A fast inference library for running LLMs locally on modern consumer-class GPUs
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).