From the course: Level up LLM applications development with LangChain and OpenAI
Unlock the full course today
Join today to access over 24,800 courses taught by industry experts.
Solution: Define tools and functions
From the course: Level up LLM applications development with LangChain and OpenAI
Solution: Define tools and functions
(upbeat music) - So now it's time for the challenge. And actually we're going to see together the solution to the challenge, which is to create a multitask agent. And we want to give this agent the capacity to search the web. And here we're going to find under web search, so this implementation, so that allows to load a data source from the internet, so which is from an HTML document. So we use web-based loader. And so we go through the same process to index the documents, split the documents into chunks, and then store the embeddings into a chroma vector store. And then we use the vector store as a retriever. We also use this prompt that gives the language model the capacity, the ability to generate responses based on a user query and also based on a specific context. And so the exercise here will be to construct the chain. So a bit just like we did for Vector Search Index, very similar. We're going to construct the chain and allow to execute this chain from using the agents. So…
Contents
-
-
-
-
-
-
-
-
-
-
(Locked)
Using agents to perform actions in chains1m 36s
-
(Locked)
Define tools5m 37s
-
(Locked)
Select the perfect prompt1m 12s
-
(Locked)
Bind tools and create agent2m 19s
-
(Locked)
Create and run the agent executor4m 41s
-
(Locked)
Challenge: Create a multitask agent5m 31s
-
(Locked)
Solution: Define tools and functions6m 9s
-
(Locked)
-
-
-