Hi ZamaAI Team,
I am currently working on a research project that relies on privacy-preserving computation using Fully Homomorphic Encryption (FHE). In particular, I am using Concrete-ML to compile encrypted circuits for feature extraction in mobile security scenarios.
I’ve read that version 1.10 of Concrete-ML introduces new quantization capabilities, which would significantly help reduce the latency of encrypted inference in my application. However, I’ve been unable to access version 1.10 through either PyPI or by building from the main branch of the GitHub repository—installation always defaults to version 1.9.0.
Could you please advise on:
Whether Concrete-ML 1.10 is publicly available at this time?
If not, is there an estimated timeline for release?
Alternatively, is there a recommended way to access the quantization support and relevant APIs in the meantime?
Thank you for your time and for the excellent work your team is doing in the FHE space. I look forward to your guidance.