This project hosts a General Assistant created specifically for benchmarking and evaluation against the GIAI (General AI Assistant) dataset. The primary goal is to assess the capabilities of large language models in a conversational context, focusing on metrics like coherence, helpfulness, and safety.
The assistant is currently in a deep development phase as we refine the core logic and expand its functionalities.