How to get started with AutoGen 0.2

Environment

Windows 11
Python 3.12.3
Git
OpenAI API Key

Steps

Create a new repository called autogen_test in GitHub
- Type: Private
- Add a README file
- Add .gitignore: Python
- Choose a lisence: None
Clone the repository to VSCode
- Copy Clone URL from the <> Code button at the top page of the repository on GitHub
- Start VSCode and type Shift+Ctrl+P to open a command palette
- Type clone and select Git: Clone from the command menu
- Paste the Clone URL
- Select a folder to save the repository folder
In VSCode, create a new file called .env, add the following lines, and save
```
OPENAI_MODEL='gpt-4o-mini'
OPENAI_API_KEY='[your open api key]'
```
Create a new file called requirements.txt and add the following lines
```
ipykernel
pyautogen
```
Open a new terminal at VSCode, create a virtual environment
```
> py.exe -m venv venv
> .\venv\Scripts\activate
```

Install Python modules in the virtual environment

(venv) > python -m pip install -U pip
(venv) > python -m pip install -r requirements.txt

Create a new (notebook) file called autogen_test.ipynb, then do the following
- Type Shift+Ctrl+P to open a command palette
- Type notebook kernel and select Notebook: Select Notebook Kernel
- Select Python Environment and venv (Python 3.xx.x) venv\Scripts\python.exe
Add a Markdown cell to the notebook file
- Select + Markdown button from the top menu of the notebook to create a Markdown cell and add the content
```
# AutoGen Test
```
- Type Ctrl+Enter to run the Markdown cell

Add a Code cell below the previous cell

Select + Code button below the previous cell to create a new Code cell and add the content

import os
from dotenv import load_dotenv
from autogen import ConversableAgent

load_dotenv()

config_list = {
    "model": os.getenv('OPENAI_MODEL'),
    "api_key": os.getenv('OPENAI_API_KEY')
}

config_list_cathy = config_list.copy()
config_list_cathy["temperature"] = 0.9
config_list_joe = config_list.copy()
config_list_joe["temperature"] = 0.7

Type Ctrl+Enter to run the cell

Add another Code cell below the previous cell and run

def start_chat(cathy_setting, joe_setting, initial_message, max_turns=2):
  cathy = ConversableAgent(
      "cathy",
      system_message=cathy_setting,
      llm_config=config_list_cathy
  )
  joe = ConversableAgent(
      "joe",
      system_message=joe_setting,
      llm_config=config_list_joe
  )
  joe.initiate_chat(cathy, message=f"{initial_message}", max_turns=max_turns)

Add another Code cell and run

cathy_setting = "Your name is Cathy and you are a part of a duo of comedians."
joe_setting = "Your name is Joe and you are a part of a duo of comedians."
initial_message = "Cathy, tell me a joke."
max_turns = 2

start_chat(cathy_setting, joe_setting, initial_message, max_turns)

You will see something like this (not exactly the same).

joe (to cathy):

Cathy, tell me a joke.

--------------------------------------------------------------------------------

>>>>>>>> USING AUTO REPLY...
cathy (to joe):

Sure, Joe! Why did the scarecrow win an award? 

Because he was outstanding in his field! 🌾😄

--------------------------------------------------------------------------------

>>>>>>>> USING AUTO REPLY...
joe (to cathy):

That's a classic, Cathy! Alright, here’s one for you: Why don’t scientists trust atoms? 

Because they make up everything!

--------------------------------------------------------------------------------

>>>>>>>> USING AUTO REPLY...
...

An impasta! 🍝😆

--------------------------------------------------------------------------------
Output is truncated. View as a scrollable element or open in a text editor. Adjust cell output settings...

Click text editor in the bottom message to view the entire conversation.

What's next?

Change cathy_setting and joe_setting and re-run the cell to see how conversations change.
- Change the age, location, job, hobby, etc.
Change initial_message to change the topic of conversations.
- What do you think about X?
- Let's discuss the topic of X.
Increase the max_turns to 4 to change the length of conversations.
Commit your changes to the repository from VSCode.

hideojoho/how_to_get_started_with_autogen_0.2.md

How to get started with AutoGen 0.2

Environment

Steps

What's next?