You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
**The Unity Machine Learning Agents Toolkit** (ML-Agents) is an open-source
8
-
Unity plugin that enables games and simulations to serve as environments for
9
-
training intelligent agents. Agents can be trained using reinforcement learning,
10
-
imitation learning, neuroevolution, or other machine learning methods through a
11
-
simple-to-use Python API. We also provide implementations (based on TensorFlow)
12
-
of state-of-the-art algorithms to enable game developers and hobbyists to easily
13
-
train intelligent agents for 2D, 3D and VR/AR games. These trained agents can be
14
-
used for multiple purposes, including controlling NPC behavior (in a variety of
15
-
settings such as multi-agent and adversarial), automated testing of game builds
16
-
and evaluating different game design decisions pre-release. The ML-Agents
17
-
toolkit is mutually beneficial for both game developers and AI researchers as it
18
-
provides a central platform where advances in AI can be evaluated on Unity’s
19
-
rich environments and then made accessible to the wider research and game
20
-
developer communities.
5
+
**The Unity Machine Learning Agents Toolkit** (ML-Agents) is an open-source Unity plugin
6
+
that enables games and simulations to serve as environments for training
7
+
intelligent agents. Agents can be trained using reinforcement learning,
8
+
imitation learning, neuroevolution, or other machine learning methods through
9
+
a simple-to-use Python API. We also provide implementations (based on
10
+
TensorFlow) of state-of-the-art algorithms to enable game developers
11
+
and hobbyists to easily train intelligent agents for 2D, 3D and VR/AR games.
12
+
These trained agents can be used for multiple purposes, including
13
+
controlling NPC behavior (in a variety of settings such as multi-agent and
14
+
adversarial), automated testing of game builds and evaluating different game
15
+
design decisions pre-release. The ML-Agents toolkit is mutually beneficial for both game
16
+
developers and AI researchers as it provides a central platform where advances
17
+
in AI can be evaluated on Unity’s rich environments and then made accessible
18
+
to the wider research and game developer communities.
21
19
22
20
## Features
23
-
24
21
* Unity environment control from Python
25
22
* 10+ sample Unity environments
26
23
* Support for multiple environment configurations and training scenarios
27
-
* Train memory-enhanced agents using deep reinforcement learning
24
+
* Train memory-enhanced Agents using deep reinforcement learning
28
25
* Easily definable Curriculum Learning scenarios
29
-
* Broadcasting of agent behavior for supervised learning
26
+
* Broadcasting of Agent behavior for supervised learning
30
27
* Built-in support for Imitation Learning
31
-
* Flexible agent control with On Demand Decision Making
28
+
* Flexible Agent control with On Demand Decision Making
32
29
* Visualizing network outputs within the environment
33
30
* Simplified set-up with Docker
34
-
* Wrap learning environments as a gym
35
31
36
32
## Documentation
37
33
38
-
* For more information, in addition to installation and usage instructions, see
39
-
our [documentation home](docs/Readme.md).
40
-
* If you are a researcher interested in a discussion of Unity as an AI platform, see a pre-print of our [reference paper on Unity and the ML-Agents Toolkit](https://arxiv.org/abs/1809.02627). Also, see below for instructions on citing this paper.
41
-
* If you have used a version of the ML-Agents toolkit prior to v0.5, we strongly
42
-
recommend our [guide on migrating from earlier versions](docs/Migrating.md).
34
+
* For more information, in addition to installation and usage
35
+
instructions, see our [documentation home](docs/Readme.md).
36
+
* If you have
37
+
used a version of the ML-Agents toolkit prior to v0.4, we strongly recommend
38
+
our [guide on migrating from earlier versions](docs/Migrating.md).
43
39
44
-
## Additional Resources
40
+
## References
45
41
46
42
We have published a series of blog posts that are relevant for ML-Agents:
*[Using Machine Learning Agents in a real game: a beginner’s guide](https://blogs.unity3d.com/2017/12/11/using-machine-learning-agents-in-a-real-game-a-beginners-guide/)
and [Q-learning](https://blogs.unity3d.com/2017/08/22/unity-ai-reinforcement-learning-with-q-learning/))
46
+
-[Using Machine Learning Agents in a real game: a beginner’s guide](https://blogs.unity3d.com/2017/12/11/using-machine-learning-agents-in-a-real-game-a-beginners-guide/)
47
+
-[Post](https://blogs.unity3d.com/2018/02/28/introducing-the-winners-of-the-first-ml-agents-challenge/) announcing the winners of our
@@ -94,14 +84,8 @@ of the documentation to one language (Chinese), but we hope to continue
94
84
translating more pages and to other languages. Consequently,
95
85
we welcome any enhancements and improvements from the community.
96
86
97
-
*[Chinese](docs/localized/zh-CN/)
87
+
-[Chinese](docs/localized/zh-CN/)
98
88
99
89
## License
100
90
101
91
[Apache License 2.0](LICENSE)
102
-
103
-
## Citation
104
-
105
-
If you use Unity or the ML-Agents Toolkit to conduct research, we ask that you cite the following paper as a reference:
106
-
107
-
Juliani, A., Berges, V., Vckay, E., Gao, Y., Henry, H., Mattar, M., Lange, D. (2018). Unity: A General Platform for Intelligent Agents. *arXiv preprint arXiv:1809.02627.*https://github.com/Unity-Technologies/ml-agents.
0 commit comments