Fix Python usage example: use llm.chat() with the chat template

Add AWQ-W4A16 quantization of Jackrong/Qwopus3.5-9B-Coder

initial commit