pinocchio  3.9.0
A fast and flexible implementation of Rigid Body Dynamics algorithms and their analytical derivatives
Loading...
Searching...
No Matches
QValueNetwork Class Reference

Public Member Functions

 __init__ (self)
 setupOptim (self)
 setupTargetAssign (self, nominalNet, tau=UPDATE_RATE)

Public Attributes

tuple gradient
list hidens = [netx1, netx2, netu1, netu2]
 optim = optim
 qref = qref
 qvalue = qvalue
 u = u
list update_variables
 variables = tf.trainable_variables()[nvars:]
 x = x

Detailed Description

Definition at line 47 of file continuous.py.

Constructor & Destructor Documentation

◆ __init__()

__init__(self)

Definition at line 48 of file continuous.py.

Member Function Documentation

◆ setupOptim()

setupOptim(self)

Definition at line 70 of file continuous.py.

◆ setupTargetAssign()

setupTargetAssign(self,
nominalNet,
tau = UPDATE_RATE )

Definition at line 84 of file continuous.py.

Member Data Documentation

◆ gradient

tuple gradient
Initial value:
= (
# Gradient of Q wrt the control dQ/du (for policy training)
gradient
)

Definition at line 78 of file continuous.py.

◆ hidens

list hidens = [netx1, netx2, netu1, netu2]

Definition at line 68 of file continuous.py.

◆ optim

optim = optim

Definition at line 77 of file continuous.py.

◆ qref

qref = qref

Definition at line 76 of file continuous.py.

◆ qvalue

qvalue = qvalue

Definition at line 65 of file continuous.py.

◆ u

u = u

Definition at line 64 of file continuous.py.

◆ update_variables

list update_variables
Initial value:
= [
target.assign(tau * ref + (1 - tau) * target)
for target, ref in zip(self.variables, nominalNet.variables)
]

Definition at line 85 of file continuous.py.

◆ variables

variables = tf.trainable_variables()[nvars:]

Definition at line 67 of file continuous.py.

◆ x

x = x

Definition at line 63 of file continuous.py.

The documentation for this class was generated from the following file: