本文整理汇总了Python中modeling.create_initializer方法的典型用法代码示例。如果您正苦于以下问题：Python modeling.create_initializer方法的具体用法？Python modeling.create_initializer怎么用？Python modeling.create_initializer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.create_initializer方法的5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: get_next_sentence_output
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import create_initializer [as 别名]
def get_next_sentence_output(bert_config, input_tensor, labels):
"""Get loss and log probs for the next sentence prediction."""
# Simple binary classification. Note that 0 is "next sentence" and 1 is
# "random sentence". This weight matrix is not used after pre-training.
with tf.variable_scope("cls/seq_relationship"):
output_weights = tf.get_variable(
"output_weights",
shape=[2, bert_config.hidden_size],
initializer=modeling.create_initializer(bert_config.initializer_range))
output_bias = tf.get_variable(
"output_bias", shape=[2], initializer=tf.zeros_initializer())
logits = tf.matmul(input_tensor, output_weights, transpose_b=True)
logits = tf.nn.bias_add(logits, output_bias)
log_probs = tf.nn.log_softmax(logits, axis=-1)
labels = tf.reshape(labels, [-1])
one_hot_labels = tf.one_hot(labels, depth=2, dtype=tf.float32)
per_example_loss = -tf.reduce_sum(one_hot_labels * log_probs, axis=-1)
loss = tf.reduce_mean(per_example_loss)
return (loss, per_example_loss, log_probs) 开发者ID:Nagakiran1，项目名称:Extending-Google-BERT-as-Question-and-Answering-model-and-Chatbot，代码行数:23，代码来源:run_pretraining.py
示例2: get_next_sentence_output
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import create_initializer [as 别名]
def get_next_sentence_output(bert_config, input_tensor, labels):
"""Get loss and log probs for the next sentence prediction."""
# Simple binary classification. Note that 0 is "next sentence" and 1 is
# "random sentence". This weight matrix is not used after pre-training.
with tf.compat.v1.variable_scope("cls/seq_relationship"):
output_weights = tf.compat.v1.get_variable(
"output_weights",
shape=[2, bert_config.hidden_size],
initializer=modeling.create_initializer(bert_config.initializer_range))
output_bias = tf.compat.v1.get_variable(
"output_bias", shape=[2], initializer=tf.compat.v1.zeros_initializer())
logits = bf.matmul(input_tensor, output_weights, transpose_b=True)
logits = bf.i_cast(logits)
logits = tf.nn.bias_add(logits, output_bias)
log_probs = tf.nn.log_softmax(logits, axis=-1)
labels = tf.reshape(labels, [-1])
one_hot_labels = tf.one_hot(labels, depth=2, dtype=tf.float32)
per_example_loss = -tf.reduce_sum(input_tensor=one_hot_labels * log_probs, axis=-1)
loss = tf.reduce_mean(input_tensor=per_example_loss)
return (loss, per_example_loss, log_probs) 示例3: get_next_sentence_output
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import create_initializer [as 别名]
def get_next_sentence_output(bert_config, input_tensor, labels):
"""Get loss and log probs for the next sentence prediction."""
# Simple binary classification. Note that 0 is "next sentence" and 1 is
# "random sentence". This weight matrix is not used after pre-training.
with tf.compat.v1.variable_scope("cls/seq_relationship"):
output_weights = tf.compat.v1.get_variable(
"output_weights",
shape=[2, bert_config.hidden_size],
initializer=modeling.create_initializer(bert_config.initializer_range))
output_bias = tf.compat.v1.get_variable(
"output_bias", shape=[2], initializer=tf.compat.v1.zeros_initializer())
input_tensor = bf.i_cast(input_tensor)
output_weights = bf.i_cast(output_weights)
logits = tf.matmul(input_tensor, output_weights, transpose_b=True)
logits = tf.nn.bias_add(logits, output_bias)
log_probs = tf.nn.log_softmax(logits, axis=-1)
labels = tf.reshape(labels, [-1])
one_hot_labels = tf.one_hot(labels, depth=2, dtype=tf.float32)
per_example_loss = -tf.reduce_sum(input_tensor=one_hot_labels * log_probs, axis=-1)
loss = tf.reduce_mean(input_tensor=per_example_loss)
return (loss, per_example_loss, log_probs) 示例4: get_next_sentence_output
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import create_initializer [as 别名]
def get_next_sentence_output(bert_config, input_tensor, labels):
"""Get loss and log probs for the next sentence prediction."""
# Simple binary classification. Note that 0 is "next sentence" and 1 is
# "random sentence". This weight matrix is not used after pre-training.
with tf.variable_scope("cls/seq_relationship"):
output_weights = tf.get_variable(
"output_weights",
shape=[2, bert_config.hidden_size],
initializer=modeling.create_initializer(
bert_config.initializer_range))
output_bias = tf.get_variable(
"output_bias", shape=[2], initializer=tf.zeros_initializer())
logits = tf.matmul(input_tensor, output_weights, transpose_b=True)
logits = tf.nn.bias_add(logits, output_bias)
log_probs = tf.nn.log_softmax(logits, axis=-1)
labels = tf.reshape(labels, [-1])
one_hot_labels = tf.one_hot(labels, depth=2, dtype=tf.float32)
per_example_loss = -tf.reduce_sum(one_hot_labels * log_probs, axis=-1)
loss = tf.reduce_mean(per_example_loss)
return (loss, per_example_loss, log_probs) 示例5: get_next_sentence_output
# 需要导入模块: import modeling [as 别名]
# 或者: from modeling import create_initializer [as 别名]
def get_next_sentence_output(bert_config, input_tensor, labels):
"""Get loss and log probs for the next sentence prediction."""
# Simple binary classification. Note that 0 is "next sentence" and 1 is
# "random sentence". This weight matrix is not used after pre-training.
with tf.variable_scope("cls/seq_relationship"):
output_weights = tf.get_variable(
"output_weights",
shape=[2, bert_config.hidden_size],
initializer=modeling.create_initializer(bert_config.initializer_range))
output_bias = tf.get_variable(
"output_bias", shape=[2], initializer=tf.zeros_initializer())
logits = tf.matmul(input_tensor, output_weights, transpose_b=True)
logits = tf.nn.bias_add(logits, output_bias)
log_probs = tf.nn.log_softmax(logits, axis=-1)
labels = tf.reshape(labels, [-1])
one_hot_labels = tf.one_hot(labels, depth=2, dtype=tf.float32)
per_example_loss = -tf.reduce_sum(one_hot_labels * log_probs, axis=-1)
loss = tf.reduce_mean(per_example_loss)
return (loss, per_example_loss, log_probs)