LLM Integration version has been released

An LLM integrated version has been released. The LLM is working locally on the CPU. So depending on your computer specs, it might take more or less time to generate the answer to your question.

To try the LLM, launch the walkie-talkie, talk, and stop it. Then, wait for an answer. The answer might be formulated in French as our dataset mainly contains french words. The LLM model is rather small (only 4GB), so its answers are not always coherent (Actually they are quite bad most of the time).

Enjoy !