NVIDIA-NeMo / NVIDIA-NeMo/Automodel
Custom registry for AutoProcessor
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 963
- Forks
- 318
- Avg merge
- 3d 20h
- Merged PRs (30d)
- 143
Description
Is your feature request related to a problem? Please describe.
For multimodal models (e.g., non-textual LLMs), preprocessing is often more involved and less standardized. As a result it would be great to have the ability to extend/implement custom preprpocessing for models that lack standardization or implementation.
Describe the solution you'd like
AutoProcessor registry similar to the custom model registry.
Describe alternatives you've considered
Fighting with yaml configs might be one option, but it would be great to have something more user friendly and less painful.
Additional context.
Contributor guide
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 locating the existing custom model registry and the AutoProcessor integration points. Compare how model registration currently works with the requested processor extension and review the YAML configuration path mentioned in the issue. Done means a documented, user-friendly way to register custom preprocessing for multimodal models without relying on YAML configuration.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100