R/optimizers.R
momentum_step.Rd
Step for SGD with momentum with `lr`
momentum_step(p, lr, grad_avg, ...)
p
learning rate
grad average
additional arguments to pass
None