Add model card with architecture details, provenance, and training metrics

Upload README.md with huggingface_hub

Upload best_model.json with huggingface_hub

initial commit