Instructions to use skoneru/OmniFusion with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use skoneru/OmniFusion with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "translation" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # 'pip install "transformers<5.0.0' from transformers import pipeline pipe = pipeline("translation", model="skoneru/OmniFusion")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("skoneru/OmniFusion", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Improve model card: Add pipeline tag, paper, code, and Hugging Face model links
#1
by nielsr HF Staff - opened
This PR enhances the model card for OmniFusion by:
- Adding the
pipeline_tag: translationto the metadata, improving discoverability on the Hugging Face Hub under the correct task (e.g., https://huggingface.co/models?pipeline_tag=translation). - Including a direct link to the Hugging Face paper page (https://huggingface.co/papers/2512.00234) in the "Publication" section, alongside the existing arXiv link.
- Adding an explicit link to the GitHub code repository (https://github.com/saikoneru/OmniFusion) for easy access to the source code.
- Adding a link to the Hugging Face model itself (https://huggingface.co/skoneru/OmniFusion), as mentioned in the GitHub README.
The existing library_name: transformers is kept, as it is consistent with the model's dependencies and integration with the Transformers library. The sample usage is also preserved as it is directly from the GitHub README and functional.
Please review and merge if these improvements are satisfactory!
skoneru changed pull request status to merged