NVIDIA-NeMo / NVIDIA-NeMo/Automodel
Include sha256 of model.safetensors.index.json in ModelRegistry
Open
Nobody has claimed this yet.
enhancement
- Dominant language
- Python
- Stars
- 963
- Forks
- 318
- Avg merge
- 3d 20h
- Merged PRs (30d)
- 143
Description
We're starting to see a lot of issues where architectures isn't enough as a lookup, and two different models with the same architectures field get instantiated with the same custom model despite having different model architectures, leading to a crash. We need to have a permanent solution to this in place
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 ModelRegistry and the handling of model.safetensors.index.json. Trace how architectures are currently used for lookup, then define and verify that including the file's SHA-256 distinguishes models with different architectures without relying on architectures alone.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100