Great video! I also follow you on O'Reilly, and your lectures are really very insightful. It seems to me that you put a lot of passion and determination into what you do. Keep it up! Also, I would like to ask you one question: is it possible to provide also structured Pydantic input to openai function calling? In other words, in your code is it possible to transform "tool_create_directory" into a Pydantic class? I did not find very much documentation on that.
Your video is fantastic ! A quick question: I use Function Calling a lot. But is this new release from OpenAI, "Structured Outputs", better than using Function Calling? Does using one or the other affect Scalability in any way? Or does it make sense to use either? Thank you very much !
@@LearningWorldChatGPT structures output is about getting an output in a Joan like structure. Function calling is about calling functions with LLMs to perform actions like web search and so on, it’s not about one or the other, it’s about what is your specific use case :)
Great video! I also follow you on O'Reilly, and your lectures are really very insightful. It seems to me that you put a lot of passion and determination into what you do. Keep it up! Also, I would like to ask you one question: is it possible to provide also structured Pydantic input to openai function calling? In other words, in your code is it possible to transform "tool_create_directory" into a Pydantic class? I did not find very much documentation on that.
@@albertoprospero4330 in function calling you get structures inputs by default. Thanks! :)
Your video is fantastic !
A quick question: I use Function Calling a lot. But is this new release from OpenAI, "Structured Outputs", better than using Function Calling? Does using one or the other affect Scalability in any way? Or does it make sense to use either?
Thank you very much !
@@LearningWorldChatGPT structures output is about getting an output in a Joan like structure. Function calling is about calling functions with LLMs to perform actions like web search and so on, it’s not about one or the other, it’s about what is your specific use case :)
great Video
Thanks!! :)
Excellent content, really helpful! I was wondering if you can share the codes?
Sure source code is here: github.com/EnkrateiaLucca/oreilly-prompt-eng/blob/main/notebooks/2.5-structured-prompting-openai-function-calling-pydantic.ipynb
Wheres the github link for the code? How are we supposed to code along without a repo or something similar?
source code! :) github.com/EnkrateiaLucca/quiz-ai-app/tree/main
Thank you,,,
God is the Best Prompt Engenier 😎