The community members are discussing the limitations of GPT-3.5-turbo and GPT-4 in handling queries with ranges in joint query engines. They are seeking a workaround or enhancement to overcome this issue while using Llama-Index. One community member has encountered an error while playing around with the SQLAutoVectorQueryEngine API and is using the GPT-3.5-turbo model, not GPT-4.
Limitation of both gpt-3.5-turbo and gpt-4 in handling queries with range in joint query engines. Do we have any work-around or enhancement to overcome this issue while using Llama-Index?
