flux? How can I handle that?

Seeker Comics

Please, I beg you, make a one-click installation for Flux in RunPod! The current setup with ComfyUI has so many problems that even learning how to troubleshoot them is a steep learning curve. I'm experiencing LoRA issues with your workflow, as well as checkpoint problems. For instance, what if I want to use Flux FP8 alongside a different checkpoint for Flux? How can I handle that?

Seeker Comics

I am an idiot NVM Civi auto installer put them in a differant spot

Mark Greenlaw

So I added tons of Loras but the Lora loader will not let me select them is there a work around for this

Mark Greenlaw

Having the same issue as well.

RenderDrgn

I'm on local comfy fresh install

Kevin O'Malley

Yes

Kevin O'Malley

have you followed the video guide? Updated comfyui + installed all the missing nodes?

Aitrepreneur

because you launched the installer as admin, that's why it installed in the system32 just launch the installer normally

Aitrepreneur

Difficult to find the issue when I can't replicate it on my side. I'll try doing something different for the V2 installer and V2 workflow that'll be released soon, so hopefully this issue will be solved with a different install process. But in the meantime make sure that you followed the process correctly and that you take notes of any errors you came across during install. Maybe a video of you installing it might help figure out where the issue lies.

Aitrepreneur

not found workflowFlorence workflowFlux Controlnet workflowNodes 3 workflowXLabs Sampler

Kevin O'Malley

Same issue - I tried this but with no luck. It's the same problem many others are experiencing, like the post below.

Sabsabinx

I apologize for the late reply! I figured out what it was. I installed this version of Flux and ComfyUI in the same folder I installed the previous version from a month ago. To fix the issue I simply installed everything in a new folder. It works like a charm

Pumpalooza

Hi, For me there is a problem with the local install, Once installed with the one click installer, nothing starts. Going to find the Comfyui folder, I tried starting it with either run_cpu.bat, run_nvidia_gpu.bat. It always gives me PermissionError: [WinError 5] Access Denied: 'C:\Windows\System32\ComfyUI_windows_portable\ComfyUI\user' C:\Windows\System32\ComfyUI_windows_portable>pause I am definitely doing something wrong, as an absolute beginner, and I apologize in advance for the trouble. Thank you!

Ludovico Celesti

https://www.patreon.com/posts/112499790

Sabsabinx

Hi, I have always the error of the impact pack custom nodes. Same as in the video you made. But after uninstalling and restarting it's still there, i tried multiple times. I'm on runpod.

Alex puliatti

Hi K and others. I have tried to install it once more time but I took special attention that I used correct version of python (included with comfyUI) and finally everything installed and imported correctly (I had to only manually install one module using "\ComfyUI_windows_portable\python_embeded\python.exe -m pip install aiohappyeyeballs" and install onnxruntime with support of CUDA "\ComfyUI_windows_portable\python_embeded\python.exe -m pip install onnxruntime-gpu" Upscaling dropped from 11 minutes to 4,5 mins :-) I thought that upscale will be much more precise like really the same image with higher res but it only took image, "read" it via flux and create a new image from description :D :D :D Anyway, thanks for your work, I will try to play aounr generation pictures :-)

BrokenBrush

@Aitrepreneur where is the json file?

Zind

Absolutely amazing, K. You’ve always made something as complicated as AI so simple and intuitive for bozos like myself, i want u to know how much we appreciate the effort you put in to removing all the stress from AI. ❤️

Fin

This is not a Zluda issue, this is probably a bad initial python install, you need to uninstall your current python installation and reinstall it correctly. Go to the add and remove programs, search for python and uninstall both the current python version and the python install program. Once this is done, go here and download this installer: https://www.python.org/ftp/python/3.10.11/python-3.10.11-amd64.exe Run it and check the “Add python 3.10 to Path” checkbox and continue with the installation. You can check that the right python version is installed by opening a new command prompt window and typing: python --version and it should give you the 3.10.11 version Then just relaunch the 1-click installer in a new folder and try again.

Aitrepreneur

Yeah, it's not working. I followed the instructions and i get an error. Error loading "E:\ComfyUI-Zluda\venv\lib\site-packages\torch\lib\caffe2_nvrtc.dll" or one of its dependencies.

Salvador Fonseca

Ok then try this: First uninstall the ComfyUI-GGUF node then go inside the ComfyUI_windows_portable folder then click on the folder path, type cmd then press enter, this will bring the command window and inside type: git clone https://github.com/city96/ComfyUI-GGUF ComfyUI/custom_nodes/ComfyUI-GGUF .\python_embeded\python.exe -s -m pip install -r .\ComfyUI\custom_nodes\ComfyUI-GGUF\requirements.txt then press enter. Then relaunch the webui.

Aitrepreneur

make sure that you updated comfyui first, then uninstall the node, restart the webui, then click install again, finally restart the webui and refresh otherwise if it still fails, you can try this: First uninstall the ComfyUI-GGUF node then. Go inside the ComfyUI_windows_portable folder then click on the folder path, type cmd then press enter, this will bring the command window and inside type: git clone https://github.com/city96/ComfyUI-GGUF ComfyUI/custom_nodes/ComfyUI-GGUF .\python_embeded\python.exe -s -m pip install -r .\ComfyUI\custom_nodes\ComfyUI-GGUF\requirements.txt then press enter. Then relaunch the webui.

Aitrepreneur

install Git for windows before launching the installer

Aitrepreneur

It does work, every AMD users that installed the ComfyUI-Zluda version works for them with the models installer

Aitrepreneur

Hi. I looked a few weeks back and didn't see any mention of this request, but an OUTPAINTING option like with A1111 would be nice, like Top, Bottom, Left and Right would be cool. Still working on my end just generating images. Thx for your installer. The only issue I ran into was on the first install it prematurely stopped, I think it was because of using my VPN during the install, FYI. I prefer the world of GUI boxes, Comfy UI is unbelievably complicated!

James Bowler

Hello Aitrepreneur! Quick question-I understand if it is not a quick answer, or if you even have time to answer at all haha: I was able to get all the custom nodes installed through comfyui's manager except "ComfyUI-GGUF". I have uninstalled it, restarted the webui, and tried to re-installing it multiple times to no avail. I made sure comfyui was updated, and I also tried updating python dependencies and ran into a few messages about things being skipped from "invalid metadata entry 'name' " (and I seem to need to update pip from 24.1.2 to 24.2, I can do that and report back)- not sure if this is leading to my mane issue of "Comfyui-GGUF not working though. Any suggestions? Thanks! Cheers,

Samantha

The ComfyUI-GGUF and x-flux-comfyui custom nodes fail to import. i've tried to instal them manualy and done so over and over again. but they still wont install.....

Arne starnd

Can I just get the workflow? I am on Mac and Linux.... I have everything already installed.

John Goad

Nice worked great! Just a few questions/possible requests for updates on the workflow. It's a great start, but wonder if it is possible to add: A batch render like forge and a111 has. Where you can pick a folder with images and run all of them during the night without needing to pick one by one? Or if there already is a feature like that, I saw batch built in to comfy, but it just seemed like a queue feature. lineart controlnet so you can color sketches, like you can in forge. Set a custom resolution on img 2 img like you can in forge for example. Anyway thanks for the first update, brought me a bit closer to using cumfy and flux. Though for my use-case forge is still more useable and gives better results with pony then flux in cumfy, but I am excited to see how it will develop. Lastly, if you have the time too, do you know where the setting is to tell comfy to look in another folder for models? Like it to be able to use the forge models without moving them. Also a question how do I load custom models in your workflow? Can't find where to select checkpoint files. Only Unet and diffisuion models, and if i past checkpoint into those folders I just get an error.

Daniel

Yeah, the zluda version doesn't work.

Salvador Fonseca

load them in the dedicated folder in runpod ( ConfyUI , models , Loras )

Bassplasma

There is no Manager Button after auto install

Long YuHe

Hi, can anyone tell me how to load my trained LoRA, I'm using runpod.

Алексей Политаев

what's your config?

Aitrepreneur

I've done exactly this multiple times and the error still pops up and other nodes don't work. Any suggestions?

Pumpalooza

Of course, I just wish they do and that you give us the news !!

H-BOX

as I explained in the video if you see a IMPORT FAILED error for one of the nodes, just uninstall it, restart the webui, then click install again for that particular node, then restart and refresh the webui again

Aitrepreneur

Hi there Aitrepreneur! I installed everything but ran into a problem with the nodes. When I go to install missing nodes, the missing nodes don't appear, only one thing: (IMPORT FAILED) ComfyUI-GGUF. I can only "Try Fix" or "Unistall" this thing. The Try Fix doesn't work. I had your previous version of FLUX and ComfyUI installed and working just fine, now it doesn't work. Sorry for any inconvenience and thanks for your help!

Pumpalooza

Man I think I figured it out, I didnt have Git installed, I dont remember seeing anything anywhere saying that was a requirement and I thought the .bat file would install everything needed. I have now installed Git and attempting the local install process again, wish me luck

ma

I think I did, but I will give it one more try later today.

Senjka Danhieux

Creating Xlabs directory... The system cannot find the path specified. The system cannot find the path specified.

ma

Folders: 2966 Files: 29520 Size: 5728853313 Compressed: 1576199635 Cloning ComfyUI-Manager... 'git' is not recognized as an internal or external command, operable program or batch file.

ma

ran the bat for the lower end gpu and the installer did not add all the necessary files and there is no cumfyui manager anywhere

ma

Zluda has been discontinued since 2022. But I'll give it a shot.

Salvador Fonseca

follow the video, I show how to start everything. Just a question of updating comfyui before installing the missing nodes

Aitrepreneur

no anyone with AMD should be using the special comfyui fork: https://github.com/patientx/ComfyUI-Zluda then put the MODELS-ULTIMATE_FLUX_INSTALL.bat file inside the comfyui folder and run it

Aitrepreneur

For people using an AMD GPU, install Anaconda, setup and run your environment with python 3.10, with your environment activated and in the main ComfyUI folder, do a "pip install torch-directml" and "pip install -r requirements.txt" command. Then run it with the --directml command when starting ComfyUI. You can create a .bat file with notepad or something like that. It should look like this below. @echo off call conda activate (Your Enrolment) python main.py --directml Then change it from a .txt file to a .bat file. Now all you must do is double click it to launch ComfyUI for an AMD GPU. Also, just for fun, I use tailscale for everything. So, if you all the --listen argument then use the tailscale IP address of the device you're running ComfyUI on with the port, you can remotely access it from anywhere. Just remember you must be using the same tailscale account on both devices. I use it on my laptop from anywhere I go from my desktop.

Salvador Fonseca

I tried it on runpod on 2 different pods, including the one that you adviced. When I hit Queue, I get this: Prompt outputs failed validation UnetLoaderGGUF: - Value not in list: unet_name: 'flux1-dev-Q8_0.gguf' not in [] DualCLIPLoaderGGUF: - Value not in list: clip_name2: 't5-v1_1-xxl-encoder-Q8_0.gguf' not in [] - Value not in list: clip_name1: 'ViT-L-14-TEXT-detail-improved-hiT-GmP-TE-only-HF.safetensors' not in [] VAELoader: - Value not in list: vae_name: 'ae.safetensors' not in ['sdxl_vae.safetensors']

Senjka Danhieux

If I used RunPod would your setup allow me to use one of the image to video models? If not, maybe you can look at doing a setup for such in the future? Thanks again!

Ewan Grantham

im getting the error when i try to queue, DualCLIPLoaderGGUF cannot mmap an empty file

Moses Son

I tried it with changing the .ini (ssl = True) but no luck so far. Same error: https://drive.google.com/file/d/1iDbVk4Jd6kC8mJcbj2oeuFkQV291OK65/view?usp=sharing But as I said, the image generation is working great and it's unbelievably fast on even on FP8. My RTX 6000 ADA is a blessing. I hope I can make the controlnets work, because that's where the real fun starts.

Violent Pixelation

Thanks the Installer ComfyUI Installer worked great!!

Otniel Ocampo

Thanks for the installer. stoked on it. However, once i installed the missing nodes when i restart the GUI window that opens in browser is blank

jack aguirre

Is it possible to merge QR code picture with a normal picture together using this? , if so please show me how.

Nabil Boulezaz

Trying it, and loving it. really, really looking forward to when we have the ability to do clothing transfer and/or reposer with Flux.

Tri3Ax

Hi, I love everything you've done, two questions. Is there a way to extract the seed from an image that has already been created in order to recreate it? How can I set the seed? When I enter it later it changes again. Thanks

Gabo

what errors?

Aitrepreneur

meah my lora's don't seem to work and I get weird errors all over when running this on runpod. bit disappointed tbh.

Robert

I don't know any paid services that allow nsfw generation but vram is king ai tools, so your system is too weak to run it at a fast speed, you might try using it on runpod though

Aitrepreneur

hi, yes of course

Aitrepreneur

what link?

Aitrepreneur

yes but you need to install the special comfyui AMD version here: https://github.com/patientx/ComfyUI-Zluda once it's done you can probably just use the MODELS-ULTIMATE_FLUX_INSTALL.bat file to install all the missing models (put the bat file inside the comfyui folder and then run it)

Aitrepreneur

Do this works with AMD Video cards?? Hopefully it works!!

Otniel Ocampo

fixed the issue saw a git error in the install logs installed latest version of git for windows and reinstalled and ui appeared correctly

Solutheis

It says link expired, can we get an updated link

CuriousCat

so running into a issue following the install video and i got to importing the json file for workflow but there is no manager button to update anything

Solutheis

I presume this is just for image generation. I ask because I'm working on a project where I have my images of my characters, and have seen some of the examples out there for CogVideoX-I2V and others. However the project display is 1920x1080 and most of the guides I read are for 720x480. I presume I could upscale with a video editor but probably lose some fidelity. In any event, I was hoping to be able to feed a base image in, say with two girls one of whom is wearing a strap-on, and get a 2-4 second loopable animation out. I've seen a lot of online services that claim to do this, but none of them seem to allow for NSFW generation. If I install locally, I have one machine that is an i9 12900H with an iGPU and 32 gigs of RAM, and another that's an i9 10900KF with 32 gigs of RAM and an RTX 3070 (8GB VRAM). With those specs, I wasn't sure if I'd do better with the newer machine since I assume the Intel iGPU uses shared RAM (so more, but slower) or the one with the dedicated VRAM if I went to install something local? Which is a long way of asking: 1) Is there any service (paid is fine if not too expensive) that will do NSFW Image to NSFW Video online? 2) If not, which machine should I do a local install on, and is the CogVideoX-I2V model my best choice if I need a 1920x1080 end resolution? Thanks in advance!

Ewan Grantham

Really depends on your whole configuration while generating image Might be due to resolution, use configuration file from Aitrepreneur and really try again. I installed Q4 model for lower vram to make it easier for my PC and it works perfectly

Filip Šenjug

I'm a new patreon member and just one to say holy cow, thank you for your effort! It's absolutely worth every penny of the subscription

Herb

Why are so many generated images blurred?

Lukas

Ok - never mind. I did a reinstall of comfy and it's all set now. Those of you having an issue with this , I recommend you just backup your current install and start fresh. Then you can move your models into the new install.

Eric Wisniewski

This is amazing work. I really appreciate it. It's one of the reasons I pay for the Patreon. I really got into your content when you started doing style training in SDLX. I'm hoping you'll have a video soon on how to train not just people into flux but styles specifically.

Kenneth Loebenberg

Hello! I installed everything and I am getting an error for ComfyUi-GGUF. I tried to reinstall and didn't work. Then tried a GIT pull and also didn't work. Any thoughts on how to get your workflow working? EDIT: Error when loading workflow: When loading the graph, the following node types were not found: FluxGuidance (In group node 'workflow/positive prompt guidance') workflow/positive prompt guidanceRemove from workflow workflow/Negative PromptRemove from workflow ModelSamplingFlux (In group node 'workflow/Other Nodes') workflow/Other NodesRemove from workflow DualCLIPLoaderGGUF UnetLoaderGGUF Nodes that have failed to load will show as red on the graph.

Eric Wisniewski

Thanks so much for this! Every idea to solve this problem is welcome. I will try this! Have a great weekend.

Violent Pixelation

send screenshots of the webui in dm, and the node connection, also what option are you trying to use?

Aitrepreneur

everything was working great a few days ago but today its spitting an error called "prompt_no_outputs" other workflows are working but not yours. I don't know what is going on. Log details are very minimal got prompt invalid prompt: {'type': 'prompt_no_outputs', 'message': 'Prompt has no outputs', 'details': '', 'extra_info': {}} Everything starting up as usual. Tried a different prompt loader that works in other workflow but still no good. Removed new nodes I had installed - still no good. Shame bc its such a great workflow

Orpheus_Reup

Looks like it did...thanks! By the way, something like this and, especially, your workflow, is EXACTLY why I became a subscriber to your patreon. This is terrific work. Thanks so much!

Ken L. Black

That's not up to me, it's up to the people making tooncrafter, if they don't release the models and code, nothing we can do

Aitrepreneur

never saw this issue before, looks like an ssl issue though, you could try this: go to the ComfyUI\custom_nodes\ComfyUI-Manager folder and edit the config.ini file and change on line 7 from bypass_ssl = False to bypass_ssl = True save the file then relaunch comfyui (and maybe also click "update all" in the manager before doing anything else)

Aitrepreneur

that's because you are running it as admin, just run the installer normally

Aitrepreneur

Ahh, I figured out activating them is in the right-click menu. It's none too obvious but I'm very new to comfyui. Thanks! Everything works just as advertised.

Dylan Barnett

Hi ! I'm not sure if you will read this, but I REALLY need to test the function "sketch reference" for tooncrafter. Could you please update about it as soon as it is available ? Thank you so much for all your work !

H-BOX

Hi, I'm able now to generate images but EVERY pre-processor brings an error (depth...etc..I also tried open-pose,....)????? Anyone with the same problem? (MaxRetryError("HTTPSConnectionPool(host='cdn-lfs.hf.co', port=443): Max retries exceeded with url: /repos/a2/11/a211ef997fa284ad67caa591ca931692f9a8a018910b18a04d9e6dff396f02e3/501f0c75b3bca7daec6b3682c5054c09b366765aef6fa3a09d03a5cb4b230853?response-content-disposition=inline%3B+filename*%3DUTF-8%27%27dpt_hybrid-midas-501f0c75.pt%3B+filename%3D%22dpt_hybrid-midas-501f0c75.pt%22%3B&Expires=1727680490&Policy=eyJTdGF0ZW1lbnQiOlt7IkNvbmRpdGlvbiI6eyJEYXRlTGVzc1RoYW4iOnsiQVdTOkVwb2NoVGltZSI6MTcyNzY4MDQ5MH19LCJSZXNvdXJjZSI6Imh0dHBzOi8vY2RuLWxmcy5oZi5jby9yZXBvcy9hMi8xMS9hMjExZWY5OTdmYTI4NGFkNjdjYWE1OTFjYTkzMTY5MmY5YThhMDE4OTEwYjE4YTA0ZDllNmRmZjM5NmYwMmUzLzUwMWYwYzc1YjNiY2E3ZGFlYzZiMzY4MmM1MDU0YzA5YjM2Njc2NWFlZjZmYTNhMDlkMDNhNWNiNGIyMzA4NTM~cmVzcG9uc2UtY29udGVudC1kaXNwb3NpdGlvbj0qIn1dfQ__&Signature=SWLAw4kvVV3Vz8NJG2GPdLYDyKcLQl1vR1858dbohoQZpnxSgATPmWA4jxhOdVh42oS33DZmwtgcB5xybkTk4ekAJQCEIulIn5RI7qYfeGmqSCOVpXDe2SCwzDE1H6YhballAr3Z~OTadfGyYjsT8Z8jQmzegRWpZ~MIK-Y612FOeCjLe7AUaHVylCkrmYjM9xeoHcNKrPV5XFuSbTdmXMcfhruWCE8VyKfF~nXBX32bd-ZFx4C-6X1Cilo2zoTzPH70YJxBLwNViVDSpXF28DF1r4AxlhSRhMZqcfgRscmnqFKyfDgJNMTVcJg430A44xI7qYub2sV6yQxD-o7i5Q__&Key-Pair-Id=K3RPWS32NSSJCE (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1006)')))"), '(Request ID: 660490b1-c267-41ab-a8a3-c8f9b6a68ef9)')

Violent Pixelation

Is there a way to run the all in one install batch file without it trying to install this all in my windows system folder? I just want it in a folder on my desktop for simplicity. Also it being there keeps causing write privilege issues. thanks Also the Xlabs stuff fails to install due to message saying "no such path" Creating Xlabs directory... The system cannot find the path specified. The system cannot find the path specified.

Rodger Hurlburt

I installed it on my work computer, and the box doesn't have those options on it either. So I'm guessing there's something I'm doing wrong.

JpaLVVmI6c8xM00

no it shouldn't impact anything, it's the exact same installation, just different models. What exactly happens now that you uninstalled 7zip from the add and remove programs? It should now reinstall 7zip from scratch

Aitrepreneur

No luck. The exact same thing happened. It leaves behind a 7-zip.exe file in the same directory, which I also then delete, try it again, and same thing. HOWEVER, if I just choose option 1, for a GPU below 12GB, then it seems to run just fine. Does it autodetect your video card somehow? (I have EXACTLY 12GB on my GPU.) I suppose I can always run your MODELS bat file to get any models I may still want to try.

Ken L. Black

uninstall 7zip completely then simply try the installer again

Aitrepreneur

Now my CMD output is missing "only" ComfyUI-GGUF: Microsoft Windows [Version 10.0.19045.4894] (c) Microsoft Corporation. Všechna práva vyhrazena. e:\NET\AI\ComfyUI_windows_portable>run_nvidia_gpu.bat e:\NET\AI\ComfyUI_windows_portable>.\python_embeded\python.exe -s ComfyUI\main.py --windows-standalone-build [START] Security scan WARNING: Ignoring invalid distribution ~arl (e:\NET\AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages) [DONE] Security scan ## ComfyUI-Manager: installing dependencies done. ** ComfyUI startup time: 2024-09-26 21:01:31.219879 ** Platform: Windows ** Python version: 3.11.8 (tags/v3.11.8:db85d51, Feb 6 2024, 22:03:32) [MSC v.1937 64 bit (AMD64)] ** Python executable: e:\NET\AI\ComfyUI_windows_portable\python_embeded\python.exe ** ComfyUI Path: e:\NET\AI\ComfyUI_windows_portable\ComfyUI ** Log path: e:\NET\AI\ComfyUI_windows_portable\comfyui.log Prestartup times for custom nodes: 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\rgthree-comfy 2.1 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager Total VRAM 12281 MB, total RAM 49082 MB pytorch version: 2.4.1+cu121 Set vram state to: NORMAL_VRAM Device: cuda:0 NVIDIA GeForce RTX 4070 : cudaMallocAsync Using pytorch cross attention e:\NET\AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\kornia\feature\lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead. @torch.cuda.amp.custom_fwd(cast_inputs=torch.float32) Traceback (most recent call last): File "e:\NET\AI\ComfyUI_windows_portable\ComfyUI\nodes.py", line 1931, in load_custom_node module_spec.loader.exec_module(module) File "", line 940, in exec_module File "", line 241, in _call_with_frames_removed File "E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF\__init__.py", line 7, in from .nodes import NODE_CLASS_MAPPINGS File "E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF\nodes.py", line 13, in from .ops import GGMLTensor, GGMLOps, move_patch_to_device File "E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF\ops.py", line 158, in class GGMLOps(comfy.ops.manual_cast): File "E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF\ops.py", line 182, in GGMLOps class Embedding(GGMLLayer, comfy.ops.manual_cast.Embedding): ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ AttributeError: type object 'manual_cast' has no attribute 'Embedding' Cannot import E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF module for custom nodes: type object 'manual_cast' has no attribute 'Embedding' ### Loading: ComfyUI-Impact-Pack (V7.6.2) ### Loading: ComfyUI-Impact-Pack (Subpack: V0.6) [Impact Pack] Wildcards loading done. Total VRAM 12281 MB, total RAM 49082 MB pytorch version: 2.4.1+cu121 Set vram state to: NORMAL_VRAM Device: cuda:0 NVIDIA GeForce RTX 4070 : cudaMallocAsync ### Loading: ComfyUI-Manager (V2.51.2) ### ComfyUI Revision: 2359 [29c2e267] | Released on '2024-07-12' e:\NET\AI\ComfyUI_windows_portable ############################################ e:\NET\AI\ComfyUI_windows_portable\custom_nodes\ComfyUI-Universal-Styler\CSV ############################################ [] ############################################ [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/model-list.json [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/github-stats.json [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/alter-list.json [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/custom-node-list.json [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/extension-node-map.json e:\NET\AI\ComfyUI_windows_portable ############################################ e:\NET\AI\ComfyUI_windows_portable\custom_nodes\ComfyUI-Universal-Styler\CSV ############################################ [] ############################################ [comfyui_controlnet_aux] | INFO -> Using ckpts path: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui_controlnet_aux\ckpts [comfyui_controlnet_aux] | INFO -> Using symlinks: False [comfyui_controlnet_aux] | INFO -> Using ort providers: ['CUDAExecutionProvider', 'DirectMLExecutionProvider', 'OpenVINOExecutionProvider', 'ROCMExecutionProvider', 'CPUExecutionProvider', 'CoreMLExecutionProvider'] E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui_controlnet_aux\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") [rgthree] Loaded 42 exciting nodes. [rgthree] Will use rgthree's optimized recursive execution. WAS Node Suite: OpenCV Python FFMPEG support is enabled WAS Node Suite Warning: `ffmpeg_bin_path` is not set in `E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\was-node-suite-comfyui\was_suite_config.json` config file. Will attempt to use system ffmpeg binaries if available. WAS Node Suite: Finished. Loaded 218 nodes successfully. "The best way to predict the future is to create it." - Peter Drucker Import times for custom nodes: 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\websocket_image_save.py 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Universal-Styler 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-NAI-styler 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Custom-Scripts 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI_UltimateSDUpscale 0.0 seconds (IMPORT FAILED): E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-KJNodes 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\rgthree-comfy 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\x-flux-comfyui 0.1 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui_controlnet_aux 0.1 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui-tensorops 0.3 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager 0.4 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Impact-Pack 0.5 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui-ollama 0.6 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Florence2 2.6 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\was-node-suite-comfyui Starting server To see the GUI go to: http://127.0.0.1:8188 FETCH DATA from: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\extension-node-map.json [DONE] []

BrokenBrush

I am trying to install ComfyUI on an external SSD file from scratch (even though I already have it installed), as you suggested. I copied the .bat file into a newly-created folder. Whenever I try running the .bat file, it tells me I do not have 7-zip installed. When I click YES to the 7-Zip installation about whether I want to make changes to my device, I say YES, and the .bat file simply stops running. I DO have 7-zip installed in a directory on that same SSD drive, but it makes no difference. I go through this same thing time after time. I've tried about 5 times and nothing works. So...I simply tried the same thing on my INTERNAL SSD drive. (Not optimal, but I'll just move the installation over when I get it installed.) Same issue. The .bat file just closes. Any suggestions?

Ken L. Black

I was able to fix ComfyUI-NAI-Styler https://github.com/KoreTeknology/ComfyUI-Universal-Styler/issues/1 Although for me fix from the link wasn't working but I just clone and renamed whole folder "ComfyUI-Universal-Styler" to "ComfyUI-NAI-styler"

BrokenBrush

Hi K, I used "ULTIMATE_FLUX-COMFYUI-MANAGER_AUTO_INSTALL.bat" but I have failed import ComfyUI-GGUF. This is the error during loading: "When loading the graph, the following node types were not found: FluxGuidance (In group node 'workflow/positive prompt guidance') workflow/positive prompt guidance workflow/Negative Prompt ModelSamplingFlux (In group node 'workflow/Other Nodes') workflow/Other Nodes DualCLIPLoaderGGUF UnetLoaderGGUF Nodes that have failed to load will show as red on the graph." I was able to fix some issues when I tried to update comfy and additional install for comfy manager as it was missing but still there is one import failed (that ComfyUI-GGUF). This is from cmd: "Prestartup times for custom nodes: 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\rgthree-comfy 3.3 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager Total VRAM 12281 MB, total RAM 49082 MB pytorch version: 2.4.1+cu121 Set vram state to: NORMAL_VRAM Device: cuda:0 NVIDIA GeForce RTX 4070 : cudaMallocAsync Using pytorch cross attention e:\NET\AI\ComfyUI_windows_portable\python_embeded\Lib\site-packages\kornia\feature\lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead. @torch.cuda.amp.custom_fwd(cast_inputs=torch.float32) Traceback (most recent call last): File "e:\NET\AI\ComfyUI_windows_portable\ComfyUI\nodes.py", line 1931, in load_custom_node module_spec.loader.exec_module(module) File "", line 940, in exec_module File "", line 241, in _call_with_frames_removed File "E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF\__init__.py", line 7, in from .nodes import NODE_CLASS_MAPPINGS File "E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF\nodes.py", line 13, in from .ops import GGMLTensor, GGMLOps, move_patch_to_device File "E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF\ops.py", line 158, in class GGMLOps(comfy.ops.manual_cast): File "E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF\ops.py", line 182, in GGMLOps class Embedding(GGMLLayer, comfy.ops.manual_cast.Embedding): ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ AttributeError: type object 'manual_cast' has no attribute 'Embedding' Cannot import E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF module for custom nodes: type object 'manual_cast' has no attribute 'Embedding' ### Loading: ComfyUI-Impact-Pack (V7.6.2) ### Loading: ComfyUI-Impact-Pack (Subpack: V0.6) [Impact Pack] Wildcards loading done. Total VRAM 12281 MB, total RAM 49082 MB pytorch version: 2.4.1+cu121 Set vram state to: NORMAL_VRAM Device: cuda:0 NVIDIA GeForce RTX 4070 : cudaMallocAsync ### Loading: ComfyUI-Manager (V2.51.1) ### ComfyUI Revision: 2359 [29c2e267] | Released on '2024-07-12' [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/model-list.json [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/alter-list.json [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/github-stats.json [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/extension-node-map.json [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/custom-node-list.json e:\NET\AI\ComfyUI_windows_portable ############################################ e:\NET\AI\ComfyUI_windows_portable\custom_nodes\ComfyUI-NAI-styler\CSV ############################################ [] ############################################ [comfyui_controlnet_aux] | INFO -> Using ckpts path: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui_controlnet_aux\ckpts [comfyui_controlnet_aux] | INFO -> Using symlinks: False [comfyui_controlnet_aux] | INFO -> Using ort providers: ['CUDAExecutionProvider', 'DirectMLExecutionProvider', 'OpenVINOExecutionProvider', 'ROCMExecutionProvider', 'CPUExecutionProvider', 'CoreMLExecutionProvider'] Traceback (most recent call last): File "e:\NET\AI\ComfyUI_windows_portable\ComfyUI\nodes.py", line 1931, in load_custom_node module_spec.loader.exec_module(module) File "", line 936, in exec_module File "", line 1073, in get_code File "", line 1130, in get_data FileNotFoundError: [Errno 2] No such file or directory: 'E:\\NET\\AI\\ComfyUI_windows_portable\\ComfyUI\\custom_nodes\\ComfyUI_NAI-mod\\__init__.py' Cannot import E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI_NAI-mod module for custom nodes: [Errno 2] No such file or directory: 'E:\\NET\\AI\\ComfyUI_windows_portable\\ComfyUI\\custom_nodes\\ComfyUI_NAI-mod\\__init__.py' [rgthree] Loaded 42 magnificent nodes. [rgthree] Will use rgthree's optimized recursive execution. WAS Node Suite: OpenCV Python FFMPEG support is enabled WAS Node Suite Warning: `ffmpeg_bin_path` is not set in `E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\was-node-suite-comfyui\was_suite_config.json` config file. Will attempt to use system ffmpeg binaries if available. WAS Node Suite: Finished. Loaded 218 nodes successfully. "The best revenge is massive success." - Frank Sinatra Import times for custom nodes: 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\websocket_image_save.py 0.0 seconds (IMPORT FAILED): E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI_NAI-mod 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Universal-Styler 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Custom-Scripts 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI_UltimateSDUpscale 0.0 seconds (IMPORT FAILED): E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\rgthree-comfy 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-KJNodes 0.0 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\x-flux-comfyui 0.1 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui_controlnet_aux 0.3 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\comfyui-tensorops 0.3 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager 0.3 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Impact-Pack 0.5 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Florence2 2.4 seconds: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\was-node-suite-comfyui Starting server To see the GUI go to: http://127.0.0.1:8188 FETCH DATA from: E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\extension-node-map.json [DONE] [ERROR] An error occurred while retrieving information for the 'Florence2ModelLoader' node. Traceback (most recent call last): File "e:\NET\AI\ComfyUI_windows_portable\ComfyUI\server.py", line 435, in get_object_info out[x] = node_info(x) ^^^^^^^^^^^^ File "e:\NET\AI\ComfyUI_windows_portable\ComfyUI\server.py", line 412, in node_info info['input'] = obj_class.INPUT_TYPES() ^^^^^^^^^^^^^^^^^^^^^^^ File "E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Florence2\nodes.py", line 148, in INPUT_TYPES "model": ([item.name for item in Path(folder_paths.models_dir, "LLM").iterdir() if item.is_dir()], {"tooltip": "models are expected to be in Comfyui/models/LLM folder"}), ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\NET\AI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Florence2\nodes.py", line 148, in "model": ([item.name for item in Path(folder_paths.models_dir, "LLM").iterdir() if item.is_dir()], {"tooltip": "models are expected to be in Comfyui/models/LLM folder"}), ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "pathlib.py", line 931, in iterdir FileNotFoundError: [WinError 3] Systém nemůže nalézt uvedenou cestu: 'E:\\NET\\AI\\ComfyUI_windows_portable\\ComfyUI\\models\\LLM' Error. No naistyles.csv found. Put your naistyles.csv in the custom_nodes/ComfyUI_NAI-mod/CSV directory of ComfyUI. Then press "Refresh". Your current root directory is: E:\NET\AI\ComfyUI_windows_portable\ComfyUI Error. No naistyles.csv found. Put your naistyles.csv in the custom_nodes/ComfyUI_NAI-mod/CSV directory of ComfyUI. Then press "Refresh". Your current root directory is: E:\NET\AI\ComfyUI_windows_portable\ComfyUI Error. No naistyles.csv found. Put your naistyles.csv in the custom_nodes/ComfyUI_NAI-mod/CSV directory of ComfyUI. Then press "Refresh". Your current root directory is: E:\NET\AI\ComfyUI_windows_portable\ComfyUI"

BrokenBrush

Sorry, super noob question. I can’t find the update comfy button so I can install the missing nodes.

JpaLVVmI6c8xM00

just install it again manually and autoinstaller will continue ;)

BrokenBrush

When I hit Queue, I get this: Prompt outputs failed validation UnetLoaderGGUF: - Value not in list: unet_name: 'flux1-dev-Q8_0.gguf' not in [] DualCLIPLoaderGGUF: - Value not in list: clip_name2: 't5-v1_1-xxl-encoder-Q8_0.gguf' not in [] - Value not in list: clip_name1: 'ViT-L-14-TEXT-detail-improved-hiT-GmP-TE-only-HF.safetensors' not in [] VAELoader: - Value not in list: vae_name: 'ae.safetensors' not in ['sdxl_vae.safetensors']

Senjka Danhieux

This is an absolutely fantastic resource! The installer worked perfectly for me and your YouTube instructions were flawless. You've made the Comfy interface manageable, which is no easy task. Outstanding work on all fronts, good sir!

Joel o' St. Pete

You should include the download to the .JSON file here, too, so that users are not confused/missing steps.

TheFoodMage

Love the workflow models though, you did most of the legwork for us, thank you!!!

Jeffrey Lenz

The problem I had was that it loaded zero models, unlike your previous autoinstallers, they had worked from the get-go.

Jeffrey Lenz

Thank you for everything! Best installation and workflow. I'm off and generating images from the get go. :D

Maxwell Leung

this is a known bug in Jupyter lab, can't do much unfortunately, except use the cd command to go inside the folder

Aitrepreneur

probably because you are running as admin rights, just run the installer normally

Aitrepreneur

I am trying to run the Pods terminal you created, but for some reason it will not let me open the Checkpoints folder in JypyterLab. Any reason why? I deleted everything and started over, but still no luck.

Daniel Lohneiss

This is an awesome one click, thank you for putting this together! I tried getting it installed but didn't realize it was using my c: drive as default. I put all my apps on my D: drive as it's huge. Is there any way to provide an option for where it'll be installed in the future? I in fact ran the .bat from my D: drive but it still apparently installs to my C: Thanks!

Serial Crush

Hi, thank you for your reply. Yes, I've done that, too. I'm in the process of reinstalling everyting but this time with full admin rights. Maybe this helps. I NEED THIS WORKFLOW IN MY LIFE! IT MUST WORK :D

Violent Pixelation

Why I'm not seeing the "Download as an archive" option when right-clicking on a folder in Jupyter Lab, even though the archive extension is installed and enabled in Aitrepreneur Comfyui pod?

MakuTu

If you have an issue with one of the nodes, do it like I showed for the runpod install. First uninstall the problem node, restart, then click install on the missing node, restart again and refresh the page with F5

Aitrepreneur

have you activated the area you want to use first?

Aitrepreneur

Same issue. Can't seem to figure it out.

Christopher Moonlight Productions

I think it installed ok but when I click Queue Prompt I get a popup saying prompt has no outputs. It looks like it has outputs but thats what it says.

Dylan Barnett

No worries I will ;)

Aitrepreneur

Hi red baron, oh I've done that several times but those nodes are still missing. I don't know if my company keeps on blocking something here!?

Violent Pixelation

yeahhhh :)

red baron

With this, its time to add a video how to create Loras locally for FLUX. Videos out there on this topic makes my brain spin. Not Aitrepreneury :D

OFF:A Our Fabulous Future Apocalypse

i have the same problem. click the restart button in "manager / install missing nodes" refresh browser

red baron

SOLVED 50%: I've solved it using my full admin account in my company and I can generate images now. BUT..... EVERY pre-processor brings an error (depth....I also tried open-pose,....)????? I'll make another comment just for this! (MaxRetryError("HTTPSConnectionPool(host='cdn-lfs.hf.co', port=443): Max retries exceeded with url: /repos/a2/11/a211ef997fa284ad67caa591ca931692f9a8a018910b18a04d9e6dff396f02e3/501f0c75b3bca7daec6b3682c5054c09b366765aef6fa3a09d03a5cb4b230853?response-content-disposition=inline%3B+filename*%3DUTF-8%27%27dpt_hybrid-midas-501f0c75.pt%3B+filename%3D%22dpt_hybrid-midas-501f0c75.pt%22%3B&Expires=1727680490&Policy=eyJTdGF0ZW1lbnQiOlt7IkNvbmRpdGlvbiI6eyJEYXRlTGVzc1RoYW4iOnsiQVdTOkVwb2NoVGltZSI6MTcyNzY4MDQ5MH19LCJSZXNvdXJjZSI6Imh0dHBzOi8vY2RuLWxmcy5oZi5jby9yZXBvcy9hMi8xMS9hMjExZWY5OTdmYTI4NGFkNjdjYWE1OTFjYTkzMTY5MmY5YThhMDE4OTEwYjE4YTA0ZDllNmRmZjM5NmYwMmUzLzUwMWYwYzc1YjNiY2E3ZGFlYzZiMzY4MmM1MDU0YzA5YjM2Njc2NWFlZjZmYTNhMDlkMDNhNWNiNGIyMzA4NTM~cmVzcG9uc2UtY29udGVudC1kaXNwb3NpdGlvbj0qIn1dfQ__&Signature=SWLAw4kvVV3Vz8NJG2GPdLYDyKcLQl1vR1858dbohoQZpnxSgATPmWA4jxhOdVh42oS33DZmwtgcB5xybkTk4ekAJQCEIulIn5RI7qYfeGmqSCOVpXDe2SCwzDE1H6YhballAr3Z~OTadfGyYjsT8Z8jQmzegRWpZ~MIK-Y612FOeCjLe7AUaHVylCkrmYjM9xeoHcNKrPV5XFuSbTdmXMcfhruWCE8VyKfF~nXBX32bd-ZFx4C-6X1Cilo2zoTzPH70YJxBLwNViVDSpXF28DF1r4AxlhSRhMZqcfgRscmnqFKyfDgJNMTVcJg430A44xI7qYub2sV6yQxD-o7i5Q__&Key-Pair-Id=K3RPWS32NSSJCE (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1006)')))"), '(Request ID: 660490b1-c267-41ab-a8a3-c8f9b6a68ef9)') EDIT: THIS SEEMS TO BE THE PROBLEM! https://drive.google.com/file/d/1ZUxT7y8rkgB-7z3kjZER0-VnExSeBWhh/view?usp=sharing And I tried to lower the security level of "config.ini" but it wasn't working, either. I think unfortunately it's my company's firewall. Hi everyone, there are still a couple of "red boxes" left after updating everything: For example the load model box is read. Has anyone an idea what I'm doing wrong? image: https://drive.google.com/file/d/1Wlwuq-RAMw1n-M-i59TfZIU-UPogRXzn/view?usp=sharing " When loading the graph, the following node types were not found: FluxGuidance (In group node 'workflow/positive prompt guidance') workflow/positive prompt guidanceRemove from workflow workflow/Negative PromptRemove from workflow ModelSamplingFlux (In group node 'workflow/Other Nodes') workflow/Other NodesRemove from workflow DualCLIPLoaderGGUF UnetLoaderGGUF Nodes that have failed to load will show as red on the graph. "

Violent Pixelation

just want to say thanks for all the work you do AItrepreneur, but what sort of times are people getting for creating an image? with FP8 it takes around 4-5 minutes and with GGUF it's around a minute on a 3080ti, with SDXL I can pump out an image in about 10 seconds at most, maybe I've done something wrong

Adders

Hi all, just one click installed worked like a champ didn't grumble, but when I hit q to generate after following the video I get failed and a attribute error 509:clipTestEncode. I am auto 1111 before first time comfy ui I am sure this is an easy I am missing but any help would be great.. okay update kick the tires and poked all the friendly buttons update drivers and reload poof all good thanks for the installer

McNutt family

I've installed Git, made some changes in PATH. I've installed on two separate computers, different directories... Please help.

Garrett deRosset

It says Installation of 7-Zip failed. Please install it manually and try again. I have it installed already

HL

Some issues: Enter your choice (A or B) and press Enter: a Downloading ComfyUI... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 100 1503M 100 1503M 0 0 15.9M 0 0:01:34 0:01:34 --:--:-- 15.6M Extracting ComfyUI... 7-Zip 24.01 (x64) : Copyright (c) 1999-2024 Igor Pavlov : 2024-01-31 Scanning the drive for archives: 1 file, 1576199635 bytes (1504 MiB) Extracting archive: ComfyUI_windows_portable_nvidia_cu121_or_cpu.7z -- Path = ComfyUI_windows_portable_nvidia_cu121_or_cpu.7z Type = 7z Physical Size = 1576199635 Headers Size = 354715 Method = LZMA2:25 LZMA:20 BCJ2 Solid = + Blocks = 1 Everything is Ok Folders: 2966 Files: 29520 Size: 5728853313 Compressed: 1576199635 Cloning ComfyUI-Manager... 'git' is not recognized as an internal or external command, operable program or batch file. Downloading VAE files... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1127 100 1127 0 0 1572 0 --:--:-- --:--:-- --:--:-- 1576 100 319M 100 319M 0 0 12.2M 0 0:00:26 0:00:26 --:--:-- 14.9M Downloading CLIP files... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1225 100 1225 0 0 1981 0 --:--:-- --:--:-- --:--:-- 1985 100 308M 100 308M 0 0 14.3M 0 0:00:21 0:00:21 --:--:-- 14.8M % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1151 100 1151 0 0 1852 0 --:--:-- --:--:-- --:--:-- 1859 100 9334M 100 9334M 0 0 16.8M 0 0:09:13 0:09:13 --:--:-- 18.3M Creating Xlabs directory... The system cannot find the path specified. The system cannot find the path specified. % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1169 100 1169 0 0 1869 0 --:--:-- --:--:-- --:--:-- 1873 100 1418M 100 1418M 0 0 16.7M 0 0:01:24 0:01:24 --:--:-- 16.0M % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1177 100 1177 0 0 1949 0 --:--:-- --:--:-- --:--:-- 1955 100 1418M 100 1418M 0 0 17.6M 0 0:01:20 0:01:20 --:--:-- 19.4M % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1167 100 1167 0 0 1857 0 --:--:-- --:--:-- --:--:-- 1864 100 1418M 100 1418M 0 0 17.1M 0 0:01:22 0:01:22 --:--:-- 17.9M Downloading the upscaler model... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1139 100 1139 0 0 1877 0 --:--:-- --:--:-- --:--:-- 1882 100 8804k 100 8804k 0 0 3591k 0 0:00:02 0:00:02 --:--:-- 7373k Downloading UNET and related models... Downloading flux1-dev-fp8.safetensors... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1155 100 1155 0 0 1887 0 --:--:-- --:--:-- --:--:-- 1896 100 11.0G 100 11.0G 0 0 13.5M 0 0:13:56 0:13:56 --:--:-- 13.4M Downloading flux1-dev-Q8_0.gguf... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1149 100 1149 0 0 1912 0 --:--:-- --:--:-- --:--:-- 1921 100 11.8G 100 11.8G 0 0 16.8M 0 0:11:59 0:11:59 --:--:-- 19.9M Downloading flux1-schnell-Q8_0.gguf... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1151 100 1151 0 0 1856 0 --:--:-- --:--:-- --:--:-- 1859 64 11.8G 64 7767M 0 0 11.6M 0 0:17:21 0:11:08 0:06:13 13.8M curl: (56) Recv failure: Connection was reset Downloading t5-v1_1-xxl-encoder-Q8_0.gguf... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 0 0 0 0 0 0 0 0 --:--:-- 0:05:15 --:--:-- 0 curl: (28) Resolving timed out after 315582 milliseconds ComfyUI and FLUX installed. Running ComfyUI... HAVE FUN ;) Total VRAM 24564 MB, total RAM 32676 MB pytorch version: 2.3.1+cu121 Set vram state to: NORMAL_VRAM Device: cuda:0 NVIDIA GeForce RTX 3090 Ti : cudaMallocAsync Using pytorch cross attention ****** User settings have been changed to be stored on the server instead of browser storage. ****** ****** For multi-user setups add the --multi-user CLI argument to enable multiple user profiles. ****** Import times for custom nodes: 0.0 seconds: C:\Users\Garrett\Downloads\ComfyUI_windows_portable\ComfyUI\custom_nodes\websocket_image_save.py Starting server To see the GUI go to: http://127.0.0.1:8188

Garrett deRosset

It works great, without any errors, but when I close the terminal and start again, I always get an error about importing some custom nodes. After selecting all failed nodes in the manager and pressing 'Try fix', and then restarting, it works again.

Stanislav

Most LORA files are not in the flux format. One solution created a custom 1.5. or SDXL flow to inject it with VAE. Is there an easier method? Also, for custom LORA files, is there an easy tool to regenerate as Flux models?

Bob O

I closed it all out to try the prompt generator, and when I started again the ultimate was gone. I grabbed the json file and put it in the comfyUI/user/default/workflows and restarted and I have it back.

Bob O

Work perfectly as charm ! it could be awesome to add reactor faceswap to it :) thank's as alway's you rock it

benoit corvol

Any change of making a Vid2Vid workflow using flux?

Kythe Druid

Brilliant. The only problems I have is getting error about memory when I click run, but I just click again and it takes off! REALLY POWERFUL!

Bob O

I do not know what is wrong but when I queue a prompt and it starts loading in the models this appears. "Loading 1 new model loaded partially" Then the image is not that close to what they are supposed to look like.

Ivan

I have the same "problem" (quotes, because I don't know yet if it will cause me any troubles or not). I've located more info in the GUI. I took a screenshot: https://i.postimg.cc/HkqQdcPB/image.png I have followed your instructions carefully and initially everything seemed fine. After finishing the setup, I started to look around in the GUI and I've used the "Update All" button, with no specific reason. It actually started to update something and after that I've noticed these same errors in the terminal, as the OP mentioned (a few more, actually). I was following only the text available here, in the post. Haven't watched the video.

Wojciech Lenartowicz

could a multitude of reasons, first open a command prompt window and drag and drop the installer file inside the window, then press enter, this should avoid the window from closing and it will at least give us an error code we can use to troubleshoot the error, make also sure that you have disabled any antivirus/firewall from running and blocking the file/connection.

Aitrepreneur

Have you updated comfyui first?

Aitrepreneur

Thanks, worked great for me :)

Scott Quartochi

Autoinstall just drops 7z-installer into the folder and closes.

HL

I push F5 and or refresh button, but then it simply craps out. I get this error: Hmmm… can't reach this page 127.0.0.1 refused to connect.

JW

work like a charm ! merci beaucoup ! ;)

Eric TUALLE

I'm getting a 'Impact Pack install failed. Bad Request' Message after loading the JSON...

James Zwadlo

Ok so the fix was easy - I installed Flux1 dev Q5 GGUF. I have a NVIDIA GeForce RTX 4060 Ti with 16gb vram. So can confirm that Q8 model didn't work for me. First render was a painful 5minutes. Second render 1minute 42secs. Time to play with the rest of the toys......

Yiluwidi Dreaming

https://www.patreon.com/posts/112499790 that link takes you to that JSON file. I had the same issue.

Ian Brumley

where is the json file??

Silly Pear

nvm figured it out, its its in the Models folder under lora , just put them in there. ComfyUI>models>lora

Lloyd Braun

How do I add a Lora to the lora loader ? its not giving me the option to upload it when I click Add Lora

Lloyd Braun

None of the workflows work????? getting error... # ComfyUI Error Report ## Error Details - **Node Type:** UnetLoaderGGUF - **Exception Type:** ValueError - **Exception Message:** cannot reshape array of size 1555082 into shape (3072,3264) ## Stack Trace ``` File "D:\Aitrepreneur - COMFY FLux\ComfyUI_windows_portable\ComfyUI\execution.py", line 323, 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)

Yiluwidi Dreaming

Hi I have a 4090 and for some reason whatever workflow I use it seems to slow to a crawl at step 3 of the workflow. It literaly takes 5 mins or more to finish an image.

Restless Mind Creations

Could you make a video about how to install ComfyUI or Ollama + webui in a Docker environment for easy deployment on other computers?

Judi Godvliet

Holy crap, thank you so much for all this work! I've been wanting to use the inpainting for a while but couldn't get my head around it from other creators instructions! You're a benevolent AI god. Thank you

Sam Shennan

I had one planned but it uses so much power and it's so slow that even on my 4090 it took like 15mins for 1 image...so I took it out

Aitrepreneur

Could you make a creative upscaler like Magnific? Would be awesome :)

Judi Godvliet

1. Have you updated comfyui? 2. Which install option did you chose? If you chose anything other than 3 then you need to select the correct model and correct text encoder before generating the image

Aitrepreneur

guidance scale is how closely the image should follow the prompt, at least in theory, for the lora and nude I can't really help with that, I'm not even sure if that nude inpainting works with other models. The lora need to be toggled on and the correct trigger needs to be used in the prompt as natural language

Aitrepreneur

if there are some import failed error for some reason, just click uninstall for that particular node, then click restart, and reinstall the node again in the install missing node section

Aitrepreneur

to avoid OOM you can just activate system fallback in the nvidia settings

Aitrepreneur

like I showed in the video, just press F5 to refresh

Aitrepreneur

any errors during install?

Aitrepreneur

Soon™

Aitrepreneur

1. Was there any errors during the install 2. Have you updated comfyui like I showed in the video 3. Have you installed all the missing custom nodes?

Aitrepreneur

Just follow the instructions A. gave in the youtube video. click the reload BUTTON in your browser

Hans

i received the following error, recommendations would be helpful Prompt outputs failed validation UnetLoaderGGUF: - Value not in list: unet_name: 'flux1-dev-Q8_0.gguf' not in [] DualCLIPLoaderGGUF: - Value not in list: clip_name2: 't5-v1_1-xxl-encoder-Q8_0.gguf' not in ['clip_l.safetensors', 't5xxl_fp16.safetensors', 't5xxl_fp8_e4m3fn.safetensors'] - Value not in list: clip_name1: 'ViT-L-14-TEXT-detail-improved-hiT-GmP-TE-only-HF.safetensors' not in ['clip_l.safetensors', 't5xxl_fp16.safetensors', 't5xxl_fp8_e4m3fn.safetensors'] VAELoader: - Value not in list: vae_name: 'ae.safetensors' not in ['ae.sft', 'taesd', 'taesdxl', 'taesd3']

E

For some reason, I am getting this issue when running the GGUF versions of Dev "UnetLoaderGGUF cannot reshape array of size 137536 into shape (3072,13056)" and on GGUF Schnell, "CLIPTextEncode Allocation on device". both the Q8.0 version(option number 3 on the installer). However on the FP8 side, both seems fine on my end. Is this due to me being low on VRAM and I should go with the Q4 version instead? I did put on "prefer no sysmem fallback". My specs, RTX 4050 laptop GPU 6gbvram, 13th gen i7-13620H 64gb of ram. Funny thing is, I swear GGUF worked before when I was testing out the Schnell model. I'll try a reinstall and try using option 1.

Junayet Miah

Aitrepreneur - Sorry for this longer post but if you could read some of it, if not all :) , and help me and guide me in some small way please :) I am playing this whole afternoon with inpainting I tried different LORA's and even without LORA's and I just cannot get a single picture without any clothes on the picture. I tried literally every single option by tweaking it from 0.01 to 1.00 What exactly is guidance scale? Aitrepreneur - would it be possible for future workflow release so that each option, for example guidance, so that it has bubble once you hover over it and then it would say how this option affects on final image? Is there any proper tutorial on how to succesfully prompt my request for nudity or at least how to get it properly so that there are zero clothing on the image? I also tried putting the word "clothes" in negative prompt to exclude it from the final image but it is still there Also, how to properly inpaint mask on my image? What I do is that I paint everything except the face and then I like to use strength of somewhere around 0.6-0.7 so that it still uses correct pose from the original image but so that it has enough freedom to create something, but as I said, clothes is always present no matter if I use LORA or not. The way I use LORA is just by typing the trigger word in positive prompt then I put "," and after that I put my prompt of what I want. Am I using LORA's incorrectly? Thanks for help in advance.

Filip Šenjug

I have a solution for you, my friend had the same problem: https://git-scm.com/downloads Install this for windows of course and you will have the manager button

Filip Šenjug

I am also having this problem, and trying to manually put GGUF via install from GitHub lacks security permissions (which i've tried enabling), and installing it standalone did nothing. Not sure about how to move forward

Jason O

Unit

First of all, AWESOME! I'm the person who was flustered by those ComfyUI workflows that looked like schematics for the space shuttle so, your workflow is brilliant thank you! One problem I have is I'm running off a 3080 laptop with 16gb and I can run using the G8 GGUF, but only once. If I do a second image generation, I get an OOM error. Which I can just hit "Queue" again and it works fine, so not really a big deal. However if I right-click on the workspace after a generation, and select "cleanup of VRAM Usage (Shift+r)" I can queue again with no error. I was wondering if there is a way to include "cleanup of VRAM Usage (Shift+r)" automatically after a generation so I can run continuously. Thanks!

TechnoCore

I have a problem, when I update comfyui and restart it doesn't show me anything in the browser, literally everything is white and stays like that. I also tried to close and reopen manually but it stays the same, I tried to reinstall everything from scratch but it still happens once I update, what should I do?

StickHatzy Art

The manager button is missing. Do you have any advice? Been a subscriber for a good while now, still haven't succeeded with the 1-clicks...

Garrett deRosset

for "locking a node" I think its just right clicking and you press "pin"

Chroma

Not to put any pressure on you but will you make a tutorial on "how to make a flux lora"

Chroma

Yo Aitrep! Excellent overall contribution! Thank you so much!!! Appreciate your YouTube videos and your patreon page. 1) Training LoRAs for Flux is now happening, right? I think your original video said it wouldn't be too realistic on consumer grade hardware but I think people are doing just that, are they not? How do I do this with comfyui? Doesn't seem to be much documentation out there. 2) Is there a way to lock each group of nodes in place? I find myself accidentally dragging nodes all over the place and that would be my only complaint about comfyui. Thanks again! Really appreciate you!!!

J

Wanted to say thank you. This works like a charm. :)

Junayet Miah

question - i got the setup done ok but then when i tried to change the model in the Unet loader section by clicking on the arrow to the right - i now only see null or undefined - any ideas why? - i see this in the console - FETCH DATA from: C:\Users\David\Desktop\flux\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Manager\extension-node-map.json [DONE] [ERROR] An error occurred while retrieving information for the 'Florence2ModelLoader' node. Traceback (most recent call last): File "C:\Users\David\Desktop\flux\ComfyUI_windows_portable\ComfyUI\server.py", line 559, in get_object_info out[x] = node_info(x) ^^^^^^^^^^^^ File "C:\Users\David\Desktop\flux\ComfyUI_windows_portable\ComfyUI\server.py", line 526, in node_info info['input'] = obj_class.INPUT_TYPES() ^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\David\Desktop\flux\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Florence2\nodes.py", line 148, in INPUT_TYPES "model": ([item.name for item in Path(folder_paths.models_dir, "LLM").iterdir() if item.is_dir()], {"tooltip": "models are expected to be in Comfyui/models/LLM folder"}), ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\David\Desktop\flux\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Florence2\nodes.py", line 148, in "model": ([item.name for item in Path(folder_paths.models_dir, "LLM").iterdir() if item.is_dir()], {"tooltip": "models are expected to be in Comfyui/models/LLM folder"}), ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "pathlib.py", line 931, in iterdir FileNotFoundError: [WinError 3] The system cannot find the path specified: 'C:\\Users\\David\\Desktop\\flux\\ComfyUI_windows_portable\\ComfyUI\\models\\LLM' - i did a reinstall and noticed this - Creating Xlabs directory... The system cannot find the path specified.

David Keane

The link in the LOCAL INSTALL only works on mobile. Not PC. There is a link to the ULTIMATE ALL-IN-ONE FLUX WORKFLOW V1: which worked. However, I did not know that had the file. It would have been nice to put the file in the post directly. Now, of course you did do that but, like I said above it only works on mobile. Not sure what happened. But at any rate, I did manage to get the file and it works so, thank you for that.

Ian Brumley

Sorry sorry! I missed that LORA module is already in there! :) OMG, this is so powerful!!! Thank you so much for doing this!

Len

I've uploaded a MODELS-ULTIMATE_FLUX_INSTALL.bat file for that just in case. Put it inside the comfyui folder and then run it, altough it's easier to just start a new install from scratch in a new folder just in case

Aitrepreneur

read the post the link is there

Aitrepreneur

have you watched the video?

Aitrepreneur

It would be amazing if you can do custom LORA workflow (without image). I just finished training LORA via https://replicate.com/ostris/flux-dev-lora-trainer/train (here's a good idea for a tutorial) on my photos and would love to do some custom avatars...

Len

Where is the JSON file?

Ian Brumley

check the first comment

SparklyPepper

As far as I know the Flux people are still working on a video creation thing. The Flux image generators don't support creating videos currently. In ComfyUI you can do image2video with SVD (stable video diffusion) but it only does about 2 second clips. The SVD ones can be quite good but 2 secs is too short, and the workflow someone gave creates the vids in .webp format that loads into a browser but not normal video players. There's a separate CogStudio that can create clips from image2video that are 6 secs long but take ages to create and they don't always come out well. But I haven't tried creating many vids. It also has 2 different models option for the text2video section (and video2video) but doesn't give the model selection for the image2video option (so I don't know if it uses the smaller or bigger model. I wondered if it could use a bigger model the image2video quality might be better).

cool1

Following. Same question.

Christopher Moonlight Productions

I already have ComfyUI on my disk installed. Does it work if I install your file in the same folder? Or does it then dublicate some files and give errors

IVERSOX

For the next version of the workflow, do you know any way to have Flux create video clips from a picture like Runway does?

pgmr185

This is fantastic. I've been trying to get controlnet working for a month and I didn't have much luck.

pgmr185

I was worried you'd say 150GB+ . C drive it goes! Thanks

SparklyPepper

yeah it's pretty big, around 50GB

Aitrepreneur

How big is the final installation? I'm asking because I'm unsure if I install on my C (ssd) or D (hdd) drive. :)

SparklyPepper

I loaded the JSON on that post above, I'm now getting these errors in the UI: When loading the graph, the following node types were not found: DownloadAndLoadFlorence2Model (In group node 'workflow/Florence') workflow/FlorenceRemove from workflow Florence2Run (In group node 'workflow/Florence') Text Find and Replace (In group node 'workflow/Florence') FluxGuidance (In group node 'workflow/positive prompt guidance') workflow/positive prompt guidanceRemove from workflow workflow/Negative PromptRemove from workflow LoadFluxControlNet (In group node 'workflow/Flux Controlnet') workflow/Flux ControlnetRemove from workflow ApplyFluxControlNet (In group node 'workflow/Flux Controlnet') XlabsSampler (In group node 'workflow/XLabs Sampler') workflow/XLabs SamplerRemove from workflow ModelSamplingFlux (In group node 'workflow/Other Nodes') workflow/Other NodesRemove from workflow ImpactGaussianBlurMask (In group node 'workflow/Nodes 3') workflow/Nodes 3Remove from workflow ShowText|pysssss MiDaS-DepthMapPreprocessor CannyEdgePreprocessor HEDPreprocessor Label (rgthree) Power Lora Loader (rgthree) LoadAndResizeImage DualCLIPLoaderGGUF UnetLoaderGGUF UltimateSDUpscale Nodes that have failed to load will show as red on the graph.

John Odum

You are not using the workflow. Workflow is here: https://www.patreon.com/posts/ultimate-all-in-112499790

Aitrepreneur

No, do a new install, it's easier this way

Aitrepreneur

I'm getting this error... Failed to validate prompt for output 9: * CheckpointLoaderSimple 4: - Value not in list: ckpt_name: 'v1-5-pruned-emaonly.ckpt' not in [] Output will be ignored invalid prompt: {'type': 'prompt_outputs_failed_validation', 'message': 'Prompt outputs failed validation', 'details': '', 'extra_info': {}}

John Odum

Sounds great Should this just be run over the existing installation done with your previous install bat?

McGenie