6e8976428d
This fixes some common incompatibilities with eager mode execution. PiperOrigin-RevId: 319258451 Change-Id: I93c66eb3b8c75f75fd9c1deb9526fbd937b93805
182 lines
7.2 KiB
Python
182 lines
7.2 KiB
Python
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
# ==============================================================================
|
|
"""Tests for Nadam."""
|
|
|
|
from __future__ import absolute_import
|
|
from __future__ import division
|
|
from __future__ import print_function
|
|
|
|
import numpy as np
|
|
|
|
from tensorflow.python.framework import constant_op
|
|
from tensorflow.python.framework import dtypes
|
|
from tensorflow.python.framework import ops
|
|
from tensorflow.python.keras.optimizer_v2 import nadam
|
|
from tensorflow.python.ops import math_ops
|
|
from tensorflow.python.ops import variables
|
|
from tensorflow.python.platform import test
|
|
|
|
|
|
def get_beta_accumulators(opt, dtype):
|
|
local_step = math_ops.cast(opt.iterations + 1, dtype)
|
|
beta_1_t = math_ops.cast(opt._get_hyper("beta_1"), dtype)
|
|
beta_1_power = math_ops.pow(beta_1_t, local_step)
|
|
beta_2_t = math_ops.cast(opt._get_hyper("beta_2"), dtype)
|
|
beta_2_power = math_ops.pow(beta_2_t, local_step)
|
|
return (beta_1_power, beta_2_power)
|
|
|
|
|
|
def update_m_cache(m_cache, t, beta1=0.9):
|
|
mu_t = beta1 * (1 - 0.5 * 0.96**(0.004 * (t + 1)))
|
|
m_cache_t = m_cache * mu_t
|
|
return m_cache_t
|
|
|
|
|
|
def nadam_update_numpy(param,
|
|
g_t,
|
|
t,
|
|
m,
|
|
v,
|
|
m_cache,
|
|
alpha=0.001,
|
|
beta1=0.9,
|
|
beta2=0.999,
|
|
epsilon=1e-8):
|
|
|
|
mu_t = beta1 * (1 - 0.5 * 0.96**(0.004 * (t + 1)))
|
|
mu_t_1 = beta1 * (1 - 0.5 * 0.96**(0.004 * (t + 2)))
|
|
m_cache_t_1 = m_cache * mu_t_1
|
|
g_prime_t = g_t / (1 - m_cache)
|
|
m_t = beta1 * m + (1 - beta1) * g_t
|
|
v_t = beta2 * v + (1 - beta2) * g_t * g_t
|
|
|
|
m_prime_t = m_t / (1 - m_cache_t_1)
|
|
v_prime_t = v_t / (1 - beta2**(t + 1))
|
|
m_bar_t = (1 - mu_t) * g_prime_t + mu_t_1 * m_prime_t
|
|
|
|
param_t = param - alpha * m_bar_t / (np.sqrt(v_prime_t) + epsilon)
|
|
return param_t, m_t, v_t
|
|
|
|
|
|
class NadamOptimizerTest(test.TestCase):
|
|
|
|
def testSparse(self):
|
|
# TODO(tanzheny, omalleyt): Fix test in eager mode.
|
|
sparse_epsilon = 1e-7
|
|
for dtype in [dtypes.half, dtypes.float32, dtypes.float64]:
|
|
with ops.Graph().as_default(), self.cached_session():
|
|
# Initialize variables for numpy implementation.
|
|
m0, v0, m1, v1, mcache = 0.0, 0.0, 0.0, 0.0, 1.0
|
|
var0_np = np.array([1.0, 1.0, 2.0], dtype=dtype.as_numpy_dtype)
|
|
grads0_np = np.array([0.1, 0, 0.1], dtype=dtype.as_numpy_dtype)
|
|
var1_np = np.array([3.0, 3.0, 4.0], dtype=dtype.as_numpy_dtype)
|
|
grads1_np = np.array([0.01, 0, 0.01], dtype=dtype.as_numpy_dtype)
|
|
|
|
var0 = variables.Variable(var0_np)
|
|
var1 = variables.Variable(var1_np)
|
|
grads0_np_indices = np.array([0, 2], dtype=np.int32)
|
|
grads0 = ops.IndexedSlices(
|
|
constant_op.constant(grads0_np[grads0_np_indices]),
|
|
constant_op.constant(grads0_np_indices), constant_op.constant([3]))
|
|
grads1_np_indices = np.array([0, 2], dtype=np.int32)
|
|
grads1 = ops.IndexedSlices(
|
|
constant_op.constant(grads1_np[grads1_np_indices]),
|
|
constant_op.constant(grads1_np_indices), constant_op.constant([3]))
|
|
opt = nadam.Nadam(epsilon=sparse_epsilon)
|
|
update = opt.apply_gradients(zip([grads0, grads1], [var0, var1]))
|
|
self.evaluate(variables.global_variables_initializer())
|
|
|
|
# Fetch params to validate initial values
|
|
self.assertAllClose([1.0, 1.0, 2.0], var0)
|
|
self.assertAllClose([3.0, 3.0, 4.0], var1)
|
|
|
|
beta1_power, beta2_power = get_beta_accumulators(opt, dtype)
|
|
|
|
# Run 3 steps of Nadam
|
|
for t in range(3):
|
|
self.assertAllCloseAccordingToType(0.9**(t + 1), beta1_power)
|
|
self.assertAllCloseAccordingToType(0.999**(t + 1), beta2_power)
|
|
update.run()
|
|
|
|
mcache = update_m_cache(mcache, t)
|
|
var0_np, m0, v0 = nadam_update_numpy(
|
|
var0_np, grads0_np, t, m0, v0, mcache, epsilon=sparse_epsilon)
|
|
var1_np, m1, v1 = nadam_update_numpy(
|
|
var1_np, grads1_np, t, m1, v1, mcache, epsilon=sparse_epsilon)
|
|
|
|
# Validate updated params
|
|
self.assertAllCloseAccordingToType(var0_np, var0)
|
|
self.assertAllCloseAccordingToType(var1_np, var1)
|
|
|
|
def testBasic(self):
|
|
# TODO(tanzheny, omalleyt): Fix test in eager mode.
|
|
for dtype in [dtypes.half, dtypes.float32, dtypes.float64]:
|
|
with ops.Graph().as_default(), self.cached_session():
|
|
# Initialize variables for numpy implementation.
|
|
m0, v0, m1, v1, mcache = 0.0, 0.0, 0.0, 0.0, 1.0
|
|
var0_np = np.array([1.0, 2.0], dtype=dtype.as_numpy_dtype)
|
|
grads0_np = np.array([0.1, 0.1], dtype=dtype.as_numpy_dtype)
|
|
var1_np = np.array([3.0, 4.0], dtype=dtype.as_numpy_dtype)
|
|
grads1_np = np.array([0.01, 0.01], dtype=dtype.as_numpy_dtype)
|
|
|
|
var0 = variables.Variable(var0_np)
|
|
var1 = variables.Variable(var1_np)
|
|
grads0 = constant_op.constant(grads0_np)
|
|
grads1 = constant_op.constant(grads1_np)
|
|
opt = nadam.Nadam()
|
|
update = opt.apply_gradients(zip([grads0, grads1], [var0, var1]))
|
|
self.evaluate(variables.global_variables_initializer())
|
|
|
|
# Fetch params to validate initial values
|
|
self.assertAllClose([1.0, 2.0], var0)
|
|
self.assertAllClose([3.0, 4.0], var1)
|
|
|
|
# Run 3 steps of Nadam
|
|
for t in range(3):
|
|
update.run()
|
|
|
|
mcache = update_m_cache(mcache, t)
|
|
var0_np, m0, v0 = nadam_update_numpy(var0_np, grads0_np, t, m0, v0,
|
|
mcache)
|
|
var1_np, m1, v1 = nadam_update_numpy(var1_np, grads1_np, t, m1, v1,
|
|
mcache)
|
|
|
|
# Validate updated params
|
|
self.assertAllCloseAccordingToType(var0_np, var0)
|
|
self.assertAllCloseAccordingToType(var1_np, var1)
|
|
|
|
def testConstructNAdamWithLR(self):
|
|
opt = nadam.Nadam(lr=1.0)
|
|
opt_2 = nadam.Nadam(learning_rate=0.1, lr=1.0)
|
|
opt_3 = nadam.Nadam(learning_rate=0.1)
|
|
self.assertIsInstance(opt.lr, variables.Variable)
|
|
self.assertIsInstance(opt_2.lr, variables.Variable)
|
|
self.assertIsInstance(opt_3.lr, variables.Variable)
|
|
|
|
self.evaluate(variables.global_variables_initializer())
|
|
self.assertAllClose(self.evaluate(opt.lr), (1.0))
|
|
self.assertAllClose(self.evaluate(opt_2.lr), (1.0))
|
|
self.assertAllClose(self.evaluate(opt_3.lr), (0.1))
|
|
|
|
def testConstructNAdamWithScheduleDecay(self):
|
|
opt = nadam.Nadam(schedule_decay=0.2)
|
|
self.assertIsInstance(opt.decay, variables.Variable)
|
|
self.evaluate(variables.global_variables_initializer())
|
|
self.assertAllClose(self.evaluate(opt.decay), (0.2))
|
|
|
|
|
|
if __name__ == "__main__":
|
|
test.main()
|