def testPixelControlLossShapes(self):
with self.assertRaisesRegexp(
ValueError, "Pixel Control values are not compatible"):
pixel_control_ops.pixel_control_loss(
self.observations_ph, self.actions_ph,
self.action_values_ph[:, :, :-1], self.cell_size,
self.discount, self.scale)
def testPixelControlLossShapes(self):
with self.assertRaisesRegexp(
ValueError, "Pixel Control values are not compatible"):
pixel_control_ops.pixel_control_loss(
self.observations_ph, self.actions_ph,
self.action_values_ph[:, :, :-1], self.cell_size, self.discount,
self.scale)
def testTensorDiscountShape(self):
with self.assertRaisesRegexp(
ValueError, "discount_factor must be a scalar or a tensor of rank 2"):
tensor_discount = tf.tile(
tf.reshape(self.discount, [1, 1, 1]),
[self.seq_length, self.batch_size, 1])
pixel_control_ops.pixel_control_loss(
self.observations_ph, self.actions_ph,
self.action_values_ph, self.cell_size, tensor_discount,
self.scale)
def testTensorDiscountShape(self):
with self.assertRaisesRegexp(
ValueError,
"discount_factor must be a scalar or a tensor of rank 2"):
tensor_discount = tf.tile(tf.reshape(self.discount, [1, 1, 1]),
[self.seq_length, self.batch_size, 1])
pixel_control_ops.pixel_control_loss(self.observations_ph,
self.actions_ph,
self.action_values_ph,
self.cell_size,
tensor_discount, self.scale)
def testPixelControlLossScalarDiscount(self):
"""Compute loss for given observations, actions, values, scalar discount."""
loss, _ = pixel_control_ops.pixel_control_loss(
self.observations_ph, self.actions_ph, self.action_values_ph,
self.cell_size, self.discount, self.scale)
init = tf.global_variables_initializer()
with self.test_session() as sess:
sess.run(init)
feed_dict = {
self.observations_ph: self.observations,
self.action_values_ph: self.action_values,
self.actions_ph: self.actions}
loss_np = sess.run(loss, feed_dict=feed_dict)
self.assertNear(loss_np, self.error, 1e-3)
def testPixelControlLossScalarDiscount(self):
"""Compute loss for given observations, actions, values, scalar discount."""
loss, _ = pixel_control_ops.pixel_control_loss(
self.observations_ph, self.actions_ph, self.action_values_ph,
self.cell_size, self.discount, self.scale)
init = tf.global_variables_initializer()
with self.test_session() as sess:
sess.run(init)
feed_dict = {
self.observations_ph: self.observations,
self.action_values_ph: self.action_values,
self.actions_ph: self.actions
}
loss_np = sess.run(loss, feed_dict=feed_dict)
self.assertNear(loss_np, self.error, 1e-3)
def testPixelControlLossTensorDiscount(self):
"""Compute loss for given observations, actions, values, tensor discount."""
zero_discount = tf.zeros((1, self.batch_size))
non_zero_discount = tf.tile(
tf.reshape(self.discount, [1, 1]),
[self.seq_length - 1, self.batch_size])
tensor_discount = tf.concat([zero_discount, non_zero_discount], axis=0)
loss, _ = pixel_control_ops.pixel_control_loss(
self.observations_ph, self.actions_ph, self.action_values_ph,
self.cell_size, tensor_discount, self.scale)
init = tf.global_variables_initializer()
with self.test_session() as sess:
sess.run(init)
feed_dict = {
self.observations_ph: self.observations,
self.action_values_ph: self.action_values,
self.actions_ph: self.actions}
loss_np = sess.run(loss, feed_dict=feed_dict)
self.assertNear(loss_np, self.error_term, 1e-3)
def testPixelControlLossTensorDiscount(self):
"""Compute loss for given observations, actions, values, tensor discount."""
zero_discount = tf.zeros((1, self.batch_size))
non_zero_discount = tf.tile(tf.reshape(self.discount, [1, 1]),
[self.seq_length - 1, self.batch_size])
tensor_discount = tf.concat([zero_discount, non_zero_discount], axis=0)
loss, _ = pixel_control_ops.pixel_control_loss(
self.observations_ph, self.actions_ph, self.action_values_ph,
self.cell_size, tensor_discount, self.scale)
init = tf.global_variables_initializer()
with self.test_session() as sess:
sess.run(init)
feed_dict = {
self.observations_ph: self.observations,
self.action_values_ph: self.action_values,
self.actions_ph: self.actions
}
loss_np = sess.run(loss, feed_dict=feed_dict)
self.assertNear(loss_np, self.error_term, 1e-3)