From 595e62e111f61966795eff05b07f62f8de1d3aeb Mon Sep 17 00:00:00 2001 From: Tongzheng Ren Date: Mon, 6 Nov 2017 15:15:44 +0800 Subject: [PATCH] architecture design --- tianshou/core/README.md | 24 ++++++++++++++++++++++++ tianshou/optimizer/README.md | 13 +++++++++++-- 2 files changed, 35 insertions(+), 2 deletions(-) create mode 100644 tianshou/core/README.md diff --git a/tianshou/core/README.md b/tianshou/core/README.md new file mode 100644 index 0000000..2ef89eb --- /dev/null +++ b/tianshou/core/README.md @@ -0,0 +1,24 @@ +# Core + +## Optimizer +TODO: + +### policy based: + +Vanilla + +Baseline + +TRPO + +PPO + +NAF + +GAE + +DPG + +### value based: + +TD diff --git a/tianshou/optimizer/README.md b/tianshou/optimizer/README.md index 31775ae..e80c0d8 100644 --- a/tianshou/optimizer/README.md +++ b/tianshou/optimizer/README.md @@ -1,11 +1,20 @@ # Optimizer for policy gradient methods TODO: + vanilla -introduce a baseline + +baseline + REINFORCE + TRPO + PPO + GAE + NAF + DPG -ACKTR \ No newline at end of file + +ACKTR \ No newline at end of file