I'm offering you a first version of my workflow adapted to WAN2.2. There's still a lot to improve, so don't hesitate to give me feedback.
Resources you need:
馃搨Files :
For base version
T2V Model : wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors and wan2.2_t2v_low_noise_14B_fp8_scaled.safetensors
In models/diffusion_models
For GGUF version
T2V Quant Model : wan2.2_t2v_high_noise_14B_QX.gguf and wan2.2_t2v_low_noise_14B_QX.gguf
In models/unet
For lightx2V version
lightx2V LoRA : Wan21_T2V_14B_lightx2v_cfg_step_distill_lora_rank64.safetensors
CLIP: umt5_xxl_fp8_e4m3fn_scaled.safetensors
in models/clip
VAE: wan_2.1_vae.safetensors
in models/vae
ANY upscale model:
Realistic : RealESRGAN_x4plus.pth
Anime : RealESRGAN_x4plus_anime_6B.pth
in models/upscale_models
馃摝Custom Nodes :
Description
base version
FAQ
Comments (8)
Thank you for your workflow, it's very useful
Great!. Do you think its possible add something like CUT the last 4 frames ? Like SwarmUI have this option. As when generating with Lightx2v lora few first or last ? frames are less saturation
Did you select the i2v Lightx2v loras on purpose or was it a mistake? Maybe it doesn't matter that much. :D
There's also the issue of slow motion generations. The way I solved this personally was by adding a couple of normal steps (>1 cfg) in the first sampler. Then the Lightx2v lora can properly be removed from the first pass. But ofcause then it's even more important to keep the output at 16, 32 or 64 fps. I see you have selected 24 fps, I believe that's only for the 5b model.
Sorry for being such a nag. :D Have a great weekend.
Awesome workflow as always. :)
Be nice if we could get a preview so we could cancel early if not working out.
Yes this! Just like the previous versions
hithatwhitham895聽I would really like to do this but WAN2.2 uses 2 model files which prevents me from doing my preview currently