YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

CheapMoE-Vul LoRA Adapters

Pre-trained LoRA adapter weights for CheapMoE-Vul vulnerability detection.

Contents

  • adapters_cm_dv/ β€” DiverseVul adapters (Qwen2.5-Coder-1.5B, 8 experts, ~1.18GB)
  • adapters_cm_xlang/ β€” CrossVul adapters (Qwen2.5-Coder-1.5B, 5 experts, ~739MB)
  • adapters_cm_dv7b/ β€” DiverseVul adapters (Qwen2.5-7B, 2 experts, ~323MB)

Usage

Download and extract to your artifact directory:

cd /path/to/artifact

# Download from HuggingFace Hub
huggingface-cli download cheapmoe-vul-anon/cheapmoe-adapters --local-dir ./hf_download

# Move to correct locations
mv hf_download/adapters_cm_dv/ ./
mv hf_download/adapters_cm_xlang/ ./
mv hf_download/adapters_cm_dv7b/ ./

# Cleanup
rm -rf hf_download

Note

These adapters are optional for verification. All paper results are pre-computed and stored in the results*/ directories of the main artifact repository.

You only need these adapters if you want to:

  • Re-run inference from scratch
  • Retrain or fine-tune the models
  • Experiment with different configurations

For quick verification of paper claims, see the main artifact README.

License

MIT

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support