289 lines
8.6 KiB
Python
289 lines
8.6 KiB
Python
"""Stochastic optimization methods for MLP
|
|
"""
|
|
|
|
# Authors: Jiyuan Qian <jq401@nyu.edu>
|
|
# License: BSD 3 clause
|
|
|
|
import numpy as np
|
|
|
|
|
|
class BaseOptimizer:
|
|
"""Base (Stochastic) gradient descent optimizer
|
|
|
|
Parameters
|
|
----------
|
|
learning_rate_init : float, default=0.1
|
|
The initial learning rate used. It controls the step-size in updating
|
|
the weights
|
|
|
|
Attributes
|
|
----------
|
|
learning_rate : float
|
|
the current learning rate
|
|
"""
|
|
|
|
def __init__(self, learning_rate_init=0.1):
|
|
self.learning_rate_init = learning_rate_init
|
|
self.learning_rate = float(learning_rate_init)
|
|
|
|
def update_params(self, params, grads):
|
|
"""Update parameters with given gradients
|
|
|
|
Parameters
|
|
----------
|
|
params : list of length = len(coefs_) + len(intercepts_)
|
|
The concatenated list containing coefs_ and intercepts_ in MLP
|
|
model. Used for initializing velocities and updating params
|
|
|
|
grads : list of length = len(params)
|
|
Containing gradients with respect to coefs_ and intercepts_ in MLP
|
|
model. So length should be aligned with params
|
|
"""
|
|
updates = self._get_updates(grads)
|
|
for param, update in zip((p for p in params), updates):
|
|
param += update
|
|
|
|
def iteration_ends(self, time_step):
|
|
"""Perform update to learning rate and potentially other states at the
|
|
end of an iteration
|
|
"""
|
|
pass
|
|
|
|
def trigger_stopping(self, msg, verbose):
|
|
"""Decides whether it is time to stop training
|
|
|
|
Parameters
|
|
----------
|
|
msg : str
|
|
Message passed in for verbose output
|
|
|
|
verbose : bool
|
|
Print message to stdin if True
|
|
|
|
Returns
|
|
-------
|
|
is_stopping : bool
|
|
True if training needs to stop
|
|
"""
|
|
if verbose:
|
|
print(msg + " Stopping.")
|
|
return True
|
|
|
|
|
|
class SGDOptimizer(BaseOptimizer):
|
|
"""Stochastic gradient descent optimizer with momentum
|
|
|
|
Parameters
|
|
----------
|
|
params : list, length = len(coefs_) + len(intercepts_)
|
|
The concatenated list containing coefs_ and intercepts_ in MLP model.
|
|
Used for initializing velocities and updating params
|
|
|
|
learning_rate_init : float, default=0.1
|
|
The initial learning rate used. It controls the step-size in updating
|
|
the weights
|
|
|
|
lr_schedule : {'constant', 'adaptive', 'invscaling'}, default='constant'
|
|
Learning rate schedule for weight updates.
|
|
|
|
-'constant', is a constant learning rate given by
|
|
'learning_rate_init'.
|
|
|
|
-'invscaling' gradually decreases the learning rate 'learning_rate_' at
|
|
each time step 't' using an inverse scaling exponent of 'power_t'.
|
|
learning_rate_ = learning_rate_init / pow(t, power_t)
|
|
|
|
-'adaptive', keeps the learning rate constant to
|
|
'learning_rate_init' as long as the training keeps decreasing.
|
|
Each time 2 consecutive epochs fail to decrease the training loss by
|
|
tol, or fail to increase validation score by tol if 'early_stopping'
|
|
is on, the current learning rate is divided by 5.
|
|
|
|
momentum : float, default=0.9
|
|
Value of momentum used, must be larger than or equal to 0
|
|
|
|
nesterov : bool, default=True
|
|
Whether to use nesterov's momentum or not. Use nesterov's if True
|
|
|
|
power_t : float, default=0.5
|
|
Power of time step 't' in inverse scaling. See `lr_schedule` for
|
|
more details.
|
|
|
|
Attributes
|
|
----------
|
|
learning_rate : float
|
|
the current learning rate
|
|
|
|
velocities : list, length = len(params)
|
|
velocities that are used to update params
|
|
"""
|
|
|
|
def __init__(
|
|
self,
|
|
params,
|
|
learning_rate_init=0.1,
|
|
lr_schedule="constant",
|
|
momentum=0.9,
|
|
nesterov=True,
|
|
power_t=0.5,
|
|
):
|
|
super().__init__(learning_rate_init)
|
|
|
|
self.lr_schedule = lr_schedule
|
|
self.momentum = momentum
|
|
self.nesterov = nesterov
|
|
self.power_t = power_t
|
|
self.velocities = [np.zeros_like(param) for param in params]
|
|
|
|
def iteration_ends(self, time_step):
|
|
"""Perform updates to learning rate and potential other states at the
|
|
end of an iteration
|
|
|
|
Parameters
|
|
----------
|
|
time_step : int
|
|
number of training samples trained on so far, used to update
|
|
learning rate for 'invscaling'
|
|
"""
|
|
if self.lr_schedule == "invscaling":
|
|
self.learning_rate = (
|
|
float(self.learning_rate_init) / (time_step + 1) ** self.power_t
|
|
)
|
|
|
|
def trigger_stopping(self, msg, verbose):
|
|
if self.lr_schedule != "adaptive":
|
|
if verbose:
|
|
print(msg + " Stopping.")
|
|
return True
|
|
|
|
if self.learning_rate <= 1e-6:
|
|
if verbose:
|
|
print(msg + " Learning rate too small. Stopping.")
|
|
return True
|
|
|
|
self.learning_rate /= 5.0
|
|
if verbose:
|
|
print(msg + " Setting learning rate to %f" % self.learning_rate)
|
|
return False
|
|
|
|
def _get_updates(self, grads):
|
|
"""Get the values used to update params with given gradients
|
|
|
|
Parameters
|
|
----------
|
|
grads : list, length = len(coefs_) + len(intercepts_)
|
|
Containing gradients with respect to coefs_ and intercepts_ in MLP
|
|
model. So length should be aligned with params
|
|
|
|
Returns
|
|
-------
|
|
updates : list, length = len(grads)
|
|
The values to add to params
|
|
"""
|
|
updates = [
|
|
self.momentum * velocity - self.learning_rate * grad
|
|
for velocity, grad in zip(self.velocities, grads)
|
|
]
|
|
self.velocities = updates
|
|
|
|
if self.nesterov:
|
|
updates = [
|
|
self.momentum * velocity - self.learning_rate * grad
|
|
for velocity, grad in zip(self.velocities, grads)
|
|
]
|
|
|
|
return updates
|
|
|
|
|
|
class AdamOptimizer(BaseOptimizer):
|
|
"""Stochastic gradient descent optimizer with Adam
|
|
|
|
Note: All default values are from the original Adam paper
|
|
|
|
Parameters
|
|
----------
|
|
params : list, length = len(coefs_) + len(intercepts_)
|
|
The concatenated list containing coefs_ and intercepts_ in MLP model.
|
|
Used for initializing velocities and updating params
|
|
|
|
learning_rate_init : float, default=0.001
|
|
The initial learning rate used. It controls the step-size in updating
|
|
the weights
|
|
|
|
beta_1 : float, default=0.9
|
|
Exponential decay rate for estimates of first moment vector, should be
|
|
in [0, 1)
|
|
|
|
beta_2 : float, default=0.999
|
|
Exponential decay rate for estimates of second moment vector, should be
|
|
in [0, 1)
|
|
|
|
epsilon : float, default=1e-8
|
|
Value for numerical stability
|
|
|
|
Attributes
|
|
----------
|
|
learning_rate : float
|
|
The current learning rate
|
|
|
|
t : int
|
|
Timestep
|
|
|
|
ms : list, length = len(params)
|
|
First moment vectors
|
|
|
|
vs : list, length = len(params)
|
|
Second moment vectors
|
|
|
|
References
|
|
----------
|
|
:arxiv:`Kingma, Diederik, and Jimmy Ba (2014) "Adam: A method for
|
|
stochastic optimization." <1412.6980>
|
|
"""
|
|
|
|
def __init__(
|
|
self, params, learning_rate_init=0.001, beta_1=0.9, beta_2=0.999, epsilon=1e-8
|
|
):
|
|
super().__init__(learning_rate_init)
|
|
|
|
self.beta_1 = beta_1
|
|
self.beta_2 = beta_2
|
|
self.epsilon = epsilon
|
|
self.t = 0
|
|
self.ms = [np.zeros_like(param) for param in params]
|
|
self.vs = [np.zeros_like(param) for param in params]
|
|
|
|
def _get_updates(self, grads):
|
|
"""Get the values used to update params with given gradients
|
|
|
|
Parameters
|
|
----------
|
|
grads : list, length = len(coefs_) + len(intercepts_)
|
|
Containing gradients with respect to coefs_ and intercepts_ in MLP
|
|
model. So length should be aligned with params
|
|
|
|
Returns
|
|
-------
|
|
updates : list, length = len(grads)
|
|
The values to add to params
|
|
"""
|
|
self.t += 1
|
|
self.ms = [
|
|
self.beta_1 * m + (1 - self.beta_1) * grad
|
|
for m, grad in zip(self.ms, grads)
|
|
]
|
|
self.vs = [
|
|
self.beta_2 * v + (1 - self.beta_2) * (grad**2)
|
|
for v, grad in zip(self.vs, grads)
|
|
]
|
|
self.learning_rate = (
|
|
self.learning_rate_init
|
|
* np.sqrt(1 - self.beta_2**self.t)
|
|
/ (1 - self.beta_1**self.t)
|
|
)
|
|
updates = [
|
|
-self.learning_rate * m / (np.sqrt(v) + self.epsilon)
|
|
for m, v in zip(self.ms, self.vs)
|
|
]
|
|
return updates
|