rk4_step_backwards {tfNeuralODE} | R Documentation |
Custom internal RK4 solver for solving the backward pass of the Neural ODE.
Description
Custom internal RK4 solver for solving the backward pass of the Neural ODE.
Usage
rk4_step_backwards(backward_dynamics, dt, state, model)
Arguments
backward_dynamics |
The backward dynamics function for the Neural ODE. |
dt |
The time step to solve the ODE on. |
state |
The current state of the differential equation. |
model |
The neural network that defines the Neural ODE. |
Value
An output list with the updated backwards state.
[Package tfNeuralODE version 0.1.0 Index]