def test_fit_simple_linear_model(self):
seed = 0x2019
np.random.seed(seed)
tf.random.set_seed(seed)
num_examples = 50000
x = np.random.standard_normal((num_examples, 3))
w = np.random.standard_normal((3, 1))
y = np.dot(x, w) + np.random.standard_normal((num_examples, 1)) * 1e-4
model = tf.keras.models.Sequential()
model.add(tf.keras.layers.Dense(input_shape=(3, ), units=1))
self.evaluate(tf.compat.v1.global_variables_initializer())
opt = MovingAverage('adam')
model.compile(opt, loss='mse')
model.fit(x, y, epochs=10)
opt.assign_average_vars(model.variables)
x = np.random.standard_normal((100, 3))
y = np.dot(x, w)
predicted = model.predict(x)
max_abs_diff = np.max(np.abs(predicted - y))
self.assertLess(max_abs_diff, 1e-3)
def test_fit_simple_linear_model():
seed = 0x2019
np.random.seed(seed)
tf.random.set_seed(seed)
num_examples = 5000
x = np.random.standard_normal((num_examples, 3))
w = np.random.standard_normal((3, 1))
y = np.dot(x, w) + np.random.standard_normal((num_examples, 1)) * 1e-4
model = tf.keras.models.Sequential()
model.add(tf.keras.layers.Dense(input_shape=(3, ), units=1))
opt = MovingAverage("sgd")
model.compile(opt, loss="mse")
model.fit(x, y, epochs=5)
opt.assign_average_vars(model.variables)
x = np.random.standard_normal((100, 3))
y = np.dot(x, w)
predicted = model.predict(x)
max_abs_diff = np.max(np.abs(predicted - y))
assert max_abs_diff < 5e-3
def test_no_average_slot():
max_features = 5000
max_len = 4
embedding_dims = 2
# Some preprocessing layers have TrackableWeightHandler.
# They are returned when using model.variables
# but it's unable to assign average slot to them.
vectorize_layer = tf.keras.layers.experimental.preprocessing.TextVectorization(
max_tokens=max_features, output_mode="int", output_sequence_length=max_len
)
vectorize_layer.adapt(["foo", "bar", "baz"])
model = tf.keras.models.Sequential(
[
tf.keras.Input(shape=(1,), dtype=tf.string),
vectorize_layer,
tf.keras.layers.Embedding(max_features + 1, embedding_dims),
tf.keras.layers.Dense(1),
]
)
optimizer = MovingAverage("sgd")
model.compile(optimizer, loss="mse")
model.fit(x=["foo", "bar", "baz"], y=[0.0, 1.0, 2.0], epochs=1)
optimizer.assign_average_vars(model.variables)
def test_run(self):
for sequential_update in [True, False]:
var0 = tf.Variable([1.0, 2.0])
var1 = tf.Variable([3.0, 4.0])
grads0 = tf.constant([0.1, 0.1])
grads1 = tf.constant([0.01, 0.01])
grads_and_vars = list(zip([grads0, grads1], [var0, var1]))
self.evaluate(tf.compat.v1.global_variables_initializer())
opt = MovingAverage(
tf.keras.optimizers.SGD(lr=2.0),
sequential_update=sequential_update,
average_decay=0.5,
)
if not tf.executing_eagerly():
update = opt.apply_gradients(grads_and_vars)
self.evaluate(tf.compat.v1.global_variables_initializer())
self.evaluate(update)
self.evaluate(update)
else:
opt.apply_gradients(grads_and_vars)
opt.apply_gradients(grads_and_vars)
self.assertAllClose(var0.read_value(), [0.6, 1.6])
self.assertAllClose(var1.read_value(), [2.96, 3.96])
ema_var0 = opt.get_slot(var0, "average")
ema_var1 = opt.get_slot(var1, "average")
if sequential_update:
self.assertAllClose(ema_var0.read_value(), [0.75, 1.75])
self.assertAllClose(ema_var1.read_value(), [2.975, 3.975])
assign = opt.assign_average_vars([var0, var1])
self.evaluate(assign)
if sequential_update:
self.assertAllClose(var0.read_value(), [0.75, 1.75])
self.assertAllClose(var1.read_value(), [2.975, 3.975])
perturb = tf.group([
var0.assign_add([1.0, 1.0]),
var1.assign_add([2.0, 2.0]),
ema_var0.assign_add([3.0, 3.0]),
ema_var1.assign_add([4.0, 4.0]),
])
self.evaluate(perturb)
if sequential_update:
self.assertAllClose(var0.read_value(), [1.75, 2.75])
self.assertAllClose(var1.read_value(), [4.975, 5.975])
self.assertAllClose(ema_var0.read_value(), [3.75, 4.75])
self.assertAllClose(ema_var1.read_value(), [6.975, 7.975])
def test_run(self):
self.skipTest(
"Wait for https://github.com/tensorflow/tensorflow/issues/31582")
for sequential_update in [True, False]:
var0 = tf.Variable([1.0, 2.0])
var1 = tf.Variable([3.0, 4.0])
grads0 = tf.constant([0.1, 0.1])
grads1 = tf.constant([0.01, 0.01])
grads_and_vars = list(zip([grads0, grads1], [var0, var1]))
opt = MovingAverage(
tf.keras.optimizers.SGD(lr=2.0),
average_decay=0.5,
sequential_update=sequential_update)
if not tf.executing_eagerly():
update = opt.apply_gradients(grads_and_vars)
self.evaluate(tf.compat.v1.global_variables_initializer())
self.evaluate(update)
self.evaluate(update)
else:
opt.apply_gradients(grads_and_vars)
opt.apply_gradients(grads_and_vars)
self.assertAllClose(var0.read_value(), [0.6, 1.6])
self.assertAllClose(var1.read_value(), [2.96, 3.96])
ema_var0 = opt._ema.average(var0) # pylint: disable=protected-access
ema_var1 = opt._ema.average(var1) # pylint: disable=protected-access
if sequential_update:
self.assertAllClose(ema_var0.read_value(), [0.75, 1.75])
self.assertAllClose(ema_var1.read_value(), [2.975, 3.975])
assign = opt.assign_average_vars([var0, var1])
self.evaluate(assign)
if sequential_update:
self.assertAllClose(var0.read_value(), [0.75, 1.75])
self.assertAllClose(var1.read_value(), [2.975, 3.975])
perturb = tf.group([
var0.assign_add([1.0, 1.0]),
var1.assign_add([2.0, 2.0]),
ema_var0.assign_add([3.0, 3.0]),
ema_var1.assign_add([4.0, 4.0])
])
self.evaluate(perturb)
if sequential_update:
self.assertAllClose(var0.read_value(), [1.75, 2.75])
self.assertAllClose(var1.read_value(), [4.975, 5.975])
self.assertAllClose(ema_var0.read_value(), [3.75, 4.75])
self.assertAllClose(ema_var1.read_value(), [6.975, 7.975])
def test_model_weights_update():
grad = tf.Variable([[0.1]])
model = tf.keras.Sequential([
tf.keras.layers.Dense(
1,
kernel_initializer=tf.keras.initializers.Constant([[1.0]]),
use_bias=False,
)
])
model.build(input_shape=[1, 1])
opt = MovingAverage(tf.keras.optimizers.SGD(lr=2.0), average_decay=0.5)
_ = opt.apply_gradients(list(zip([grad], model.variables)))
np.testing.assert_allclose(model.variables[0].read_value(), [[0.8]])
_ = opt.assign_average_vars(model.variables)
np.testing.assert_allclose(model.variables[0].read_value(), [[0.9]])
def test_run(sequential_update):
var0 = tf.Variable([1.0, 2.0])
var1 = tf.Variable([3.0, 4.0])
grads0 = tf.constant([0.1, 0.1])
grads1 = tf.constant([0.01, 0.01])
grads_and_vars = list(zip([grads0, grads1], [var0, var1]))
opt = MovingAverage(
tf.keras.optimizers.SGD(lr=2.0),
sequential_update=sequential_update,
average_decay=0.5,
)
opt.apply_gradients(grads_and_vars)
opt.apply_gradients(grads_and_vars)
np.testing.assert_allclose(var0.read_value(), [0.6, 1.6])
np.testing.assert_allclose(var1.read_value(), [2.96, 3.96])
ema_var0 = opt.get_slot(var0, "average")
ema_var1 = opt.get_slot(var1, "average")
if sequential_update:
np.testing.assert_allclose(ema_var0.read_value(), [0.75, 1.75])
np.testing.assert_allclose(ema_var1.read_value(), [2.975, 3.975])
_ = opt.assign_average_vars([var0, var1])
if sequential_update:
np.testing.assert_allclose(var0.read_value(), [0.75, 1.75])
np.testing.assert_allclose(var1.read_value(), [2.975, 3.975])
var0.assign_add([1.0, 1.0]),
var1.assign_add([2.0, 2.0]),
ema_var0.assign_add([3.0, 3.0]),
ema_var1.assign_add([4.0, 4.0]),
if sequential_update:
np.testing.assert_allclose(var0.read_value(), [1.75, 2.75])
np.testing.assert_allclose(var1.read_value(), [4.975, 5.975])
np.testing.assert_allclose(ema_var0.read_value(), [3.75, 4.75])
np.testing.assert_allclose(ema_var1.read_value(), [6.975, 7.975])
def test_model_weights_update(self):
grad = tf.Variable([[0.1]])
model = tf.keras.Sequential([
tf.keras.layers.Dense(
1,
kernel_initializer=tf.keras.initializers.Constant([[1.0]]),
use_bias=False)
])
model.build(input_shape=[1, 1])
opt = MovingAverage(tf.keras.optimizers.SGD(lr=2.0), 0.5)
update = opt.apply_gradients(list(zip([grad], model.variables)))
self.evaluate(tf.compat.v1.global_variables_initializer())
self.evaluate(update)
self.assertAllClose(model.variables[0].read_value(), [[0.8]])
mean_update = opt.assign_average_vars(model.variables)
self.evaluate(mean_update)
self.assertAllClose(model.variables[0].read_value(), [[0.9]])
