Build a board game app with policy gradient (Reinforcement learning with TensorFlow Agents) @TensorFlow

Uploaded February 2022 | Updated September 2026, 1 week ago

Wei Wei, a Developer Advocate for TensorFlow, walks through how to build a board game app that leverages TF Agents to train a REINFORCE agent for a custom environment, convert the model to TFLite, and deploy it on Android.

Resources:
Building a board game app with TensorFlow: a new TensorFlow Lite reference app →
https://goo.gle/3oy7Z9K

TF Agents checkpointer and PolicySaver → https://goo.gle/3HELl6X

Train a REINFORCE agent → https://goo.gle/3oB6gAi

Simple statistical gradient-following algorithms for connectionist reinforcement learning → https://goo.gle/3uwtThr

TensorFlow Lite homepage → https://goo.gle/2QCwBjG

Github → https://goo.gle/3sptGKe

Chapters:
00:00 Introduction
00:40 About the game
01:23 Using Reinforce (policy gradient)
02:08 Creating a TF Agents custom environment
04:35 TensorFlow Lite
05:28 Summary and resources

Watch more Reinforcement learning with TensorFlow Agents episodes → https://goo.gle/reinforcement-learning
Subscribe to TensorFlow → https://goo.gle/TensorFlow
Ask your questions on the TF Forum → https://goo.gle/discuss_tensorflow

#TensorFlow #MachineLearning #ML

product: TensorFlow - General; fullname: Wei Wei;

Build a board game app with policy gradient (Reinforcement learning with TensorFlow Agents)

Radiological image segmentation with Medseg - Made with TensorFlow.js

Community matters: 8 reasons why you should be involved with Kaggle

Im a developer, how can I become an ML developer? | Breakout session

Item-to-item recommendation and sequential recommendation

Tulsee Doshi - Responsible AI and human centered technology

Simon Tokumine - Gemini and the future of Generative AI tools

Building Neural Collaborative Filtering recommendation model

3D MRI brain segmentation - Made with TensorFlow.js

The “spell check” of design systems - Made with TensorFlow.js