Unity 机器学习代理工具包 (ML-Agents) 是一个开源项目,它使游戏和模拟能够作为训练智能代理的环境。
您最多选择25个主题 主题必须以中文或者字母或数字开头,可以包含连字符 (-),并且长度不得超过35个字符
 
 
 
 
 
GitHub b2fa2268 Merge pull request #2648 from Unity-Technologies/release-0.10.0 5 年前
..
proto/mlagents/envs/communicator_objects Merge pull request #2648 from Unity-Technologies/release-0.10.0 5 年前
README.md Merge branch 'develop' into protobuf_update 5 年前
make.bat Enable mypy in precommit checks (#2177) 5 年前
make_for_win.bat Update make_for_win_.bat to match make.bat 5 年前

README.md

Unity ML-Agents Protobuf Definitions

Contains relevant definitions needed to generate probobuf files used in ML-Agents Toolkit.

Requirements

  • protobuf 3.6.0
  • grpcio-tools 1.11.1
  • Grpc.Tools 1.14.1

Set-up & Installation

First we will follow these steps once install protobuf and grpcio-tools via your terminal. Assume the ml-agents repository is checked out to a folder named $MLAGENTS_ROOT. Note: If you're using Anaconda, don't forget to activate the ml-agents environment first.

pip install protobuf==3.6.0 --force

pip install grpcio-tools==1.11.1

pip install mypy-protobuf

If you don't have it already, download the latest version of nuget. Navigate to your installation of nuget and run the following:

nuget install Grpc.Tools -Version 1.14.1 -OutputDirectory $MLAGENTS_ROOT\protobuf-definitions

Installing Protobuf Compiler

On Mac: brew install protobuf

Running

Whenever you change the fields of a custom message, you must follow the steps below to create C# and Python files corresponding to the new message.

  1. Open a terminal. Note: If you're using Anaconda, don't forget to activate the ml-agents environment first.
  2. Un-comment line 7 in make.bat (for Windows, use make_for_win.bat), and set to correct Grpc.Tools sub-directory.
  3. Run the .bat from the terminal by navigating to $MLAGENTS_ROOT\protobuf-definitions and entering make.bat (for Windows, use make_for_win.bat)
  4. Note any errors generated that may result from setting the wrong directory in step 2.
  5. In the generated UnityToExternalGrpc.cs file in the $MLAGENTS_ROOT/UnitySDK/Assets/ML-Agents/Scripts/CommunicatorObjects folder, you will need to add the following to the beginning of the file:
# if UNITY_EDITOR || UNITY_STANDALONE_WIN || UNITY_STANDALONE_OSX || UNITY_STANDALONE_LINUX

and the following line to the end

#endif

This is to make sure the generated code does not try to access the Grpc library on platforms that are not supported by Grpc.

Finally, re-install the mlagents packages by running the following commands from the same $MLAGENTS_ROOT\protobuf-definitions directory.

cd ..
cd ml-agents-envs
pip install -e .
cd ..
cd ml-agents
pip install -e .
mlagents-learn

The final line will test if everything was generated and installed correctly. If it worked, you should see the Unity logo.