04. Agent utilizing Claude, Gemini, Ollama, Together.ai

Tool calling agent other than OpenAI

In addition to OpenAI Anthropic , Google Gemini , Together.ai , Ollama , Mistral Supports a wider range of supplier implementations such as.

In this chapter, we will look at how to create and run tool calling agents using various LLMs.

Reference link

# Configuration file for managing API keys as environment variables
from dotenv import load_dotenv

# Load API key information
load_dotenv()
 True 
# Set up LangSmith tracking. https://smith.langchain.com
# !pip install -qU langchain-teddynote
from langchain_teddynote import logging

# Enter a project name.
logging.langsmith("CH15-Agents")
 Start tracking LangSmith. 
[Project name] 
CH15-Agents

Create prompt for Agent

  • chat_history : Variables that store previous conversations (if you don't support multi-turns, you can omit them)

  • agent_scratchpad : Variables that the agent temporarily saves

  • input : User input

Various LLM lists supporting Tool Calling

You need to set the content below to proceed with the lab.

Anthropic

Gemini

Together AI

Ollama

langchain-ollama installation

Generate Agent based on LLM.

Execute after creating AgentExecutor and confirm results

Run the agent using various llm.

Here is a function that creates and executes Agent using the entered llm to output the result.

Generate and run agents with each llm star to output results.

Previous03. AgentNext05. Iteration function and human intervention (Human-in-the-loop)

Last updated