Instructions to use Kijai/WanVideo_comfy_fp8_scaled with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusion Single File
How to use Kijai/WanVideo_comfy_fp8_scaled with Diffusion Single File:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
LoRA loading errors with Wan2_2-Animate-14B_fp8_scaled_e4m3fn_KJ_v2.safetensors
I'm getting some console errors with the lightx2v_i2v and wananimate_relight loras that come with the native comfyui wan 2.2 animate workflow:
The lightx2v_i2v generates this error when loading:
unet missing: ['face_encoder.out_proj.scale_weight']
unet unexpected: ['img_emb.proj.0.scale_weight', 'img_emb.proj.4.scale_weight']
And the wananimate_relight generates this error:lora key not loaded (followed by many missing keys)
Are these errors to be expected, or are there better loras for this particular model and quantization? will the loras still work?
The latest v2 seems to have fixed the unet errors on lightx2v_i2v.
I'm still getting the lora keys not loaded on wananimate_relight.
The latest v2 seems to have fixed the unet errors on lightx2v_i2v.
I'm still getting the lora keys not loaded on wananimate_relight.
That fixed it, no errors. Thanks!
That fixed it, no errors. Thanks!
Ok, good to know, thanks, fixed the loras here for native now too.