Could you kindly trigger the CUDA build for v0.3.2 and publish them to GitHub [Releases](https://github.com/abetlen/llama-cpp-python/releases)? Thank you 🙏!