Instructions to use lightx2v/Wan2.2-Distill-Loras with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use lightx2v/Wan2.2-Distill-Loras with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline from diffusers.utils import load_image, export_to_video # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Wan-AI/Wan2.2-I2V-A14B", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("lightx2v/Wan2.2-Distill-Loras") prompt = "A man with short gray hair plays a red electric guitar." input_image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/guitar-man.png") image = pipe(image=input_image, prompt=prompt).frames[0] export_to_video(output, "output.mp4") - Diffusion Single File
How to use lightx2v/Wan2.2-Distill-Loras 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
- Inference
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
Distill 1030 lora??
is there any plan to release the 1030 lora??
It's available on Civitai.
Yeah, I know , but that LoRA was extracted by Kijai (if I’m not mistaken), not the LightX2V team.
In my tests, it can cause color shifts and flashes of light in some generations and It has some minor issues when used alongside the distilled model (Yeah the distilled model and the distilled LoRA can work together well and surprisingly can improve motion and prompt adherence slightly). I’m looking for the 1030 distilled LoRA from the original team, the latest one is still 1022, which is perfect and has none of those issues when used alongside the distilled model.
Could you explain the differences between “Wan2.2-Distill-Loras” and “Wan2.2-Lightning-Loras”?
It's available on Civitai.
Yeah, I know , but that LoRA was extracted by Kijai (if I’m not mistaken), not the LightX2V team.
In my tests, it can cause color shifts and flashes of light in some generations and It has some minor issues when used alongside the distilled model (Yeah the distilled model and the distilled LoRA can work together well and surprisingly can improve motion and prompt adherence slightly). I’m looking for the 1030 distilled LoRA from the original team, the latest one is still 1022, which is perfect and has none of those issues when used alongside the distilled model.
for you the best lightx2v for wan animate to preserve face identity what's the one for you? and coupled with what (scheduler, sampler etc) if you wanna connect with me on telegram im doing stress test these few weeks so @jinbeers