def custom_constant_layer(const_size, name, np_constant=None):
if np_constant is None:
np_constant = np.zeros((const_size, 1))
else:
np_constant = np.reshape(np_constant, (*np_constant.shape, 1))
const_fake_extender = keras_constant_layer(np_constant, name=name)
# tf_constant = tf.keras.backend.constant(np_constant, dtype='float32')
# const_fake_extender = Input(tensor=tf_constant, shape=np_constant.shape, dtype='float32', name=name)
# TODO: reshaping constant input??
return tf.keras.layers.Lambda(lambda x: tf.keras.backend.reshape(x, const_fake_extender.shape))(
const_fake_extender), const_fake_extender
def build_active_passive_network(roles, dual_roles, fillers, tree_shape):
filler_len = fillers[0].shape[0]
max_depth = len(tree_shape) - 1
branch = build_classification_branch(roles=dual_roles,
fillers=fillers,
tree_shape=tree_shape,
role_extraction_order=[1, 0, 0],
stop_level=0)
classification_const_inputs, apnet_variable_input, classification_output, _ = branch
scalar_mul = tf.keras.layers.Lambda(lambda tensors: tensors[0] * tensors[1])(
[apnet_variable_input, classification_output])
passive_branch_const_inputs, passive_branch_output = extract_semantic_tree_from_passive_voice_branch(
input_layer=scalar_mul,
roles=roles,
dual_roles=dual_roles,
filler_len=filler_len,
max_depth=max_depth)
np_constant = np.array([-1])
const_neg_1 = keras_constant_layer(np_constant, name='active_voice_neg_1')
sum_is_passive_const_neg_1 = tf.keras.layers.Add()([classification_output, const_neg_1])
is_active = tf.keras.layers.Multiply()([sum_is_passive_const_neg_1, const_neg_1])
active_branch_input = tf.keras.layers.Lambda(lambda tensors: tensors[0] * tensors[1])(
[apnet_variable_input, is_active])
active_branch_const_inputs, active_branch_output = extract_semantic_tree_from_active_voice_branch(
input_layer=active_branch_input,
roles=roles,
dual_roles=dual_roles,
filler_len=filler_len,
max_depth=max_depth)
sum_branches = tf.keras.layers.Add()([passive_branch_output, active_branch_output])
return tf.keras.Model(
inputs=[
*classification_const_inputs,
*passive_branch_const_inputs,
*active_branch_const_inputs,
const_neg_1,
apnet_variable_input,
],
outputs=sum_branches)
def constant_input(role, filler_len, max_depth, name, matrix_creator):
np_constant = matrix_creator(role, filler_len, max_depth, name)
return keras_constant_layer(np_constant, name=name)
def extract_semantic_tree_from_passive_voice_branch(input_layer, roles,
dual_roles, filler_len,
max_depth):
stop_level_for_verb = 0
verb_branch = build_one_level_extraction_branch(
model_input=input_layer,
roles=dual_roles,
filler_len=filler_len,
max_depth=max_depth,
stop_level=stop_level_for_verb,
role_extraction_order=[1, 0, 1],
prefix='passive_verb_extract')
verb_extraction_const_inputs, verb_extraction_output, _ = verb_branch
stop_level_for_agent = 0
agent_branch = build_one_level_extraction_branch(
model_input=input_layer,
roles=dual_roles,
filler_len=filler_len,
max_depth=max_depth,
stop_level=stop_level_for_agent,
role_extraction_order=[1, 1, 1],
prefix='passive_agent_extract')
agent_extraction_const_inputs, agent_extraction_output, _ = agent_branch
stop_level_for_p = max_depth - 1
p_branch = build_one_level_extraction_branch(model_input=input_layer,
roles=dual_roles,
filler_len=filler_len,
max_depth=max_depth,
stop_level=stop_level_for_p,
role_extraction_order=[0],
prefix='passive_p_extract')
p_extraction_const_inputs, p_raw_output, current_num_elements = p_branch
_, flattened_num_elements = unshift_matrix(roles[0], filler_len,
stop_level_for_p).shape
# TODO: insert cropping here
reshape_for_crop = tf.keras.layers.Lambda(
lambda x: tf.keras.backend.reshape(x, (1, flattened_num_elements +
filler_len, 1)))(p_raw_output)
clip_first_level = tf.keras.layers.Cropping1D(
cropping=(0, flattened_num_elements))(reshape_for_crop)
p_extraction_output = tf.keras.layers.Lambda(
lambda x: tf.keras.backend.reshape(x, (1, filler_len, 1)))(
clip_first_level)
# TODO: define how to tackle extractions not till the bottom of structure
# given that we have all fillers maximum joining depth is equal to 1
agentxr0_pxr1_const_inputs, agentxr0_pxr1_output = build_join_branch(
roles=roles,
filler_len=filler_len,
max_depth=1,
inputs=[agent_extraction_output, p_extraction_output],
prefix='passive_join(agent,p)')
# later we have to join two subtrees of different depth. for that we have to
# make filler of verb of the same depth - make fake constant layer
np_constant = np.zeros((filler_len, 1))
const_fake_extender = keras_constant_layer(
np_constant, name='passive_fake_extender_verb_agent')
concatenate_verb = tf.keras.layers.Concatenate(axis=0)(
[verb_extraction_output, const_fake_extender, const_fake_extender])
# TODO: why is there a constant 3?
reshaped_verb = tf.keras.layers.Lambda(lambda x: tf.keras.backend.reshape(
x, (1, filler_len * 3, 1)))(concatenate_verb)
# TODO: reshape by 2, why is there a constant 2?
tmp_reshaped_agentxr0_pxr1 = tf.keras.layers.Lambda(
lambda x: tf.keras.backend.reshape(x, (filler_len * 2, 1)))(
agentxr0_pxr1_output)
# TODO: reshaping constant input??
tmp_reshaped_fake = tf.keras.layers.Lambda(
lambda x: tf.keras.backend.reshape(x, (filler_len, 1)))(
const_fake_extender)
concatenate_agentxr0_pxr1 = tf.keras.layers.Concatenate(axis=0)(
[tmp_reshaped_fake, tmp_reshaped_agentxr0_pxr1])
# TODO: why is there a constant 3?
reshaped_agentxr0_pxr1 = tf.keras.layers.Lambda(
lambda x: tf.keras.backend.reshape(x, (1, filler_len * 3, 1)))(
concatenate_agentxr0_pxr1)
semantic_tree_const_inputs, semantic_tree_output = build_join_branch(
roles=roles,
filler_len=filler_len,
max_depth=2,
inputs=[reshaped_verb, reshaped_agentxr0_pxr1],
prefix='passive_join(verb, join(agent,p))')
return [
*verb_extraction_const_inputs, *agent_extraction_const_inputs,
*p_extraction_const_inputs, *agentxr0_pxr1_const_inputs,
*semantic_tree_const_inputs, const_fake_extender
], semantic_tree_output
def check_if_zero_branch(flag_input, block_id):
np_constant = np.array([-1])
const_neg_1 = keras_constant_layer(np_constant, 'increment_neg_{}'.format(block_id))
sum_is_zero_const = tf.keras.layers.Add()([flag_input, const_neg_1])
return const_neg_1, tf.keras.layers.Multiply()([sum_is_zero_const, const_neg_1])