Skip to content

A programming framework for agentic AI

guide on the usage of docker (#1111)

Qingyun Wu643a0310622f
diff --git a/README.md b/README.md
index 052c41f85..99131cb46 100644
--- a/README.md
+++ b/README.md
@@ -62,18 +62,12 @@ The easiest way to start playing is
3. Start playing with the notebooks!
*NOTE*: OAI_CONFIG_LIST_sample lists GPT-4 as the default model, as this represents our current recommendation, and is known to work well with AutoGen. If you use a model other than GPT-4, you may need to revise various system prompts (especially if using weaker models like GPT-3.5-turbo). Moreover, if you use models other than those hosted by OpenAI or Azure, you may incur additional risks related to alignment and safety. Proceed with caution if updating this default.
+## [Installation](https://microsoft.github.io/autogen/docs/Installation)
+### Option 1. Install and Run AutoGen in Docker
-## Using existing docker image
-Install docker, save your oai key into an environment variable name OPENAI_API_KEY, and then run the following.
+Find detailed instructions for users [here](https://microsoft.github.io/autogen/docs/Installation#option-1-install-and-run-autogen-in-docker), and for developers [here](https://microsoft.github.io/autogen/docs/Contribute#docker).
-```
-docker pull yuandongtian/autogen:latest
-docker run -it -e OPENAI_API_KEY=$OPENAI_API_KEY -p 8081:8081 docker.io/yuandongtian/autogen:latest
-```
-
-Then open `http://localhost:8081/` in your browser to use AutoGen. The UI is from `./samples/apps/autogen-assistant`. See docker hub [link](https://hub.docker.com/r/yuandongtian/autogen) for more details.
-
-## Installation
+### Option 2. Install AutoGen Locally
AutoGen requires **Python version >= 3.8, < 3.12**. It can be installed from pip:
@@ -88,11 +82,11 @@ Minimal dependencies are installed without extra options. You can install extra
pip install "pyautogen[blendsearch]"
``` -->
-Find more options in [Installation](https://microsoft.github.io/autogen/docs/Installation).
+Find more options in [Installation](https://microsoft.github.io/autogen/docs/Installation#option-2-install-autogen-locally-using-virtual-environment).
<!-- Each of the [`notebook examples`](https://github.com/microsoft/autogen/tree/main/notebook) may require a specific option to be installed. -->
-For [code execution](https://microsoft.github.io/autogen/docs/FAQ/#code-execution), we strongly recommend installing the Python docker package and using docker.
+Even if you are installing AutoGen locally out of docker, we recommend performing [code execution](https://microsoft.github.io/autogen/docs/FAQ/#code-execution) in docker. Find more instructions [here](https://microsoft.github.io/autogen/docs/Installation#docker).
For LLM inference configurations, check the [FAQs](https://microsoft.github.io/autogen/docs/FAQ#set-your-api-endpoints).
diff --git a/samples/dockers/Dockerfile.base b/samples/dockers/Dockerfile.base
new file mode 100644
index 000000000..4d08608ff
--- /dev/null
+++ b/samples/dockers/Dockerfile.base
@@ -0,0 +1,21 @@
+FROM python:3.11-slim-bookworm
+
+RUN : \
+ && apt-get update \
+ && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
+ software-properties-common \
+ && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
+ python3-venv \
+ && apt-get clean \
+ && rm -rf /var/lib/apt/lists/* \
+ && :
+
+RUN python3 -m venv /venv
+ENV PATH=/venv/bin:$PATH
+EXPOSE 8081
+
+RUN cd /venv; pip install pyautogen
+# Pre-load popular packages as per https://learnpython.com/blog/most-popular-python-packages/
+RUN pip install numpy pandas matplotlib seaborn scikit-learn requests urllib3 nltk pillow pytest beautifulsoup4
+
+ENTRYPOINT []
diff --git a/Dockerfile b/samples/dockers/Dockerfile.dev
similarity index 62%
rename from Dockerfile
rename to samples/dockers/Dockerfile.dev
index 7f85ac9e5..07c0bb357 100644
--- a/Dockerfile
+++ b/samples/dockers/Dockerfile.dev
@@ -1,5 +1,5 @@
# basic setup
-FROM python:3.10
+FROM python:3.11-slim-bookworm
RUN apt-get update && apt-get -y update
RUN apt-get install -y sudo git npm
@@ -13,8 +13,8 @@ USER autogen-dev
RUN cd /home/autogen-dev && git clone https://github.com/microsoft/autogen.git
WORKDIR /home/autogen-dev/autogen
-# Install autogen (Note: extra components can be installed if needed)
-RUN sudo pip install -e .[test]
+# Install autogen in editable mode (Note: extra components can be installed if needed)
+RUN sudo pip install -e .[test,teachable,lmm,retrievechat,mathchat,blendsearch]
# Install precommit hooks
RUN pre-commit install
@@ -25,6 +25,9 @@ RUN sudo pip install pydoc-markdown
RUN cd website
RUN yarn install --frozen-lockfile --ignore-engines
+# Pre-load popular packages as per https://learnpython.com/blog/most-popular-python-packages/
+RUN pip install numpy pandas matplotlib seaborn scikit-learn requests urllib3 nltk pillow pytest beautifulsoup4
+
# override default image starting point
CMD /bin/bash
ENTRYPOINT []
diff --git a/samples/dockers/Dockerfile.full b/samples/dockers/Dockerfile.full
new file mode 100644
index 000000000..4307177d0
--- /dev/null
+++ b/samples/dockers/Dockerfile.full
@@ -0,0 +1,21 @@
+FROM python:3.11-slim-bookworm
+
+RUN : \
+ && apt-get update \
+ && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
+ software-properties-common \
+ && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
+ python3-venv \
+ && apt-get clean \
+ && rm -rf /var/lib/apt/lists/* \
+ && :
+
+RUN python3 -m venv /venv
+ENV PATH=/venv/bin:$PATH
+EXPOSE 8081
+
+RUN cd /venv; pip install pyautogen[teachable,lmm,retrievechat,mathchat,blendsearch] autogenra
+# Pre-load popular packages as per https://learnpython.com/blog/most-popular-python-packages/
+RUN pip install numpy pandas matplotlib seaborn scikit-learn requests urllib3 nltk pillow pytest beautifulsoup4
+
+ENTRYPOINT []
diff --git a/website/docs/Contribute.md b/website/docs/Contribute.md
index 3b13d2f8e..eb8305e7d 100644
--- a/website/docs/Contribute.md
+++ b/website/docs/Contribute.md
@@ -97,13 +97,19 @@ pip install -e autogen
### Docker
-We provide a simple [Dockerfile](https://github.com/microsoft/autogen/blob/main/Dockerfile).
+We provide [Dockerfiles](https://github.com/microsoft/autogen/blob/main/samples/dockers/Dockerfile.dev) for developers to use.
-```bash
-docker build https://github.com/microsoft/autogen.git#main -t autogen-dev
-docker run -it autogen-dev
+Use the following command line to build and run a docker image.
+
+```
+docker build -f samples/dockers/Dockerfile.dev -t autogen_dev_img https://github.com/microsoft/autogen.git#main
+
+docker run -it autogen_dev_img
```
+Detailed instructions can be found [here](Installation.md#option-1-install-and-run-autogen-in-docker).
+
+
### Develop in Remote Container
If you use vscode, you can open the autogen folder in a [Container](https://code.visualstudio.com/docs/remote/containers).
diff --git a/website/docs/FAQ.md b/website/docs/FAQ.md
index b63df3a18..a54d0ab24 100644
--- a/website/docs/FAQ.md
+++ b/website/docs/FAQ.md
@@ -71,8 +71,8 @@ The `AssistantAgent` doesn't save all the code by default, because there are cas
We strongly recommend using docker to execute code. There are two ways to use docker:
-1. Run autogen in a docker container. For example, when developing in GitHub codespace, the autogen runs in a docker container.
-2. Run autogen outside of a docker, while perform code execution with a docker container. For this option, make sure the python package `docker` is installed. When it is not installed and `use_docker` is omitted in `code_execution_config`, the code will be executed locally (this behavior is subject to change in future).
+1. Run AutoGen in a docker container. For example, when developing in [GitHub codespace](https://codespaces.new/microsoft/autogen?quickstart=1), AutoGen runs in a docker container. If you are not developing in Github codespace, follow instructions [here](Installation.md#option-1-install-and-run-autogen-in-docker) to install and run AutoGen in docker.
+2. Run AutoGen outside of a docker, while performing code execution with a docker container. For this option, set up docker and make sure the python package `docker` is installed. When not installed and `use_docker` is omitted in `code_execution_config`, the code will be executed locally (this behavior is subject to change in future).
### Enable Python 3 docker image
diff --git a/website/docs/Installation.md b/website/docs/Installation.md
index 0535ea2d3..647986e9f 100644
--- a/website/docs/Installation.md
+++ b/website/docs/Installation.md
@@ -1,10 +1,50 @@
# Installation
-## Setup Virtual Environment
+## Option 1: Install and Run AutoGen in Docker
-When not using a docker container, we recommend using a virtual environment to install AutoGen. This will ensure that the dependencies for AutoGen are isolated from the rest of your system.
+[Docker](https://www.docker.com/) 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](https://codespaces.new/microsoft/autogen?quickstart=1) or follow the instructions below to do so.
-### Option 1: venv
+#### Step 1. Install Docker.
+
+Install docker following [this instruction](https://docs.docker.com/get-docker/).
+
+For Mac users, alternatively you may choose to install [colima](https://smallsharpsoftwaretools.com/tutorials/use-colima-to-run-docker-containers-on-macos/) to run docker containers, if there is any issues with starting the docker daemon.
+
+#### Step 2. Build a docker image
+
+AutoGen provides [dockerfiles](https://github.com/microsoft/autogen/tree/main/samples/dockers/) 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](https://github.com/microsoft/autogen/blob/main/test/twoagent.py)) in a folder named `myapp`. With the command line below, you can mont your folder and run the application in docker.
+
+```python
+# 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
+```
+
+<!-- You may also run [AutoGen Studio](https://github.com/microsoft/autogen/tree/main/samples/apps/autogen-studio) (assuming that you have built a docker image named `autogen_full_img` with `Dockerfile.full` and you have set the environment variable `OPENAI_API_KEY` to your OpenAI API key) as below:
+
+```
+docker run -it -e OPENAI_API_KEY=$OPENAI_API_KEY -p 8081:8081 autogen_full_img:latest autogenra ui --host 0.0.0.0
+```
+Then open `http://localhost:8081/` in your browser to use AutoGen Studio. -->
+
+## 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:
```bash
@@ -17,9 +57,9 @@ The following command will deactivate the current `venv` environment:
deactivate
```
-### Option 2: conda
+### Option b: conda
-Another option is with `Conda`, Conda works better at solving dependency conflicts than pip. You can install it by following [this doc](https://docs.conda.io/projects/conda/en/stable/user-guide/install/index.html),
+Another option is with `Conda`. You can install it by following [this doc](https://docs.conda.io/projects/conda/en/stable/user-guide/install/index.html),
and then create a virtual environment as below:
```bash
conda create -n pyautogen python=3.10 # python 3.10 is recommended as it's stable and not too old
@@ -31,7 +71,7 @@ The following command will deactivate the current `conda` environment:
conda deactivate
```
-### Option 3: poetry
+### Option c: poetry
Another option is with `poetry`, which is a dependency manager for Python.
@@ -93,13 +133,26 @@ Inference parameter tuning can be done via [`flaml.tune`](https://microsoft.gith
### Optional Dependencies
- #### docker
-For the best user experience and seamless code execution, we highly recommend using Docker with AutoGen. Docker is a containerization platform that simplifies the setup and execution of your code. Developing in a docker container, such as GitHub Codespace, also makes the development convenient.
+Even if you install AutoGen locally, we highly recommend using Docker for [code execution](FAQ.md#enable-python-3-docker-image).
-When running AutoGen out of a docker container, to use docker for code execution, you also need to install the python package `docker`:
+To use docker for code execution, you also need to install the python package `docker`:
```bash
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.:
+```python
+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](https://arxiv.org/abs/2303.04673) for tuning Large Language Models. Please install with the [blendsearch] option to use it.