aigc-apps / aigc-apps/VideoX-Fun
WanTransformer3DModel from diffusers 0.31 not yet support GGUF, but official diffusers from 0.32, 0.33, has started to support to load GGUF
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 2.3k
- Forks
- 188
- Avg merge
- 2d 22h
- Merged PRs (30d)
- 3
Description
I've just tested to load GGUF by new diffusers 0.33, it works. But of course, diffusers' version not support multi-GPU yet.
from huggingface_hub import hf_hub_download
path = hf_hub_download(
repo_id="city96/Wan2.1-I2V-14B-480P-gguf", filename="wan2.1-i2v-14b-480p-Q8_0.gguf" # or 720P file
)
transformer = WanTransformer3DModel.from_single_file(
path,
quantization_config=GGUFQuantizationConfig(compute_dtype=torch.bfloat16),
torch_dtype=torch.bfloat16,
)
But this still not working with VideoX-Fun's WanTransformer3DModel version with latest code I just cloned from VideoX-Fun yesterday.
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by comparing VideoX-Fun's WanTransformer3DModel with the diffusers 0.33 WanTransformer3DModel and the GGUF loading example in the issue. No project files or tests are named, so first locate the model implementation and its loading tests; done means the VideoX-Fun model can load the referenced GGUF files without breaking its existing behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100