| Safe Haskell | None |
|---|---|
| Language | Haskell2010 |
RPB.MEM
Description
PPO Style Replay Memory and Memory Loaders
Synopsis
- data Buffer a = Buffer {}
- mkBuffer :: Buffer Tensor
- empty :: Buffer Tensor
- size :: Buffer Tensor -> Int
- push :: Buffer Tensor -> Tensor -> Tensor -> Tensor -> Tensor -> Tensor -> Tensor -> Buffer Tensor
- push' :: Buffer Tensor -> Buffer Tensor -> Buffer Tensor
- gae :: Tensor -> Tensor -> Tensor -> Tensor -> Tensor -> Tensor -> Tensor
- data Loader a = Loader {
- states' :: !a
- actions' :: !a
- logProbs' :: !a
- returns' :: !a
- advantages' :: !a
- mkLoader :: Buffer Tensor -> Int -> Tensor -> Tensor -> Loader [Tensor]
- size' :: Loader [Tensor] -> Int
Documentation
Replay Memory
Constructors
| Buffer | |
push :: Buffer Tensor -> Tensor -> Tensor -> Tensor -> Tensor -> Tensor -> Tensor -> Buffer Tensor Source #
Push new memories into Buffer
push' :: Buffer Tensor -> Buffer Tensor -> Buffer Tensor Source #
Pushing one buffer into another one
gae :: Tensor -> Tensor -> Tensor -> Tensor -> Tensor -> Tensor -> Tensor Source #
Generalized Advantage Estimate
Memory Data Loader
Constructors
| Loader | |
Fields
| |