Completions API requests containing 'function_call" and “functions” are rejected as invalid when using the gpt-4-vision-preview model, with the following error message :
“2 validation errors for Request\nbody → function_call\n extra fields not permitted (type=value_error.extra)\nbody → functions\n extra fields not permitted (type=value_error.extra)”,“type”:“invalid_request_error”
Same call works with the gpt-4-1106-preview model.
Yes, it seems the vision model may not support function calling yet, despite this from the description of the vision model: “Ability to understand images, in addition to all other GPT-4 Turbo capabilties.”
Same issue here.
Wondering when function calling will come to Vision model.
It would also be easier to switch between models if the api could
just ignore unknown parameters instead of erroring out.
This! What a pain, that’s caused me to have to add another otherwise redundant setting and If statement …