#Did OpenAi had an update on Function Calling? The mapping is diferent...

1 messages · Page 1 of 1 (latest)

old cypress
#

Did OpenAi had an update on Function Calling? The mapping is diferent...

lime karmaBOT
#

Hi @old cypress. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:

#

Yes, OpenAI has made updates to the function calling feature. Here are some key points from the updates:

  1. Deprecation of Parameters: The parameters named functions and function_call have been deprecated in favor of using tools and tool_choice. This change is reflected in the way function calling is configured and monitored, as seen in the function calling guide (https://platform.openai.com/docs/guides/function-calling).

  2. Structured Outputs: In June 2024, OpenAI introduced Structured Outputs, which can be enabled by setting strict: true in your function definition. This ensures that the arguments generated by the model for a function call exactly match the JSON Schema you provided. More details can be found in the function calling guide (https://platform.openai.com/docs/guides/function-calling).

  3. Parallel Function Calling: Parallel function calling is supported on models released on or after November 6, 2023. This feature allows for more efficient function execution and is enabled by default, but can be disabled using parallel_tool_calls: false in the training example.

  4. Function Calling Use Cases: Function calling is useful for various applications, such as enabling AI assistants to fetch data, take actions, perform computations, and build rich workflows. These use cases are detailed in the function calling guide (https://platform.openai.com/docs/guides/function-calling).

For more detailed information, you can refer to the function calling guide (https://platform.openai.com/docs/guides/function-calling) and related documentation. If you have specific questions about the changes, feel free to ask!

Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: