Training Atari games with VPG

moabitcoin / cherry-pytorch

Reinforcement Learning Tutorials & other bedtime stories in PyTorch

MIT License

11 stars 1 forks source link

Training Atari games with VPG #36

Closed sandhawalia closed 4 years ago

sandhawalia commented 4 years ago

We currently train Atari models with DQN + DDQN. VPG (Vanilla Policy Gradient) have shown to be a better structured agent (as tested on Control problems / Health gathering in Doom).

Generate model + hyper params which can solve Breakout Atari retro game

sandhawalia commented 4 years ago

Being addressed here https://github.com/moabitcoin/cherry-pytorch/pull/40