1-Click INSTALL WAN IMAGE-TO-VIDEO & TEXT-TO-VIDEO ComfyUI WebUI!

I am also experiencing this issue and would like to know if there is a fix.

Nampot

I think the issue a few of us are having has to do with this Part of step 8. Load the IMAGE_TO_VIDEO-TEXT_TO_VIDEO-ALL.json file. I'm not exactly sure what that means. Do I upload it to the Comfyui folder in Jupyter or is there something I'm supposed to do from the manager interface after launching the GUI?

Darkwan75

Runpod: Each time I get to the Update & Restart steps (step 8), I update and it completes, but when I restart ComfyUI, nothing happens and I notice in JupyterLab update that it says "libcudnn.so.9: cannot open shared object file: No such file or directory". This happens everytime I click restart, and also when I refresh the browser it then tells me the port is not ready. Anyone experience this and know how to fix?

PW Reclaim

just use another gpu

Aitrepreneur

I am having issues getting the comfyui image. I get "create container: still fetching image aitrepreneur/comfyui:2.3.5" over and over again. It did work with a 4090 but not with the L40S.

Rekcah Keeg

I have run into some weird errors with the ui that were very hard to pinpoint and I am not sure what causes them but some other problems described here could be cause by that. Often after activating one of the bypassed parts comfyui will just completely mess up all the values in the workflow. For example putting 1024 as batch size and bf16 as model name. I am not exactly sure what causes this. Maybe because the workflow was exported by an older version, but this would not explain why it works at first most of the time. (I am using a 5090 and have installed cu 128 myself but I dont think that this is the root cause)

VinPre

check the model that you are using, without seeing what you are using difficult to say which incorrect model you are using here

Aitrepreneur

I get the following error: Prompt outputs failed validation: UNETLoader: - Value not in list: unet_name: 'wan2.1_i2v_480p_14B_bf16.safetensors' not in ['hunyuan_video_720_cfgdistill_fp8_e4m3fn.safetensors', 'hunyuan_video_FastVideo_720_fp8_e4m3fn.safetensors', 'skyreels_hunyuan_i2v_fp8_e4m3fn.safetensors', 'wan2.1_t2v_1.3B_bf16.safetensors'] Any tips how on how to fix it?

Bogdan

Jupp, it's also not working for me on my 5090

CheruPrime

I have it running but everytime I try to generate an image to video I get this error. "CUDA error: no kernel image is available for execution on the device"........ I don't know how to solve this and nothing is able to generate. Im using the 5090

jonnylpg1993

Not sure if it's related to the RTX5080, but I was having similar issues with a 5090. I know Cuda needs to be 12.8... but it's beyond me how to do that or of I'm even talking any sense here.

toony

I managed to install everything but when I try to make a video I get this error message # ComfyUI Error Report ## Error Details - **Node ID:** N/A - **Node Type:** N/A - **Exception Type:** Prompt execution failed - **Exception Message:** Prompt outputs failed validation: CLIPLoader: - Value not in list: clip_name: 'umt5_xxl_fp8_e4m3fn_scaled.safetensors' not in [] CLIPVisionLoader: - Value not in list: clip_name: 'clip_vision_h.safetensors' not in [] LoadImage: - Custom validation failed for node: image - Invalid image file: WOMAN.png UnetLoaderGGUF: - Value not in list: unet_name: 'wan2.1-i2v-14b-480p-Q8_0.gguf' not in [] VAELoader: - Value not in list: vae_name: 'wan_2.1_vae.safetensors' not in ['taesd', 'taesdxl', 'taesd3', 'taef1'] ## Stack Trace ``` PromptExecutionError@http://127.0.0.1:8188/assets/index-DIZ3S7TP.js:95645:5 queuePrompt@http://127.0.0.1:8188/assets/index-DIZ3S7TP.js:95974:13 ``` ## System Information - **ComfyUI Version:** 0.3.32 - **Arguments:** ComfyUI\main.py - **OS:** nt - **Python Version:** 3.12.10 (tags/v3.12.10:0cc8128, Apr 8 2025, 12:21:36) [MSC v.1943 64 bit (AMD64)] - **Embedded Python:** true - **PyTorch Version:** 2.7.0+cu128 ## Devices - **Name:** cuda:0 NVIDIA GeForce RTX 5080 : cudaMallocAsync - **Type:** cuda - **VRAM Total:** 17094475776 - **VRAM Free:** 15589179392 - **Torch VRAM Total:** 0 - **Torch VRAM Free:** 0

Rafi

I keep getting the error, "prompt has no output." invalid prompt: {'type': 'prompt_no_outputs', 'message': 'Prompt has no outputs', 'details': '', 'extra_info': {}} I have tried several prompts: invalid prompt: {'type': 'prompt_outputs_failed_validation', 'message': 'Prompt outputs failed validation', 'details': 'Required input is missing: text', 'extra_info': {}}

VanillaBryce

I keep getting this warning below when comfyui loads and the though a few of the HUNYUAN workflows will work none of the GGUF or WAN workflows will work because the workflowWANparam sections wont load. They're on the page but they have a RED border. I've tried updating comfyui and all nodes but I can't get those portions to not have the red border. workflowLTXPARAMTXTFP workflowLTXIMAGETOV workflowwanparam workflowskyreelslogic workflowWANIMAGEPARAM workflowLTXPARAMTEXTOVIDGGUF

Dylan Barnett

whats the command to start comfyui after stopping the pod and running it again?

Vince Celebrado

workflowskyreelsparam workflowskyreelslogic workflowWANIMAGEPARAM

Juan Romero

Even with this new update, im still missing these nodes

Juan Romero

After about five seconds, the results become unreliable, with noticeable degradation. To maintain quality, the last good frame needs to be fed into a new workflow to continue the process. "Chain proceess". There are a few tutorials on YouTube

Diego Vargas Escobar

How do you increase the length of the video output?

Aerys

been noticing this as a common error - refreshing the page usually works for me

Chris Meerdo

Running on Windows with the Nvidia version (clean install with batch file), I get an empty screen when when loading the workflow. I verified the JSON is good, it just doesn't do.. anything? Nothing to update or install since it thinks the workflow is empty. Any ideas on what I can try? Thank you!

Lewd Dewd

ive tried to do this but the missing nodes are still there, anyone have a fix?

YEs

Missing Node Types When loading the graph, the following node types were not found workflowLTXPARAMTEXTOVIDGGUF workflowLTXPARAMTXTFP workflowLTXIMAGETOV workflowskyreelsparam workflowskyreelslogic

Axel

I was able to get around it by manually installing filterpy (pip install filterpy)

Kris Raves

I'm getting this error when trying to run the autoinstall on runpod: Downloading filterpy-1.4.5.zip (177 kB) Installing build dependencies ... done Getting requirements to build wheel ... error error: subprocess-exited-with-error × Getting requirements to build wheel did not run successfully. │ exit code: 1 ╰─> [31 lines of output] Traceback (most recent call last): File "/usr/local/lib/python3.11/dist-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 353, in main() File "/usr/local/lib/python3.11/dist-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 335, in main json_out['return_val'] = hook(**hook_input['kwargs']) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.11/dist-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 118, in get_requires_for_build_wheel return hook(config_settings) ^^^^^^^^^^^^^^^^^^^^^ File "/tmp/pip-build-env-obooyyss/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 334, in get_requires_for_build_wheel return self._get_build_requires(config_settings, requirements=[]) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/tmp/pip-build-env-obooyyss/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 304, in _get_build_requires self.run_setup() File "/tmp/pip-build-env-obooyyss/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 522, in run_setup super().run_setup(setup_script=setup_script) File "/tmp/pip-build-env-obooyyss/overlay/local/lib/python3.11/dist-packages/setuptools/build_meta.py", line 320, in run_setup exec(code, locals()) File "", line 12, in File "/tmp/pip-build-env-obooyyss/overlay/local/lib/python3.11/dist-packages/setuptools/__init__.py", line 116, in setup _install_setup_requires(attrs) File "/tmp/pip-build-env-obooyyss/overlay/local/lib/python3.11/dist-packages/setuptools/__init__.py", line 87, in _install_setup_requires dist.parse_config_files(ignore_option_errors=True) File "/tmp/pip-build-env-obooyyss/overlay/local/lib/python3.11/dist-packages/setuptools/dist.py", line 730, in parse_config_files self._parse_config_files(filenames=inifiles) File "/tmp/pip-build-env-obooyyss/overlay/local/lib/python3.11/dist-packages/setuptools/dist.py", line 599, in _parse_config_files opt = self._enforce_underscore(opt, section) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/tmp/pip-build-env-obooyyss/overlay/local/lib/python3.11/dist-packages/setuptools/dist.py", line 629, in _enforce_underscore raise InvalidConfigError( setuptools.errors.InvalidConfigError: Invalid dash-separated key 'description-file' in 'metadata' (setup.cfg), please use the underscore name 'description_file' instead. [end of output] note: This error originates from a subprocess, and is likely not a problem with pip.

Kris Raves

Where can I find these workflows?

BitterPasta

i got this message while reinstalling on windows after all my exports started turning out as glitchy videos unable to compare versions for numpy = 1.17

Ashley Tan

i have an error: KSampler Sizes of tensors must match except in dimension 1. Expected size 3 but got size 5 for tensor number 1 in the list.

riffsandskulls

Look into log in /workspace/logs/comfyui... -> For me it was the missing 'propcache', I have to install it with "pip install propcache" everytime I restart my runPod.

Klaus The.

It does not make any sense if we need to reinstall it again every time we use runpod

Diego Vargas Escobar

I need the way to fix it :(

Jan Marko

same here, it shows the web terminal is running but services are never ready to use

Diego Vargas Escobar

Hi. I have error on ffmpeg ### ComfyUI Revision: 3183 [b07f116d] *DETACHED | Released on '2025-02-26' [VideoHelperSuite] - WARNING - Failed to import imageio_ffmpeg [VideoHelperSuite] - ERROR - No valid ffmpeg found. How to install this?

Łukasz Pelczar

Hey, all was working well until I stopped the pod and then when I launched it again I can’t open the port 3000, could you help me guys?

Jan Marko

Need this also

Jan Marko

i think this: TEXT-IMAGE-TO_VIDEO-MODELS_INSTALL-UPDATED.bat ... run it in the Comfyui folder

Janne Kallio

In RunPod, I use your comfyui-template and the wan-video-installer from this patreon. Everytime I restart the pod, the comfyui-logs show "ModuleNotFoundError: No module named 'propcache'" and I have to install it with "pip install propcache". After restarting comfyui it works like a charm. Just wanted to let you know.

Klaus The.

When you installed the missing nodes with the comfyui manager you have to restart comfyui and then reload the comfyui website/browser tab. After this, close the shown workflow and drag-n-drop the workflow.json-file onto the comfyui gui.

Klaus The.

How do we fix this?

Toxic

you have to look for it and re do it and check if you have the last comfyui version, not the last manager, the comfyui

Diego Vargas Escobar

Getting a missing node error, anybody know a solution to this?: "Missing Node Types When loading the graph, the following node types were not found WanImageToVideo (In group node 'workflow>WANIMAGEPARAM') workflow>WANIMAGEPARAM workflowWANIMAGEPARAM"

Imran Bukhari

I'm getting the same: "Missing Node Types When loading the graph, the following node types were not found WanImageToVideo (In group node 'workflow>WANIMAGEPARAM') workflow>WANIMAGEPARAM workflowWANIMAGEPARAM"

Imran Bukhari

can you please point me in the right direction? Got the ffmpeg issue too

Micahel DeLeo

Thank you so much for this setup. Do you know why I can't get loras to work correctly? I'm getting a lora key not loaded error. Apologies this is my first time using Comfyui. Thanks!

Jay Park

Im still waiting to move from Sora to this when i get a build and VIM method working well.

Wally The artist

Is multi gpu possible with these models to split the large vram demand?

Aurora.luxe.og

Try clearing the cache folder in users/comfyuimanager

jonescrusher1

Your windows 1 click installer has a typo in the file "TEXT-IMAGE-TO_VIDEO-COMFYUI-MANAGER_AUTO_INSTALL-UPDATED.bat" on line 286. I assume that line is supposed to be some kind of curl to get the 480p model from huggingface. Also, your install is missing the "imageio-ffmpeg" python module, so trying to make a video in comfyUI after install fails because the video can't be put together...

Pelu

I'm constantly getting this error after refreshing ComfyUI. This has completely broken the install every time and made me start fresh again ModuleNotFoundError: No module named 'propcache'

Ricky Hoffman

Seems to work: Uninstall torch by opening terminal inside the main folder (where run_nvidia_gpu.bat is located) and: python_embeded\python.exe -m pip uninstall torch torchvision torchaudio ...make sure torch, torchvision and torchaudio with all their dependencies are uninstalled. Then do a fresh install: python_embeded\python.exe -m pip install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu128 ...yes, i'm aware that nightly builds are a bad idea, but it's better that the video generation works for now than not at all. This makes it easier to wait for the stable version ;-)

J. Lehmann

Hi thank you so much for bringing Wan video Generation to us. :-) Your installation scripts work perfectly. Without any problems my cat ran through the snowy forest after only two hours and 23 minutes of computing time. The video was calculated via run_cpu.bat (AMD 9800X3D). ;-) But when I now try to switch to GPU (run_nvidia_gpu.bat), I get CLIPTextEncode and CUDA errors. How can I update your ComfyUI standalone build so that it works with Blackwell hardware with CUDA 12.8? Is it enough to update PyTorch to the latest nightly build? Any advice? Please?

J. Lehmann

I have the same thing, loads of custom nodes missing

Murylo Batista

cant find this custom node any help on how to find it WANIMAGEPARAM for image to video

Mike Contreras

Where's json file??

Suzuri Ligrness

Hi, I created Storage for this project. I have a question, what command do I start POD after removing it to use host3000? (in runpod)

Kamil Tatus

comfy UI manager not showing up on the right side - Just a tiny Queue window how do I access it in order to update all the missing models / nodes for the .Json workflow? When I load the .Json it just gives options to remove from the workflow & not update?? please assist. =(

Vato Vegas

Lora training for WAN i2v next? https://github.com/modelscope/DiffSynth-Studio/tree/main/examples/wanvideo

Thyandyr

very strange, i did that. i know it may be my noob user error. I had initially installed on my C drive, but it didnt have enough memory. so then i reinstalled in D drive. but for some reason that didnt work. so now i've combined my drive to just C. and will reinstall in C. hope this works!

Dat Nguyen

Tried to install both update all in one and only models but yet getting this error. Prompt outputs failed validation UNETLoader: - Value not in list: unet_name: 'wan2.1_i2v_480p_14B_bf16.safetensors' not in ['hunyuan_video_720_cfgdistill_fp8_e4m3fn.safetensors', 'hunyuan_video_FastVideo_720_fp8_e4m3fn.safetensors', 'skyreels_hunyuan_i2v_fp8_e4m3fn.safetensors', 'wan2.1_t2v_1.3B_bf16.safetensors']

Elrond

Its ok I resolved it, it was the pingpong option. Thanks

Admiral Gadoosh

How hard would it be to add this https://github.com/kijai/ComfyUI-WanVideoWrapper to the IMAGE_TO_VIDEO-TEXT_TO_VIDEO-ALL.json? It allows you to make longer videos! Thanks in advance.

ED-3D

try copying the requirements txt into the relevant folder (where it is looking for it)?

Michael Moser

Sorry my mistake ! I meant the text to video, tried it on both FP8 and GGUF

Admiral Gadoosh

is anyone else having issues updating comfyui? i ahve tried installing and reinstalling serveral times and everytyime i am unable to update.

Dat Nguyen

which hunyuan image to video?

Aitrepreneur

maybe yeah, try another browser

Aitrepreneur

Feddie Xtzeth

I'm just probably being a noob with all this but why do my hunyuan image to videos loop just after a few seconds regardless on length. Leaving at default 61 does it and tried increasing to a video around 10 seconds but again just loops somewhere around 3-5 seconds. I read it's capable of 8-12 seconds , is there some settings I'm not changing or is that more or less it. No complaints though great workflow thanks.

Admiral Gadoosh

My terminal on runpod is all pixelated and distorted like it can't render the font or something. Can't read shit lol I'm thinking it's my browser or something, using Librewolf. Any ideas?

Kuzo

yea, thats wrong. i "drag and drop" wahtever file you are talking about and its empty. please provide the files you state. where is the "file" you want us to drag and drop?!?! 1 click install ==== FALSE ADVERTISNG

Dat Nguyen

i have no idea what you are talking about. i cant even find where your program installed anything. can you please provide instructions? i cant find the file you're talking about. theres no instructions on anything.

Dat Nguyen

where do you "LOAD" the file from?! i dont see any directories or files

Dat Nguyen

I've run the runpod install a few times now and the install script should finish in less than 30 minutes. If not I would start a new build on a new pod. The install does fail fairly often and I have to just start over.

J Rosen

I'm scared to stop, start over and take all this time again. It's taking my dollars away.

Maurício Velasquez

I think the Cumfyui installation on Runpod is in a loop. I've been installing it since 2pm (it's already 9pm here in Brazil) and it never finishes. I've installed it other times and it didn't take that long. I think it's an error lol

Maurício Velasquez

Further details here. This error is generated by KSampler node. Also, this happens when using the GGUF WAN Image To Video workflow. The only settings I've changed otherwise is the Width, Height, Length and Config property in the WANIMAGEPARAM node. I've also set Loop to 1 on the Video Combin VHS node.

Garnett Gilchrest

if you've used the updated installer, just re drag and drop the workflow inside comfyui, all the nodes are installed automatically now

Aitrepreneur

anyone else not able to install? i installed, but the "4. The webui will launch, load the IMAGE_TO_VIDEO-TEXT_TO_VIDEO-ALL.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 then redrag and drop the workflow once again in comfyui and there you go😎" step never appeared..... what do i do?

Dat Nguyen

I get the exact same thing. Tried a fresh install two different times with the updated installer. Still the same outcome.

Derek Waters

Ok, I've run the UPDATED bat file installer for a fresh install twice now, and I've watched the troubleshooting video as well, but I'm getting an error I can't seem to resolve. I tested the runpod script, and it works just fine. but when installing locally (3090 24GB VRAM) I get this error every time I try to run the queue: "The size of tensor a (128) must match the size of tensor b (16) at non-singleton dimension 1" How can I resolve this?

Garnett Gilchrest

Can I also use that .sh install file in Vast.AI or is it specific to Runpod only?

toxrednil

I'm sorry to bother you, but is it possible to use the Blackwell GPU's with the RVC video you posted? I'm having trouble getting it to work

...

Thank you guys you're more helpfull than the creator!

Einziger

Thanks !

Einziger

Hi , just wondering one thing... I really don't feel comfy with comfyui, and am still using the stable diffusion webui... But it seems like this kind of functions won't exist for sdwebui.. Do you suggest completely moving to comfy ui ?

H-BOX

I'm trying to install onto runpod and I downloaded the new script "WAN_ONLY-TXT+IMG_AUTO_INSTALL-RUNPOD.sh" today onto the pod and ran it successfully. However, Comfyui is still showing missing custom nodes. The VAE and Ksampler is missing (they are red and undefined) and the comfyui manager is not installing the missing nodes. How can I fix manually. I have tried running script this several times on two different pods. I'm seeing this error in the comfyui.log file: [ERROR] An error occurred while retrieving information for the 'DiffusionModelLoaderKJ' node. Traceback (most recent call last): File "/workspace/ComfyUI/server.py", line 443, in get_object_info out[x] = node_info(x) ^^^^^^^^^^^^ File "/workspace/ComfyUI/server.py", line 420, in node_info info['input'] = obj_class.INPUT_TYPES() ^^^^^^^^^^^^^^^^^^^^^^^ File "/workspace/ComfyUI/custom_nodes/ComfyUI-KJNodes/nodes/model_optimization_nodes.py", line 169, in INPUT_TYPES "model_name": (folder_paths.get_filename_list("diffusion_models"), {"tooltip": "The name of the checkpoint (model) to load."}), ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/workspace/ComfyUI/folder_paths.py", line 228, in get_filename_list out = get_filename_list_(folder_name) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/workspace/ComfyUI/folder_paths.py", line 195, in get_filename_list_ folders = folder_names_and_paths[folder_name] ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^ KeyError: 'diffusion_models'

J Rosen

Nice job man, I was going to try to look online to set this up and I noticed you already did it for me lol was great timing. Now if people can come up with a good way to extend the 5 seconds length like if you set it to a 2 min video the software could just auto-save the generations in smaller videos and at the end stitch them all together and save it as one file, then we can have longer videos because 5 seconds is kinda meh lol

Danny Smithers

Click on the run_nvidia_gpu.bat file to launch ComfyUI again

Pink Floyd

So there was an extra port that has a "Launch ComfyUI" button on it that I was unaware of. All set. TYVM for your responses.

Snoogins

Create a new, separate ComfyUI folder and run the installer from inside that folder. This will allow you to have multiple installations that won't mess with each other.

ChudJones

I tried that. It keeps saying it's not connected, like ComfyUI is shut down and I can't figure out how to get it to launch again.

Snoogins

yes

Alex C

don't run the command again, just click the connect button on runpod

Alex C

What if we installed everything and are now returning to create more, what is the command to just start up ComfyUI again without having to redownload all the models (on RUNPOD)? All it does when I run the command again, it redownloads all the models, then gives me an error at Insightface because it's already there.

Snoogins

I ran TEXT-IMAGE-TO_VIDEO-COMFYUI-MANAGER_AUTO_INSTALL-UPDATED.bat within the folder: ...\ComfyUI_windows_portable\ComfyUI I also so saw that there are some WAN models in GGUF in this folder: ...\ComfyUI_windows_portable\ComfyUI\ComfyUI_windows_portable\ComfyUI\models\unet Oh God now that I see this path I realize there must have been something wrong on the location where I ran the .bat file, or something wrong I did earlier Folder \unet contains the following: fast-hunyuan-video-t2v-720p-Q8_0.gguf ltx-video-2b-v0.9-Q8_0.gguf skyreels-hunyuan-l2V-Q8_0.gguf wan2.1-i2v-14b-480p-Q8_0.gguf wan2.1-i2v-14b-720p-Q8_0.gguf wan2.1-t2v-14b-Q8_0.gguf

Alex Arangon

Bro please reply: Is it possible to install this while keeping an old version for the images? Like 2 separate systems? (I did a lot of customization on my system for image generation and I don't want to spend time to do it again, at least for now) I asked this 2 day ago and no reply. It's a simple question.

Einziger

Hello, I have several issues now. When I run GGUF WAN TEXTO TO VIDEO with Skyreels and fats.Hunyuan-video, and maybe more, I get this message: HunyuanVideo.forward() missing 1 required positional argument: 'y' And when I run Running LTX-video-2b-v0.9 I get this error: LTXVModel.forward() missing 1 required positional argument: 'attention_mask' And whn I run GGUF WAN IMAGE TO VIDEO I get this Comfyui error: Prompt outputs failed validation LoadImage: - Custom validation failed for node: image - Invalid image file: WOMAN.png UnetLoaderGGUF: - Value not in list: unet_name: 'wan2.1-i2v-14b-480p-Q8_0.gguf' not in ['fast-hunyuan-video-t2v-720p-Q5_K_S.gguf', 'ltx-video-2b-v0.9-Q5_K_S.gguf', 'skyreels-hunyuan-I2V-Q5_K_S.gguf', 'wan2.1-i2v-14b-480p-Q5_K_S.gguf', 'wan2.1-i2v-14b-720p-Q5_K_S.gguf', 'wan2.1-t2v-14b-Q5_K_S.gguf'] When I installed, I got some errors trying to download some things, maybe that was the problema, but I don´t know what to do now. I haven´t try other things yet. Please, help me. Thank you in advance.

Juan Forero

Are we talking about framerate? I doubled the framerate and it made the video even shorter??

Swarls

What updated installers? Do I need todo a clean install or can I run the comfy installer above and overwrite the current files and add the missing nodes. I can’t search for them either share using the V-3 flux installer and the updated models installer above

Martin Thoburn

woops did not see that, thanks for the reply will give it a shot

studio@brooklynpostcards.com

Redownload this installer: https://www.patreon.com/file?h=123400460&m=431239932 and try again

Aitrepreneur

Juan Forero

are all the models donwloaded correctly? Seems like the models were not downloaded, have you used the updated installer?

Aitrepreneur

I decided to only make the gguf models dl since these are the optimized models and the only models 99% of GPU could run

Aitrepreneur

use the updated installers, all missing nodes will be installed automatically, don't forget to re drag and drop the workflow inside comfyui

Aitrepreneur

yes increase the number of frames

Aitrepreneur

send me a dm

Aitrepreneur

yes, you use the TEXT-IMAGE-TO_VIDEO-MODELS_INSTALL-UPDATED.bat put that inside the comfyui folder and run it

Aitrepreneur

use the updated installers, all missing nodes will be installed automatically, don't forget to re drag and drop the workflow inside comfyui

Aitrepreneur

use the gguf side of the workflow

Aitrepreneur

just increase the number of frames

Aitrepreneur

Getting these errors when trying to run Wan: Prompt outputs failed validation CLIPLoader: - Value not in list: clip_name: 'umt5_xxl_fp8_e4m3fn_scaled.safetensors' not in ['ViT-L-14-TEXT-detail-improved-hiT-GmP-TE-only-HF.safetensors', 'clip-vit-large-patch14\\model.safetensors', 'clip_l.safetensors', 'llava_llama3_fp8_scaled.safetensors', 'longclip-L.pt', 't5xxl_fp16.safetensors', 't5xxl_fp8_e4m3fn.safetensors'] CLIPVisionLoader: - Value not in list: clip_name: 'clip_vision_h.safetensors' not in ['sigclip_vision_patch14_384.safetensors'] UNETLoader: - Value not in list: unet_name: 'wan2.1_i2v_480p_14B_bf16.safetensors' not in ['flux1-canny-dev-fp8.safetensors', 'flux1-depth-dev-fp8.safetensors', 'flux1-dev-fp8.safetensors', 'flux1-fill-dev_fp8.safetensors', 'hunyuan_video_720_cfgdistill_fp8_e4m3fn.safetensors', 'hunyuan_video_FastVideo_720_fp8_e4m3fn.safetensors'] VAELoader: - Value not in list: vae_name: 'wan_2.1_vae.safetensors' not in ['LTX_vae.safetensors', 'ae.safetensors', 'ae.sft', 'hunyuan_video_vae_bf16.safetensors', 'taesd', 'taesdxl', 'taesd3', 'taef1']

Alex Arangon

In the FP8 WAN image2video node, i can't seem to load a WAN image2vid model. In the GGUF node I can load the GGUF model, but I can't see a model for the FP8 WAN node. Anyone else having this problem?

fjui

I've gotten most of the bat file to correctly load the models and dependencies for comfy portable, however, on trying to use the workflow I'm getting a missing node error that comfy manager can't resolve: "When loading the graph, the following node types were not found workflowskyreelslogic" Any help would be greatly appreciated.

John Lee

Successfully created a short video (2s), are we able to create longer than that?

Swarls

Oh I found the info, it was there. Sorry :)

Alex Arangon

Alex Arangon

Install ffmpeg with pip, then go to the lib>site-packages of your python, copy ffmpeg libraries in there and then go to the folders where ComfyUi is should be another folder call python_embeded go there go into lib and then site-packages again and paste the libraries there, you may need to do the same with imageio-ffmpeg

lolno

I am trying the fresh install and get thet folllowing. [INFO] Winget installed successfully. [INFO] Installing Microsoft.VCRedist.2015+.x64... 'winget' is not recognized as an internal or external command, operable program or batch file. [INFO] Installing Microsoft.VCRedist.2015+.x86... 'winget' is not recognized as an internal or external command, operable program or batch file. Switched to another PC all seems to install, but the manager does not show. Will review your other help video. Thanks for all your work!

Scott Quartochi

I am getting a lot of failed downloads, and then when I try and run the installer it skips them and says the models are already installed. Is there a way to force redownloads of the models for this installer file?

studio@brooklynpostcards.com

I'm getting stuck after opening the json file. I click on the "Install missing custom nodes" and it just spins and never displays anything. I have rebuild twice on runpod with same issue.

J Rosen

I completed all the runpod steps, but the Wan2.1_i2v model files are missing from the ComfyUI directory. I something must have gone wrong during the installation phase. Please advise.

Priya Patel

This worked of rhte WAN only workflow but still missing many nodes in the Video-All Workflow

Martin Thoburn

What updated installer? Like a clean install? My comfyUI no longer finds missing nodes. I used the video models installer, as I already had ComfyUi running using the V3 Ultimate Flux installer.

Martin Thoburn

hello guys, i've problem only with img2 video fp8. i loaded image and wrote the prompt, but after the generation gave me another image.

Fuzzyvega

How do I fix FFMPEG errors? I installed ffmpeg and also put the exe in the comfui portable folder

Swarls

Yes, would appreciate knowing this as well. I've seen it in other workflows but its missing here.

Priya Patel

What is the option to increase the duration in I2V?

Saul Mendoza Martinez

It worked, thank you!

Zafira LPSTV

Click on the fit to view icon on the bottom right

Aitrepreneur

Someone also have problem with ComfyUI that doesn't load workflow ? i have empty page.

Zafira LPSTV

this was tried (and re-tried) in a cache-less incognito browser instance. (Nothing to do with browser cache)

Derek Waters

Thank you for your valuable help! I believe that wan2.1_i2v_480p_14B_fp8_e4m3fn.safetensors will do the job. Once again, a big thank you for your support, which helped me find the right solution. Take care and see you soon! 😊

3dAA

Absolutely not no, this is the full model but you're the one who asked for it

Aitrepreneur

I noticed that it has 30Gb15. Is that normal for a model to use with a 3060/12Gb?

3dAA

Here: https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/resolve/main/split_files/diffusion_models/wan2.1_i2v_480p_14B_bf16.safetensors?download=true

Aitrepreneur

I need some help! If anyone has a direct link to download the "wan2.1_I2V_480P_14B_bf16.saftensors" model, I’d really appreciate it. I can’t find it on Hugging Face or similar sites. Thanks in advance!

3dAA

You just need to re drag and drop the workflow inside comfyui again, it's a comfyui display bug that stays in your cache

Aitrepreneur

Same. I tried the updated installers from scratch and it still misses the WANIMAGEPARAM and workflowskyreelslogic pieces. I tried "Update All" afterwords. Still missing. Tried the troubleshooting video, and it doesn't mention this other than the "update all" (which I already tried). That's too bad, looked cool.

Derek Waters

try the new updated installer, it will install all missing nodes now

Aitrepreneur

yes that worked, Thanks!

Tonic69

watch the troubleshooting video I linked in the post, it's always the same thing when loading a new workflow. ALso I just updated the installers so you can try the new ones, they will install all nodes automatically now

Aitrepreneur

Everything works out of the box! Amazing work

ZZZ0mbieSSS

No go unfortunately, Getting ton of missing node errors. :( Did another Fresh install with this script and when it finally loaded was greeted with a huge list of missing node errors. Trying to run Install missing nodes, Update All doesnt resovle the issue. I can already tell why this didnt happen to you. In the video tutorial where you followed the procedure... you kinda didn't, looking at your command window, it was telling you a few times the certain thing is already installed. Meaning it was run on an already existing installation, not only that but it skipped download/updating stuff because you already have it installed. I would recommend, going to blank folder, popping the BAT file in there and try to run it. See if it can download install everything from scratch and run fine. Didn't for me unfortunately. Really sad, looks like a awesome worflow, unfortunately an absolute nightmare to get up and running. I'm sitting on day 10 of failing to get this installed lol. My internet traffic for those 10 days are insane, im over 2 terrabytes haha. But thats additionally youtube and other stuff ontop of redownloading the entire Comfy/models etc. Everytime I try a fresh install

Spoff Ninja

The webUI didn't launch after I ran the install. Where can I find it to open it?

Pink Floyd

I get this error and can't seem to find a fix (have updated everything)" workflowWANIMAGEPARAM

Bob Winberry

What LoRa can we use? Or does it work with any LoRa's?

Zoran Kanti-Paul

After successfully running it on runpod, I went back the next day to start a pod on my network volume. It tells me this port is not open when I click HTTP 3000. Do I have to run the auto install every time? When I look in the comfyui log it says ModuleNotFoundError: No module named 'propcache'

Dylan

Was having workflow>WANIMAGEPARAM missing missing node #166 Solved by the following. Your results may vary, but posting here to help. As Kubilay göze commented 1. Make sure you updated ComfyUI to the latest version (right now 0.3.18) and drag the workflow again. 2. If you updated from Manager and you're still getting the error; MOST LIKELY your update didn't happen and you didn't realize. Go to the "ComfyUI_windows_portable\update" folder and click "update_comfyui.bat" and if it still doesn't work click "update_comfyui_and_python_dependencies.bat" Go to Settings-About and check if the version number is at least 0.3.18 Additional steps - Update Wan custom nodes 1. Open ComfyUI Manager 2. Open Custom Nodes Manager 3. Search "Wanvideo" 4. Install Comfyui_patches_ii 5. Restart ComfyUI

Uncc234

below the height and wight in comfy

Janne Kallio

Is it possible to install this while keeping an old version for the images? Like 2 separate systems? (I did a lot of customization on my system for image generation and I don't want to spend time to do it again, at least for now)

Einziger

After this fresh installation, again. Are we supposed to do an Update All or Update ComfyUI?! Thx!

James Bowler

In the video after V2 he said to do Update ComfyUI, not Update All, because on V2 to V3, when I hit Update all it broke the Facepoke feature so I had to reinstall and just do Update ComfyUI. And at least I had a working version of Face Poke. But this time, when I just updated and downloaded the video models, I tried Update ComfyUI again and that did not work. It blew up my V3. So in the end, I just did another fresh download, are we supposed to do an Update All? Or Update ComfyUI?

James Bowler

Is there a setting to adjust the video length? I feel like there should be, but I can't find it. Specifically talking about image-to-video conversion. Thanks for the help!

Reapax

Anybody getting "workflow>WANIMAGEPARAM loading error" : 1. Make sure you updated ComfyUI to the latest version (right now 0.3.18) and drag the workflow again. 2. If you updated from Manager and you're still getting the error; MOST LIKELY your update didn't happen and you didn't realize. Go to the "ComfyUI_windows_portable\update" folder and click "update_comfyui.bat" and if it still doesn't work click "update_comfyui_and_python_dependencies.bat" Go to Settings-About and check if the version number is at least 0.3.18

Kubilay göze

sounds like you are getting the same Missing Node Types error that I am

Darkshepherd

I am facing this problem I dont see any .venv on Comfyui and when i install the newer vrsion of huggingface hub it stills giving the same error :( ImportError: huggingface-hub>=0.26.0,<1.0 is required for a normal functioning of this module, but found huggingface-hub==0.25.2. Try: `pip install transformers -U` or `pip install -e '.[dev]'` if you're working with git main

Airsoft Fow

yep same error on 480 one and the 720p version gives me the error "cannot mmap an empty file"

Tonic69

Works now. I used the models bat file. I had to execute the bat file multiple times to keep downloading models. After it wouldn't download anymore. it works now.

Ian Brumley

What package from are the workflowskyreelsparam and workflowskyreelslogic

Thyandyr

K should add this to the Workflow 3.5 ore something xD

Glen

Same issue and this did the trick. Thank you!

Slap Dash Dolt

I seem to get nothing but static with text to video. I've tried with different models. Local and on Runpod. Same outcome. Any ideas?

Bruce

I did do that. but I'll try the new one you just dropped maybe it will work better

Ian Brumley

You actually need to have the newest comfyui install, you can update it by going inside the update folder, recently comfyui has an issue updating from the manager but using the update bat file works

Aitrepreneur

everything should be downloaded, make sure that you chose yes to download the wan models in the beginning

Aitrepreneur

watch my troubleshooting video as I explained (the link is in the post) if you don't know what to do for the first time. Also these missing nodes are for the ultimate V3 workflow, in that case you can just use the newest updated V3 installer

Aitrepreneur

I updated the V3 installer, it should work fine now

Aitrepreneur

UPDATE: Fixed this by clicking "update all" in the manager and when it finished in console, i restarted from the menu.

Moses Son

"Unexpected architecture type in GGUF file, expected one of flux, sd1, sdxl, t5encoder but got 'wan'"

Moses Son

Maybe there was a issue during the model install and it corrupted it? TRY deleting it and use the models bat file installer (put it inside the comfyui folder then run it) to dl the missing models

Aitrepreneur

Unfortunately the code is not available for wan just yet, maybe in a few days. Only hunyuan and cogvideo are supported for now.

Aitrepreneur

Theres actually a new AI just released which is designed to extend existing video. Check AI Search's latest video, he covered it. Its looks pretty good. Gonna give that a go after I get this up and running. It's already got a Comfy workflow and everything. Can also be finetuned for better results, but they say the out of box model should be able to do practically everything. But finetuning on certain concepts/characters would make better results.

Spoff Ninja

I have the exact same error

Pete

is there a setting to make the video longer ? like 10 or 20 seconds long?

Jaulustus

you need to watch the older vids to make sure you config'd correctly, but basically dropping the json into the interface will implement each nodes config

J

The error occurs in \ComfyUI\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-GGUF\loader.py, line 52, within the gguf_sd_loader function. Tried to update in Manager but no luck. opened loader.py then went to https://github.com/city96/ComfyUI-GGUF/blob/main/loader.py copied file and pasted into loader.py and saved. restarted comfyui and works like a charm

john doe

send me a dm

Aitrepreneur

Just click update all in the manager then re drag and drop the workflow inside comfyui again

Aitrepreneur

maybe using the full model instead of the gguf but even then 5090 might not be enough actually

Aitrepreneur

? It doesn't matter, it will just download all the best models depending on the installer you use, 200g is indeed to much but it doesn't matter what you put as long as you have enough spaces

Aitrepreneur

Yes, as I said you can use this installer: https://www.patreon.com/file?h=123400460&m=429645609 put it inside the ComfyUI_windows_portable\ComfyUI folder and then run it

Aitrepreneur

Update - I had ComfyUI already installed, so just loaded the models. The issue, the version I had was 0.3.13 and the wan option was only added, as I understand, after 0.3.15. I tried to update comfyui but it would not move past 0.3.13 even when trying to switch via the Manager panel (If there is a way to force this, it would be of great help to know. An absolute fresh install using the TEXT-IMAGE-TO_VIDEO-COMFYUI-MANAGER_AUTO_INSTALL.bat in a new location seems to have fixed the issue as I now see the ComfyUI version as 0.3.18 and the 'wan' option is there within the Load Clip node. Also, no node errors after running 'Install Missing Nodes' so essentially it works well..but I'd rather be able to do this on my old installation.

Walker4k

tried that. did not work unfortunately

Ian Brumley

Everything I tested so far works fine except the model "wan2.1-i2v-14b-480p-Q5_K_S.gguf" I did not change your workflow, uploaded an image and always got the following Error: "UnetLoaderGGUF cannot reshape array of size 5369813 into shape (5120,3520)"

Adaptalabor

Try going into Manager and choose "update ComfyUI"., then reimport the workflow

Greg

Try updating the node itself by going into Manager, Custom Nodes Manager, and find ComfyUI-Manager and select "Try Update" to update it to the latest nightly build.

Greg

Ok, I found out that this is because I don't have the models installed. I thought those were supposed to come with the installer. But at any rate, I installed them but it is still saying the sd1_tokenizer is not found, Permission Denied #13

Ian Brumley

Any advice regards the following error - Prompt outputs failed validation CLIPLoader: - Value not in list: type: 'wan' not in ['stable_diffusion', 'stable_cascade', 'sd3', 'stable_audio', 'mochi', 'ltxv', 'pixart', 'cosmos']

Walker4k

Doesn't work. I get the following error: Prompt outputs failed validation CLIPLoader: - Value not in list: clip_name: 'umt5_xxl_fp8_e4m3fn_scaled.safetensors' not in [] UNETLoader: - Value not in list: unet_name: 'wan2.1_t2v_1.3B_bf16.safetensors' not in [] VAELoader: - Value not in list: vae_name: 'wan_2.1_vae.safetensors' not in ['taesd', 'taesdxl', 'taesd3', 'taef1']

Ian Brumley

https://github.com/comfyanonymous/ComfyUI/discussions/6643 50XX cards have to install a different version of ComfyUI (for now?) Not too educated on what's going on, but i know this works. I've got a 5080

Junbi

All the video models in one place. This is amazing! Thank you

ZZZ0mbieSSS

Thanks! This fixes the dreaded missing workflowreelslogic error. Did not occur to me to reimport the workflow itself, restarting and reloading did not fix.

Lone Striker

Update ComfyUI in the Manager. It requires the newest one to work with WAN, then reimport the workflow.

Greg

Umm what is this BS? Followed everything in your video. Came here and became a subscriber. Setup runpod and can see comfyui. NO workflow there. So I guessed that needed to be uploaded to the /default/workflow folder in Jupiter. OK so did that and refreshed the comfyui..The worfkflow is a mass of red boxes with nodes that cannot be found? What am I doing wrong? LTXAttnOverride (In group node 'workflow>LTXPARAMTEXTOVIDGGUF') workflow>LTXPARAMTEXTOVIDGGUF ModifyLTXModel (In group node 'workflow>LTXPARAMTEXTOVIDGGUF') LTXPerturbedAttention (In group node 'workflow>LTXPARAMTEXTOVIDGGUF') workflow>LTXPARAMTXTFP workflow>LTXIMAGETOV INTConstant (In group node 'workflow>skyreelsparam') workflow>skyreelsparam FloatConstant (In group node 'workflow>skyreelsparam') Int-🔬 (In group node 'workflow>skyreelslogic') workflow>skyreelslogic MathExpression|pysssss (In group node 'workflow>skyreelslogic') GetImageSizeAndCount (In group node 'workflow>skyreelslogic') Compare-🔬 (In group node 'workflow>skyreelslogic') If ANY return A else B-🔬 (In group node 'workflow>skyreelslogic') ImageResizeKJ (In group node 'workflow>skyreelslogic') Power Lora Loader (rgthree) workflowLTXPARAMTEXTOVIDGGUF workflowLTXPARAMTXTFP UnetLoaderGGUF VHS_SelectFilename VHS_VideoCombine workflowLTXIMAGETOV VHS_LoadVideoPath workflowskyreelsparam workflowskyreelslogic DualCLIPLoaderGGUF Label (rgthree) RIFE VFI VHS_LoadVideo Fast Groups Bypasser (rgthree)

Zoran Kanti-Paul

this is great. it would be great to get a video on how to train loras for these video models

latentpunk

Running on a new 5070TI got this error - CUDA error: no kernel image is available for execution on the device CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Tim Mandahl

same thing on my end with also the workflowreelslogic missing.

Verratanectu

Thanks for this release mate! looks very promising. I am having an issue at the moment with the GGUF WAN Image to Video. 1- Missing Node Types When loading the graph, the following node types were not found workflowskyreelslogic workflowWANIMAGEPARAM When I try to queue up I got this Cannot execute because a node is missing the class_type property.: Node ID '#166' As usual thanks in advance!

Verratanectu

I managed to brute force through this by repeatedly loading the workflow until the nodes finally showed up, even though the error message still shows up. You may have to load the Ultimate Flux All In One first too.

Samuel

Tried to reinstall 3 different ways Missing Node Types When loading the graph, the following node types were not found INTConstant (In group node 'workflow>skyreelsparam') workflow>skyreelsparam FloatConstant (In group node 'workflow>skyreelsparam') Int-🔬 (In group node 'workflow>skyreelslogic') workflow>skyreelslogic MathExpression|pysssss (In group node 'workflow>skyreelslogic') GetImageSizeAndCount (In group node 'workflow>skyreelslogic') Compare-🔬 (In group node 'workflow>skyreelslogic') If ANY return A else B-🔬 (In group node 'workflow>skyreelslogic') ImageResizeKJ (In group node 'workflow>skyreelslogic') Power Lora Loader (rgthree) UnetLoaderGGUF VHS_SelectFilename VHS_VideoCombine VHS_LoadVideoPath workflowskyreelsparam workflowskyreelslogic DualCLIPLoaderGGUF Label (rgthree) RIFE VFI VHS_LoadVideo Fast Groups Bypasser (rgthree)

jesus bernal

check a comment from Dani in the last V3 tutorial, that how you fix that problem, for some reason the bat file doesnt install all the models.

Ronald Solis

Getting "Missing Node Types When loading the graph, the following node types were not found". message. Already updated comfi and missing nodes. It list many nodes this is just the top of the list "DownloadAndLoadFlorence2Model (In group node 'workflow>Florence') workflow>Florence Florence2Run (In group node 'workflow>Florence') LoadFluxControlNet (In group node 'workflow>Flux Controlnet') workflow>Flux Controlnet ApplyFluxControlNet (In group node 'workflow>Flux Controlnet') XlabsSampler (In group node 'workflow>XLabs Sampler') workflow>XLabs Sampler ReActorFaceBoost (In group node 'workflow>REACTORFACESWAP') workflow>REACTORFACESWAP ReActorFaceSwap (In group node 'workflow>REACTORFACESWAP')". Any ideas anyone?

Aurora.luxe.og

I got error with Cannot execute because a node is missing the class_type property.: Node ID '#18' after installing all of it then using the .json template on comfyui

Derppy

When I've run it on V3 then I get this error: " (In group node 'workflow>skyreelslogic') workflow>skyreelslogic Compare-🔬 (In group node 'workflow>skyreelslogic') If ANY return A else B-🔬 (In group node 'workflow>skyreelslogic') RIFE VFI".

Michał

Anything to help streamline and take advantage of a 4090 or soon a 5090 with a bit more vram?

GenTsoCkn

Missing Node Types When loading the graph, the following node types were not found WanImageToVideo (In group node 'workflow>WANIMAGEPARAM') workflow>WANIMAGEPARAM

Stephen Murray

I am getting this error

Stephen Murray

You said to make a runpod with 200GB on disk and on workspace. It seems quite wasteful, shouldn't just having 200GB on workspace and optimizing how it download be better ?

FlameVeritas

Ok i was having trouble with the ultimate workflow, i started from scratch like suggested in this post, is there a bat file to just install all the models i need for ultimate v3? i got this one installed and afraid if i run the one from feb 2 it will give me old version

Tha WOLF

GOAT!

Alex Bejko

If you used the V3 installer then yes

Aitrepreneur

Can I install this on an existing pod that already has the V3 ultimate workflow?

Daniel Kutz

You seem to be releasing the good stuff all at once right when I don't have access to my computer XD. Anyway, keep up the great work.

VinPre