BRAND NEW OpenAI Function Calling | LangChain OpenAI functions agent

Exciting news from OpenAI! They've just introduced function calling. Check out my latest video where I walk through a simple example, a complex example, LangChain's impressive timely support for functions, and LangChain's fresh OpenAI functions agent.

00:00 intro
00:40 simple example
05:45 complex example to pull data from SQL database
07:39 covert LangChain tools to OpenAI functions
08:48 LangChain OpenAI functions agent

⭐ Stay in touch ⭐
💚 #AI

[ashinidealwis]

So grateful for this video - a clear, concise walkthrough that's been helpful for a beginner like me! A refreshing change from some of the other AI content out there that assumes you already know what they are talking about :D Thanks Sophia!

[nntun]

best explaination and correct script..other channels rushed it and got some lines wrong..

[codeAI]

I just learned how to use new OpenAI's function calling capability in the Chat Completions API when my LLM generates the SQL code to query a (Postgres) Database. Nice way to connect GPT's capabilities with external tools /postgres and APIs. great demo!

[andfanilo]

So fast! Thanks for the demo, I was too lazy to read the actual blog post 😜

[OliviaChen-krjf]

Great video. Love to see more female YouTuber doing AI content. Go girl!!

[jiqiyang]

Hi Sophia, great work! I am curious when we use the current weather function, which external api does openai goes to in order to get the internet data?

[matemarschalko]

Interestingly, the assistant seems to just use the function output almost "as is". I'm unable to get the assistant to use the function output as context only and focus on the users actual question...

Example: my function returns all the information about a property, and the assistant seems to just output the whole block of information instead of using it as a context/knowledge base to answer the users question. Nothing I put in the system message affects how the assistant uses the information from the function call.

Any idea how this could be achieved?

[ahmadalis]

Thanks, but I can't find LangChain example :(

[gileneusz]

now we can call Cats As A Service API directly from gpt-4 😍

[kajasheriff]

What is the different between OpenAi Plugins vs OpenAi Function Calling?

[SrishtiJaiswal-YearBTechElectr]

Srishti Jaiswal 1:19 AM
Hey Sophia I want to know can I pass the prompt to agent while initializing it.
example agent = initialize_agent(
tools, llm, agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION, verbose=True
) if not how can I use prompt or prompt template with agent

[williamipark]

I've been having issues with it not sticking to the correct casing that I provide in enums, in the JSON schema. Doesn't seem like it can be steered very well with system prompts to fix this issue either. Anyone else?

[sgramstrup]

A bit sad that openai gets to shape so much of the development ecosystem. They don't have the ecology/participants in mind when they evolve their standards. We probably can't do it fully right now, but as soon as possible, we should boycott openai and all their closed proprietary shenannigans.

[klammer]

Does this take langchain tool use out of the loop entirely? Why would I use a langchain agent with tools when I can just do a function call directly through the openai endpoint? Thanks for the explanation and great video as always😎🥳🦾