Hybrid for Classification
Overview
A small Hybrid implementation for Classification, packaged with an explicit configuration and an initialization checkpoint. The base variant is a reproducible starting point, not a trained model release.
Quick check
Run python inference.py --help for the built-in smoke test. Because this is a custom implementation, generic automatic loading APIs require an explicit adapter before use.
Limitations
The checkpoint is untrained and unaudited. Treat the repository as an experimental starting point; anything learned from a future trained version must be documented separately.
Default experiment recipe
Defaults: novograd optimizer, linear warmup schedule. Treat these as starting points; a real run should sweep learning rate and batch size before drawing conclusions.
Training details
- Weights initialized with a xavier uniform scheme.
- Runs on a single consumer GPU.
- Full smoke test takes minutes, not hours.
- No distributed setup required.
Repository status
- Single-file implementation; the
__main__block runs a smoke test. config.jsonandtraining_args.jsonhold the generated settings.model.safetensorsis freshly initialized — plumbing, not results.- No benchmark numbers are reported on purpose.
Evaluation guidance
A first evaluation could use a task-specific labeled split. Report the metric across at least three seeds and include a matched-capacity baseline; keep logs and library versions with any published number.
Files
| File | Purpose |
|---|---|
inference.py |
primary artifact |
README.md |
this documentation |
config.json |
architecture configuration |
training_args.json |
default experiment settings |
model.safetensors |
initialization checkpoint |
License
Released under mpl-2.0. Review the source-data terms separately when this repository is used with external datasets.
- Downloads last month
- -