autogen/website/docs/Installation.md

Installation

Option 1: Install and Run AutoGen in Docker

Docker is a containerization platform that simplifies the setup and execution of your code. A properly built docker image could provide isolated and consistent environment to run your code securely across platforms. One option of using AutoGen is to install and run it in a docker container. You can do that in Github codespace or follow the instructions below to do so.

Step 1. Install Docker.

Install docker following this instruction.

For Mac users, alternatively you may choose to install colima to run docker containers, if there is any issues with starting the docker daemon.

Step 2. Build a docker image

AutoGen provides dockerfiles that could be used to build docker images. Use the following command line to build a docker image named autogen_img (or other names you prefer) from one of the provided dockerfiles named Dockerfile.base:

docker build -f samples/dockers/Dockerfile.base -t autogen_img https://github.com/microsoft/autogen.git#main

which includes some common python libraries and essential dependencies of AutoGen, or build from Dockerfile.full which include additional dependencies for more advanced features of AutoGen with the following command line:

docker build -f samples/dockers/Dockerfile.full -t autogen_full_img https://github.com/microsoft/autogen.git

Once you build the docker image, you can use docker images to check whether it has been created successfully.

Step 3. Run applications built with AutoGen from a docker image.

Mount your code to the docker image and run your application from there: Now suppose you have your application built with AutoGen in a main script named twoagent.py (example) in a folder named myapp. With the command line below, you can mont your folder and run the application in docker.

# Mount the local folder `myapp` into docker image and run the script named "twoagent.py" in the docker.
docker run -it -v `pwd`/myapp:/myapp autogen_img:latest python /myapp/main_twoagent.py

Option 2: Install AutoGen Locally Using Virtual Environment

When installing AutoGen locally, we recommend using a virtual environment for the installation. This will ensure that the dependencies for AutoGen are isolated from the rest of your system.

Option a: venv

You can create a virtual environment with venv as below:

python3 -m venv pyautogen
source pyautogen/bin/activate

The following command will deactivate the current venv environment:

deactivate

Option b: conda

Another option is with Conda. You can install it by following this doc, and then create a virtual environment as below:

conda create -n pyautogen python=3.10  # python 3.10 is recommended as it's stable and not too old
conda activate pyautogen

The following command will deactivate the current conda environment:

conda deactivate

Option c: poetry

Another option is with poetry, which is a dependency manager for Python.

Poetry is a tool for dependency management and packaging in Python. It allows you to declare the libraries your project depends on and it will manage (install/update) them for you. Poetry offers a lockfile to ensure repeatable installs, and can build your project for distribution.

You can install it by following this doc, and then create a virtual environment as below:

poetry init
poetry shell

poetry add pyautogen

The following command will deactivate the current poetry environment:

exit

Now, you're ready to install AutoGen in the virtual environment you've just created.

Python

AutoGen requires Python version >= 3.8, < 3.12. It can be installed from pip:

pip install pyautogen

pyautogen<0.2 requires openai<1. Starting from pyautogen v0.2, openai>=1 is required.

Migration guide to v0.2

openai v1 is a total rewrite of the library with many breaking changes. For example, the inference requires instantiating a client, instead of using a global class method. Therefore, some changes are required for users of pyautogen<0.2.

• api_base -> base_url, request_timeout -> timeout in llm_config and config_list. max_retry_period and retry_wait_time are deprecated. max_retries can be set for each client.
• MathChat is unsupported until it is tested in future release.
• autogen.Completion and autogen.ChatCompletion are deprecated. The essential functionalities are moved to autogen.OpenAIWrapper:

from autogen import OpenAIWrapper
client = OpenAIWrapper(config_list=config_list)
response = client.create(messages=[{"role": "user", "content": "2+2="}])
print(client.extract_text_or_completion_object(response))

• Inference parameter tuning and inference logging features are currently unavailable in OpenAIWrapper. Logging will be added in a future release. Inference parameter tuning can be done via flaml.tune.
• seed in autogen is renamed into cache_seed to accommodate the newly added seed param in openai chat completion api. use_cache is removed as a kwarg in OpenAIWrapper.create() for being automatically decided by cache_seed: int | None. The difference between autogen's cache_seed and openai's seed is that:
  • autogen uses local disk cache to guarantee the exactly same output is produced for the same input and when cache is hit, no openai api call will be made.
  • openai's seed is a best-effort deterministic sampling with no guarantee of determinism. When using openai's seed with cache_seed set to None, even for the same input, an openai api call will be made and there is no guarantee for getting exactly the same output.

Optional Dependencies

• docker

Even if you install AutoGen locally, we highly recommend using Docker for code execution.

To use docker for code execution, you also need to install the python package docker:

pip install docker

You might want to override the default docker image used for code execution. To do that set use_docker key of code_execution_config property to the name of the image. E.g.:

user_proxy = autogen.UserProxyAgent(
    name="agent",
    human_input_mode="TERMINATE",
    max_consecutive_auto_reply=10,
    code_execution_config={"work_dir":"_output", "use_docker":"python:3"},
    llm_config=llm_config,
    system_message=""""Reply TERMINATE if the task has been solved at full satisfaction.
Otherwise, reply CONTINUE, or the reason why the task is not solved yet."""
)

• blendsearch

pyautogen<0.2 offers a cost-effective hyperparameter optimization technique EcoOptiGen for tuning Large Language Models. Please install with the [blendsearch] option to use it.

pip install "pyautogen[blendsearch]<0.2"

Example notebooks:

Optimize for Code Generation

Optimize for Math

• retrievechat

pyautogen supports retrieval-augmented generation tasks such as question answering and code generation with RAG agents. Please install with the [retrievechat] option to use it.

pip install "pyautogen[retrievechat]"

RetrieveChat can handle various types of documents. By default, it can process plain text and PDF files, including formats such as 'txt', 'json', 'csv', 'tsv', 'md', 'html', 'htm', 'rtf', 'rst', 'jsonl', 'log', 'xml', 'yaml', 'yml' and 'pdf'. If you install unstructured (pip install "unstructured[all-docs]"), additional document types such as 'docx', 'doc', 'odt', 'pptx', 'ppt', 'xlsx', 'eml', 'msg', 'epub' will also be supported.

You can find a list of all supported document types by using autogen.retrieve_utils.TEXT_FORMATS.

Example notebooks:

Automated Code Generation and Question Answering with Retrieval Augmented Agents

Group Chat with Retrieval Augmented Generation (with 5 group member agents and 1 manager agent)

Automated Code Generation and Question Answering with Qdrant based Retrieval Augmented Agents

• Teachability

To use Teachability, please install AutoGen with the [teachable] option.

pip install "pyautogen[teachable]"

Example notebook: Chatting with a teachable agent

• Large Multimodal Model (LMM) Agents

We offered Multimodal Conversable Agent and LLaVA Agent. Please install with the [lmm] option to use it.

pip install "pyautogen[lmm]"

Example notebooks:

LLaVA Agent

• mathchat

pyautogen<0.2 offers an experimental agent for math problem solving. Please install with the [mathchat] option to use it.

pip install "pyautogen[mathchat]<0.2"

Example notebooks:

Using MathChat to Solve Math Problems