Instructions to use lightx2v/Wan2.2-Distill-Models with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use lightx2v/Wan2.2-Distill-Models with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("lightx2v/Wan2.2-Distill-Models", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Diffusion Single File
How to use lightx2v/Wan2.2-Distill-Models 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
Question
agree, even 8 steps is not that fast,almost same as wan. 4 steps will be a killer
LTX itself is faster than Wan, even at 8 steps. But I’d like to see a 4-step version.
wan 480p 1.5minuates, while LTX 1-2minuates
Yes… You said it. LTX takes 8 steps, while WAN with Lightning LoRA only takes 4 also doesn't include audio in the output. It's clear that LTX is faster and includes audio in its generations.
Yes… You said it. LTX takes 8 steps, while WAN with Lightning LoRA only takes 4 also doesn't include audio in the output. It's clear that LTX is faster and includes audio in its generations.
sorry to be off topic but do you wondered what's the best speed lora to use in wan animate for locking the fidelity of the ref image character? does wan animate need high and low nise the lora select multi node or not?