0.19.1

• FIxed bug where llm.get_models() and llm.get_async_models() returned the same model multiple times. #667