Object oriented optimization
This library provides object oriented optimization. This allows...
1. using theoretic values (such as the strong convexity parameter)
2. object-oriented definitions, both for models and optimization algorithms. This allows...
* interacting with the optimization as an object. Want to compute some
value partway through? Want to change the values as time goes on?
* getting results intermediately (or in the presence of a keyboard
* having callbacks, etc
A typical example:
model = Model()
opt = SGD(model.loss)
data = 
for _ in range(10):
data += [get_stats(model)]
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.