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

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 → TF Agents checkpointer and PolicySaver → Train a REINFORCE agent → Simple statistical gradient-following algorithms for connectionist reinforcement learning → TensorFlow Lite homepage → Github → 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 → Subscribe to TensorFlow → Ask your ques