Skip to content

added learning rate bias correction and corrected m_t #15

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
dchatterjee172 wants to merge 1 commit into
base: master
Choose a base branch
from
Open

added learning rate bias correction and corrected m_t #15

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 9 additions & 8 deletions tf_utils/adamax.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,14 +8,12 @@

class AdamaxOptimizer(optimizer.Optimizer):
"""Optimizer that implements the Adamax algorithm.

See [Kingma et. al., 2014](http://arxiv.org/abs/1412.6980)
([pdf](http://arxiv.org/pdf/1412.6980.pdf)).

@@__init__
"""

def __init__(self, learning_rate=0.001, beta1=0.9, beta2=0.999, use_locking=False, name="Adamax"):
def __init__(self, learning_rate=0.002, beta1=0.9, beta2=0.999, use_locking=False, name="Adamax"):
super(AdamaxOptimizer, self).__init__(use_locking, name)
self._lr = learning_rate
self._beta1 = beta1
Expand All @@ -36,6 +34,7 @@ def _create_slots(self, var_list):
for v in var_list:
self._zeros_slot(v, "m", self._name)
self._zeros_slot(v, "v", self._name)
self._zeros_slot(v, "t", self._name)

def _apply_dense(self, grad, var):
lr_t = math_ops.cast(self._lr_t, var.dtype.base_dtype)
Expand All @@ -45,15 +44,17 @@ def _apply_dense(self, grad, var):
eps = 1e-7 # Can't use 1e-8 due to underflow -- not sure if it makes a big difference.
else:
eps = 1e-8

v = self.get_slot(var, "v")
v_t = v.assign(beta1_t * v + (1. - beta1_t) * grad)
m = self.get_slot(var, "m")
m_t = m.assign(tf.maximum(beta2_t * m + eps, tf.abs(grad)))
g_t = v_t / m_t
m_t = m.assign(tf.maximum(beta2_t * m,tf.abs(grad)+eps))
t=self.get_slot(var,"t")
t_t=t.assign(t+1)
g_t = lr_t/(1-beta1_t**t_t)*v_t/m_t

var_update = state_ops.assign_sub(var, lr_t * g_t)
return control_flow_ops.group(*[var_update, m_t, v_t])
var_update = state_ops.assign_sub(var,g_t)
return control_flow_ops.group(*[var_update, t_t ,m_t, v_t])

def _apply_sparse(self, grad, var):
raise NotImplementedError("Sparse gradient updates are not supported.")