search

rgilman33 / simple-A2C-PPO

Actor-critic trained w PPO on OpenAI's Procgen Benchmark (PyTorch). Built from scratch.
100 stars 26 forks source link

critic and actor have same hidden layers #1

Open pjcorp opened 6 years ago

pjcorp commented 6 years ago

Hello, Correct me if I'm wrong, I'm under the impression that the critic and the actor share the same hidden layers in the tutorial notebook, why that constraint? Thanks

rgilman33 commented 6 years ago

Not so much a constraint as a consolidation. There's a large overlap in what the Actor and the Critic need to learn--no need for them each to learn it from scratch.

pjcorp notifications@github.com 于2018年8月17日周五 下午4:41写道：

Hello, Correct me if I'm wrong, I'm under the impression that the critic and the actor share the same hidden layers in the tutorial notebook, why that constraint? Thanks

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/rgilman33/simple-A2C/issues/1, or mute the thread https://github.com/notifications/unsubscribe-auth/ANW0Hng61jM5Nad1ua-XdkGCGB6S4slTks5uR0aagaJpZM4WCSXR .