Update comfyui and WanVideoWrapper!
VAE (models/vae):
https://huggingface.co/Kijai/WanVideo_comfy/blob/main/Wan2_1_VAE_fp32.safetensors
Model (models/diffusion_models):
https://huggingface.co/Kijai/WanVideo_comfy/blob/main/Wan2_1-I2V-14B-720P_fp8_e4m3fn.safetensors
Text Encoder (models/text_encoders):
Clip Vision (models/clip_vision):
Upscaler Model (models/upscale_models):
https://openmodeldb.info/models/4x-ClearRealityV1
https://openmodeldb.info/models/1x-SkinContrast-High-SuperUltraCompact
Description
i2v Workflow
FAQ
Comments (23)
Hi, thank you so much for this workflow. This is correct for me. But I have a question?. 驴How do I should generate a video from IMG without losing original face composition?
higher steps, higher resolution and slower movement (in the prompt) can help
Can you please explain why
Pony Realism 馃敭
Checkpoint
馃寪 v2.2 Main + VAE
is included in your WAN Video examples for Image2video?
because the image was made with PonyRealism and then ImageToVideo with Wan
Getting "expected str, bytes or os.PathLike object, not NoneType"
any thoughts? Comfy UI interface.
fixed it... :)
WanVideoModelLoader
'patch_embedding.weight' having problem with this can you tell me how to fix or whats wrong
maybe you have the Lora nodes enabled but no correct lora selected?
Hey man, love this workflow and I have a lot of fun with it. A question: is there an easy way to add LORAs to it? I'm a complete noob to ComfyUI and I have no idea of how to do this correctly.
nodes it tried to install destroyed my comfyui setup/w sageatten,
register_finder(pkgutil.ImpImporter, find_on_path)
^^^^^^^^^^^^^^^^^^^
AttributeError: module 'pkgutil' has no attribute 'ImpImporter'. Did you mean: 'zipimporter'?
This fixed it
python -m ensurepip --upgrade
python -m pip install --upgrade setuptools
python -m pip install --upgrade pip
An error pops up, please tell me how to fix it?Thanks.An error pops up, please tell me how to fix it?Thanks.
WanVideoImageClipEncode
module 'tensorflow' has no attribute 'Tensor'
# ComfyUI Error Report
## Error Details
- Node ID: 184
- Node Type: WanVideoImageClipEncode
- Exception Type: AttributeError
- Exception Message: module 'tensorflow' has no attribute 'Tensor'
## Stack Trace
It's giving me out of memory error, what to do?
You can try:
- set blockswap to 40
- reduce resolution
- reduce video length (num of frames)
Thanks for your workflow letting me generate my first animation.I'm a newbie and I've been studying for over 48 hours for this video, which is a good starting point for me.Thanks again
Mine has been stuck on 80% for the past hour and a half, is that normal?
Might have something to do with out of memory? Not quite sure
same for me I am constantly running into CUDA OOM errors with this workflow at around 70-80%
Probably impossible to run on something like a RTX4070 Super... need way more VRAM I guess.
Is there a way to switch to a GGUF Model with this workflow for less VRAM usage?
Even the nodes for less VRAM usage arent helping.
I'm not getting any error, it simply runs for 0.01 s and it doesn't generate anything. Could someone help pls?
hi i get this error:
When loading the graph, the following node types were not found:
JWStringConcat
Fast Groups Bypasser (rgthree)
WAS Node Suite
Author WASasquatch
Seems to be a problem with Mixlab nodes
Can anyone help please?
and when i close this window and try to run than i get this:
Cannot execute because a node is missing the class_type property.: Node ID '#50'