Version 1.1
This update is a complete overhaul of the workflow, focused on clarity, performance, and output quality.
The entire UI and node layout have been redesigned to significantly improve readability and ease of use. Navigation is now more intuitive, making the workflow much more comfortable to work with, whether you're a beginner or an advanced user.
The core structure of the workflow has been rebuilt from the ground up. This new version takes inspiration from several highly efficient workflows, especially in areas like upscaling and FPS enhancement, resulting in a more optimized and streamlined pipeline.
Major improvements have been made to:
Upscaling quality and consistency
Frame interpolation and FPS boosting
Generation speed and overall performance
Additionally, advanced techniques adapted from other proven workflows have been integrated to:
Accelerate image and video generation
Maximize visual quality
Improve coherence and stability across frames, ensuring the output stays faithful to the input image throughout the entire video
This version is faster, cleaner, and produces more consistent high-quality results across the board.
Use the WanVideo model and LoRA technology. The perfect workflow to get fast NSFW video whatever the style may be (anime or realistic) or the action.
Checkpoints :
WAN 2.2 Enhanced NSFW | SVI | camera prompt adherence (Lightning Edition) I2V and T2V fp8 GGUF :
Version 1.0
Ressource needed (Be sur to download the high model and the low model) :
Checkpoints :
Wan Video 2.2
i2v_low_noise_14B_fp8_scd
i2v_high_noise_14B_fp8_scd
https://civarchive.com/models/1817671/wan-video-22
Loras :
WAN DR34ML4Y - All-In-One NSFW
https://civarchive.com/models/1811313?modelVersionId=2303113
(wan 2.2 experimental) WAN General NSFW model
https://civarchive.com/models/1307155?modelVersionId=2073605
Wan 2.2 Lightning LoRAs
https://huggingface.co/lightx2v/Wan2.2-Official-Models/tree/main
Description
FAQ
Comments (11)
where might i find the fastmove V2 H & L .safetensors? Google gives me no help here.
@Mittoshura Thank you so much. I checked out that link earlier, but this time I took a closer look.
Can you tell me how to make a loop?
Thank you for the workflow, though I am having trouble getting it to work. I have followed all the steps but when I get to Lora, I add in the 2 I am using, click run but get an error due to the WanVideo Lora Select. - do I just set these to bypass? not sure what they do sorry.
edit - I bypassed the node and got another error saying that Low Mem Load does not work, disable Merge Loras. is that normal.
Also does this workflow save the last image?
Sorry for the hassle. Thank you again.
You’re running into two separate issues:
Don’t bypass the WanVideo LoRA Select node — it’s required for proper LoRA handling in the workflow.
The error you’re getting is because:
You’re using Low Mem Load
AND Merge LoRAs at the same time
These are incompatible in WAN 2.2.
Try this to fix it:
Either disable “Merge LoRAs” (recommended)
Or disable Low Mem Load if you have enough VRAM
@Mittoshura Thank you for your reply, I have tried that but I am still getting the error:
9fe8fc80-e1ca-40e4-b23c-1e3eef852c99.safetensors missing within the single Lora window (the one next to the multi select. I have two loras within them, the Lightx2v and a M4crom4sti4. Is it because
9fe8fc80-e1ca-40e4-b23c-1e3eef852c99.safetensors refers to the experimental lora?
Thank you again,
@Dracken1986 Disable these 2, they are just internal loras i tried you don't need them make it work, and try to use the new version of the workflow
@Mittoshura For god sake >_<, sorry no idea how I managed to download the older version, must have clicked on 1.0 by mistake. Thank you, will try it out now :D (done this a few times now with other things and always ends up to be the same reason, really need to check what I am doing more XD).
The wf is great and all,ty for sharing it!
But i can't generate any extended videos no matter what i did. I bypassed them,enabled them,selected different loras,selected many seeds but stil nothing
The required models and nodes have been set up, and the video can be generated now. However, the generated video is blurry and has a lot of snow-like noise. Could you please explain why?