ML-Agents Python Setup with Anaconda

In this tutorial, you will learn:

How to download and install Anaconda (for easy Python management)
How to set up a Python environment for Unity ML-Agents

Anaconda Installation

While there are other ways to install Python, I find that Anaconda is the easiest way to manage multiple Python environments. It’s completely free and works on Windows, Mac, and Linux

Navigate to https://www.anaconda.com/distribution/
Choose your operating system (e.g. Windows)
Download the Python 3.7 version
Run the installer (Anaconda installation documentation)

Anaconda Environment Setup

Once Anaconda is installed, you will need to set up a new environment for ML-Agents.

The first thing you'll need to do is determine which release of ML-Agents you're working with. This is a little tricky because they have multiple versioning systems, one for the Package Manager, and one for GitHub.

Unity Package Manager: ML-Agents

Find which Package version you're using by looking in the Package Manager. In my case, the version I'm using is 2.0.1
Go to the ML-Agents Releases page
Find the release with the closest match to your package version. In my case it's Release 17, which uses com.unity.ml-agents (C#) v2.0.0, which isn't exact, but is the only 2.0.x release.
Look in the table for the ml-agents (Python) release. In my case it's v0.26.0

So now we have the info we need: Release 17, ml-agents (Python) v0.26.0.

Open the newly installed “Anaconda Prompt” (Anaconda prompt documentation)

Run the following command

conda create -n ml-agents-r17 python=3.7

This will create a new Python 3.7 environment called “ml-agents-r17” for release 17. You can name it to whatever you like, I just like to give myself a hint in the future when I have multiple environments with different releases installed.

Anaconda prompt window: Environment is waiting on you to type Y to proceed

Type “y” and press Enter to proceed

Anaconda prompt window: Environment activation

Follow the instructions to activate the environment. In my case, I ran
```
conda activate ml-agents-r17
```
Confirm that the environment is active by looking for “(ml-agents-r17)” on the left side of the command prompt

Windows users will need to install PyTorch

pip3 install torch~=1.7.1 -f https://download.pytorch.org/whl/torch_stable.html

Run the following command to install the correct version of mlagents (v0.26.0 in my case):
```
pip3 install mlagents==0.26.0
```

You should now be ready to train your ML-Agents!

3D, Deep Learning, UnityAdam KellyFebruary 8, 2021Unity ML Agents, Artificial Intelligence6 Comments