本文整理汇总了Python中modeling.get_shape_list方法的典型用法代码示例。如果您正苦于以下问题:Python modeling.get_shape_list方法的具体用法?Python modeling.get_shape_list怎么用?Python modeling.get_shape_list使用的例子?那么, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类modeling
的用法示例。
在下文中一共展示了modeling.get_shape_list方法的9个代码示例,这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞,您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: gather_indexes
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import get_shape_list [as 别名]
def gather_indexes(sequence_tensor, positions):
"""
Gathers the vectors at the specific positions over a minibatch.
sequence_tensor: [batch, seq_length, width]
positions: [batch, n]
"""
sequence_shape = modeling.get_shape_list(sequence_tensor, expected_rank=3)
batch_size = sequence_shape[0]
seq_length = sequence_shape[1]
width = sequence_shape[2]
flat_offsets = tf.reshape(tf.range(0, batch_size, dtype=tf.int32) * seq_length, [-1, 1])
flat_positions = tf.reshape(positions + flat_offsets, [-1])
flat_sequence_tensor = tf.reshape(sequence_tensor, [batch_size * seq_length, width])
output_tensor = tf.gather(flat_sequence_tensor, flat_positions)
return output_tensor
#
示例2: extract_span_tensor
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import get_shape_list [as 别名]
def extract_span_tensor(bert_config, sequence_tensor, output_span_mask,
start_positions, end_positions,
scope=None):
sequence_shape = modeling.get_shape_list(sequence_tensor, expected_rank=3)
seq_length = sequence_shape[1]
output_span_mask = tf.cast(output_span_mask, tf.float32)
filtered_sequence_tensor = sequence_tensor * tf.expand_dims(output_span_mask, -1)
with tf.variable_scope(scope, default_name="span_loss"):
fw_output_tensor = gather_indexes(filtered_sequence_tensor, tf.expand_dims(start_positions, -1))
bw_output_tensor = gather_indexes(filtered_sequence_tensor, tf.expand_dims(end_positions, -1))
att_output_tensor = simple_attention(filtered_sequence_tensor, filtered_sequence_tensor, output_span_mask, self_adaptive=True)
output_tensor = tf.concat([fw_output_tensor, bw_output_tensor, att_output_tensor], axis=-1)
return output_tensor
#
示例3: create_model
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import get_shape_list [as 别名]
def create_model(bert_config, is_training, input_ids, input_mask, segment_ids,
use_one_hot_embeddings):
"""Creates a classification model."""
model = modeling.BertModel(
config=bert_config,
is_training=is_training,
input_ids=input_ids,
input_mask=input_mask,
token_type_ids=segment_ids,
use_one_hot_embeddings=use_one_hot_embeddings)
final_hidden = model.get_sequence_output()
final_hidden_shape = modeling.get_shape_list(final_hidden, expected_rank=3)
batch_size = final_hidden_shape[0]
seq_length = final_hidden_shape[1]
hidden_size = final_hidden_shape[2]
output_weights = tf.get_variable(
"cls/squad/output_weights", [2, hidden_size],
initializer=tf.truncated_normal_initializer(stddev=0.02))
output_bias = tf.get_variable(
"cls/squad/output_bias", [2], initializer=tf.zeros_initializer())
final_hidden_matrix = tf.reshape(final_hidden,
[batch_size * seq_length, hidden_size])
logits = tf.matmul(final_hidden_matrix, output_weights, transpose_b=True)
logits = tf.nn.bias_add(logits, output_bias)
logits = tf.reshape(logits, [batch_size, seq_length, 2])
logits = tf.transpose(logits, [2, 0, 1])
unstacked_logits = tf.unstack(logits, axis=0)
(start_logits, end_logits) = (unstacked_logits[0], unstacked_logits[1])
return (start_logits, end_logits)
开发者ID:Nagakiran1,项目名称:Extending-Google-BERT-as-Question-and-Answering-model-and-Chatbot,代码行数:40,代码来源:run_squad.py
示例4: gather_indexes
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import get_shape_list [as 别名]
def gather_indexes(sequence_tensor, positions):
"""Gathers the vectors at the specific positions over a minibatch."""
sequence_shape = modeling.get_shape_list(sequence_tensor, expected_rank=3)
batch_size = sequence_shape[0]
seq_length = sequence_shape[1]
width = sequence_shape[2]
flat_offsets = tf.reshape(
tf.range(0, batch_size, dtype=tf.int32) * seq_length, [-1, 1])
flat_positions = tf.reshape(positions + flat_offsets, [-1])
flat_sequence_tensor = tf.reshape(sequence_tensor,
[batch_size * seq_length, width])
output_tensor = tf.gather(flat_sequence_tensor, flat_positions)
return output_tensor
开发者ID:Nagakiran1,项目名称:Extending-Google-BERT-as-Question-and-Answering-model-and-Chatbot,代码行数:16,代码来源:run_pretraining.py
示例5: span_output_layer
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import get_shape_list [as 别名]
def span_output_layer(bert_config, input_tensor, input_span_mask=None, scope=None):
input_tensor_shape = modeling.get_shape_list(input_tensor, expected_rank=3)
batch_size = input_tensor_shape[0]
seq_length = input_tensor_shape[1]
hidden_size = input_tensor_shape[2]
# output layers
with tf.variable_scope(scope, default_name="cls/squad"):
output_weights = tf.get_variable("output_weights", [2, hidden_size],
initializer=tf.truncated_normal_initializer(stddev=0.02))
output_bias = tf.get_variable("output_bias", [2],
initializer=tf.zeros_initializer())
final_hidden_matrix = tf.reshape(input_tensor, [batch_size * seq_length, hidden_size])
logits = tf.matmul(final_hidden_matrix, output_weights, transpose_b=True)
logits = tf.nn.bias_add(logits, output_bias)
logits = tf.reshape(logits, [batch_size, seq_length, 2])
logits = tf.transpose(logits, [2, 0, 1])
unstacked_logits = tf.unstack(logits, axis=0)
(start_logits, end_logits) = (unstacked_logits[0], unstacked_logits[1])
if input_span_mask is not None:
adder = (1.0 - tf.cast(input_span_mask, tf.float32)) * -10000.0
start_logits += adder
end_logits += adder
return (start_logits, end_logits)
#
示例6: create_model
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import get_shape_list [as 别名]
def create_model(bert_config, is_training, input_ids, input_mask, segment_ids,
use_one_hot_embeddings):
"""Creates a classification model."""
model = modeling.BertModel(
config=bert_config,
is_training=is_training,
input_ids=input_ids,
input_mask=input_mask,
token_type_ids=segment_ids,
use_one_hot_embeddings=use_one_hot_embeddings)
final_hidden = model.get_sequence_output()
final_hidden_shape = modeling.get_shape_list(final_hidden, expected_rank=3)
batch_size = final_hidden_shape[0]
seq_length = final_hidden_shape[1]
hidden_size = final_hidden_shape[2]
output_weights = tf.compat.v1.get_variable(
"cls/squad/output_weights", [2, hidden_size],
initializer=tf.compat.v1.truncated_normal_initializer(stddev=0.02))
output_bias = tf.compat.v1.get_variable(
"cls/squad/output_bias", [2], initializer=tf.compat.v1.zeros_initializer())
final_hidden_matrix = tf.reshape(final_hidden,
[batch_size * seq_length, hidden_size])
logits = tf.matmul(final_hidden_matrix, output_weights, transpose_b=True)
logits = tf.nn.bias_add(logits, output_bias)
logits = tf.reshape(logits, [batch_size, seq_length, 2])
logits = tf.transpose(a=logits, perm=[2, 0, 1])
unstacked_logits = tf.unstack(logits, axis=0)
(start_logits, end_logits) = (unstacked_logits[0], unstacked_logits[1])
return (start_logits, end_logits)
示例7: create_model
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import get_shape_list [as 别名]
def create_model(bert_config, is_training, input_ids, input_mask, segment_ids,
use_one_hot_embeddings):
"""Creates a classification model."""
model = modeling.BertModel(
config=bert_config,
is_training=is_training,
input_ids=input_ids,
input_mask=input_mask,
token_type_ids=segment_ids,
use_one_hot_embeddings=use_one_hot_embeddings)
final_hidden = model.get_sequence_output()
final_hidden_shape = modeling.get_shape_list(final_hidden, expected_rank=3)
batch_size = final_hidden_shape[0]
seq_length = final_hidden_shape[1]
hidden_size = final_hidden_shape[2]
output_weights = tf.get_variable(
"cls/squad/output_weights", [2, hidden_size],
initializer=tf.truncated_normal_initializer(stddev=0.02))
output_bias = tf.get_variable(
"cls/squad/output_bias", [2], initializer=tf.zeros_initializer())
final_hidden_matrix = tf.reshape(final_hidden,
[batch_size * seq_length, hidden_size])
logits = tf.matmul(final_hidden_matrix, output_weights, transpose_b=True)
logits = tf.nn.bias_add(logits, output_bias)
logits = tf.reshape(logits, [batch_size, seq_length, 2])
logits = tf.transpose(logits, [2, 0, 1])
unstacked_logits = tf.unstack(logits, axis=0)
(start_logits, end_logits) = (unstacked_logits[0], unstacked_logits[1])
return (start_logits, end_logits)
示例8: gather_indexes
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import get_shape_list [as 别名]
def gather_indexes(sequence_tensor, positions):
"""Gathers the vectors at the specific positions over a minibatch."""
sequence_shape = modeling.get_shape_list(sequence_tensor, expected_rank=3)
batch_size = sequence_shape[0]
seq_length = sequence_shape[1]
width = sequence_shape[2]
flat_offsets = tf.reshape(
tf.range(0, batch_size, dtype=tf.int32) * seq_length, [-1, 1])
flat_positions = tf.reshape(positions + flat_offsets, [-1])
flat_sequence_tensor = tf.reshape(sequence_tensor,
[batch_size * seq_length, width])
output_tensor = tf.gather(flat_sequence_tensor, flat_positions)
return output_tensor
示例9: create_model
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import get_shape_list [as 别名]
def create_model(self, bert_config, is_training, input_ids, input_mask, segment_ids, use_one_hot_embeddings):
"""Creates a classification model."""
model = modeling.BertModel(
config=bert_config,
is_training=is_training,
input_ids=input_ids,
input_mask=input_mask,
token_type_ids=segment_ids,
use_one_hot_embeddings=use_one_hot_embeddings,
compute_type=tf.float32)
final_hidden = model.get_sequence_output()
final_hidden_shape = modeling.get_shape_list(final_hidden, expected_rank=3)
batch_size = final_hidden_shape[0]
seq_length = final_hidden_shape[1]
hidden_size = final_hidden_shape[2]
output_weights = tf.get_variable(
"cls/squad/output_weights", [2, hidden_size],
initializer=tf.truncated_normal_initializer(stddev=0.02))
output_bias = tf.get_variable(
"cls/squad/output_bias", [2], initializer=tf.zeros_initializer())
final_hidden_matrix = tf.reshape(final_hidden, [batch_size * seq_length, hidden_size])
logits = tf.matmul(final_hidden_matrix, output_weights, transpose_b=True)
logits = tf.nn.bias_add(logits, output_bias)
logits = tf.reshape(logits, [batch_size, seq_length, 2])
return logits
# logits = tf.transpose(logits, [2, 0, 1])
# unstacked_logits = tf.unstack(logits, axis=0, name='unstack')
# (start_logits, end_logits) = (unstacked_logits[0], unstacked_logits[1])
# return (start_logits, end_logits)