python create_pretraining_data.py
–input_file=./sample_text.txt
–output_file=/tmp/tf_examples.tfrecord
–vocab_file=$BERT_BASE_DIR/vocab.txt
–do_lower_case=True
–max_seq_length=128
–max_predictions_per_seq=20
–masked_lm_prob=0.15
–random_seed=12345
–dupe_factor=5
def create_training_instances(input_files, tokenizer, max_seq_length,
dupe_factor, short_seq_prob, masked_lm_prob,
max_predictions_per_seq, rng):
"""Create `TrainingInstance`s from raw text."""
all_documents = [[]]
# Input file format:
# (1) One sentence per line. These should ideally be actual sentences, not
# entire paragraphs or arbitrary spans of text. (Because we use the
# sentence boundaries for the "next sentence prediction" task).
# (2) Blank lines between documents. Document boundaries are needed so
# that the "next sentence prediction" task doesn't span between documents.
for input_file in input_files:
with tf.gfile.GFile(input_file, "r") as reader:
while True:
line = tokenization.convert_to_unicode(reader.readline())
if not line:
break
line = line.strip()
# Empty lines are used as document delimiters
if not line:
all_documents.append([])
tokens = tokenizer.tokenize(line) # token化
if tokens:
all_documents[-1].append(tokens)
# Remove empty documents
all_documents = [x for x in all_documents if x]
rng.shuffle(all_documents) # 随机shuffle
vocab_words = list(tokenizer.vocab.keys())
instances = []
for _ in range(dupe_factor):
for document_index in range(len(all_documents)):
instances.extend(
create_instances_from_document(
all_documents, document_index, max_seq_length, short_seq_prob,
masked_lm_prob, max_predictions_per_seq, vocab_words, rng))
rng.shuffle(instances)
return instances
输入文件要求(next sentence prediction):1. 一个句子一行;2. 文档之间用空白行隔空
由两部分构成:BasicTokenizer和WordpieceTokenizer
其中BasicTokenizer将中文按照字进行切分,英文按照标点符号进行切分。WordpieceTokenizer对BasicTokenizer切分的每个词按照longest-match-first前向最长查找vocabulary,不是开头匹配的词加入##标示,比如"unaffable",最长匹配词为un, ##aff, ##able,输出[“un”, “##aff”, “##able”]。
for token in whitespace_tokenize(text):
chars = list(token)
# 超出词最大输入字符的部分用unk_token替代
if len(chars) > self.max_input_chars_per_word:
output_tokens.append(self.unk_token)
continue
is_bad = False
start = 0
sub_tokens = []
while start < len(chars):
end = len(chars)
cur_substr = None
# 从end开始查找start:end之间的词是否在vocab中,如果存在则找到,如果不存在则依次将end减1
while start < end:
substr = "".join(chars[start:end])
if start > 0:
substr = "##" + substr
if substr in self.vocab:
cur_substr = substr
break
end -= 1
if cur_substr is None:
is_bad = True
break
sub_tokens.append(cur_substr)
start = end
if is_bad:
output_tokens.append(self.unk_token)
else:
output_tokens.extend(sub_tokens)
return output_tokens
def create_instances_from_document(
all_documents, document_index, max_seq_length, short_seq_prob,
masked_lm_prob, max_predictions_per_seq, vocab_words, rng):
"""Creates `TrainingInstance`s for a single document."""
document = all_documents[document_index]
# Account for [CLS], [SEP], [SEP]
max_num_tokens = max_seq_length - 3
# We *usually* want to fill up the entire sequence since we are padding
# to `max_seq_length` anyways, so short sequences are generally wasted
# computation. However, we *sometimes*
# (i.e., short_seq_prob == 0.1 == 10% of the time) want to use shorter
# sequences to minimize the mismatch between pre-training and fine-tuning.
# The `target_seq_length` is just a rough target however, whereas
# `max_seq_length` is a hard limit.
target_seq_length = max_num_tokens
if rng.random() < short_seq_prob:
target_seq_length = rng.randint(2, max_num_tokens)
# We DON'T just concatenate all of the tokens from a document into a long
# sequence and choose an arbitrary split point because this would make the
# next sentence prediction task too easy. Instead, we split the input into
# segments "A" and "B" based on the actual "sentences" provided by the user
# input.
# 创新点1:下一句预测
instances = []
current_chunk = []
current_length = 0
i = 0
while i < len(document):
segment = document[i]
current_chunk.append(segment)
current_length += len(segment)
if i == len(document) - 1 or current_length >= target_seq_length:
if current_chunk:
# `a_end` is how many segments from `current_chunk` go into the `A`
# (first) sentence.
# 拆分A、B句作next sentence prediction
a_end = 1
# 随机采样当前句a_end
if len(current_chunk) >= 2:
a_end = rng.randint(1, len(current_chunk) - 1)
tokens_a = []
for j in range(a_end):
tokens_a.extend(current_chunk[j])
tokens_b = []
# Random next
is_random_next = False
# 如果文档只有一个segment一句话,则随机从其他文档采样得到下一句。
# 50%几率随机从其他文档采样(随机长度句子)得到下一句,50%几率使用真实的下一句作为下一句
if len(current_chunk) == 1 or rng.random() < 0.5:
is_random_next = True
target_b_length = target_seq_length - len(tokens_a)
# This should rarely go for more than one iteration for large
# corpora. However, just to be careful, we try to make sure that
# the random document is not the same as the document
# we're processing.
# 为了避免random的文档和原文档一样
for _ in range(10):
random_document_index = rng.randint(0, len(all_documents) - 1)
if random_document_index != document_index:
break
random_document = all_documents[random_document_index]
random_start = rng.randint(0, len(random_document) - 1)
# 随机产生采样的开始点
for j in range(random_start, len(random_document)):
tokens_b.extend(random_document[j])
if len(tokens_b) >= target_b_length:
break
# We didn't actually use these segments so we "put them back" so
# they don't go to waste.
num_unused_segments = len(current_chunk) - a_end
i -= num_unused_segments
# Actual next
else:
is_random_next = False
for j in range(a_end, len(current_chunk)):
tokens_b.extend(current_chunk[j])
# 截断使其满足max_num_tokens
truncate_seq_pair(tokens_a, tokens_b, max_num_tokens, rng)
assert len(tokens_a) >= 1
assert len(tokens_b) >= 1
# 生成 [[CLS]+第一句+[SEP]+下一句+[SEP]]
tokens = []
segment_ids = []
tokens.append("[CLS]")
segment_ids.append(0)
for token in tokens_a:
tokens.append(token)
segment_ids.append(0)
tokens.append("[SEP]")
segment_ids.append(0)
for token in tokens_b:
tokens.append(token)
segment_ids.append(1)
tokens.append("[SEP]")
segment_ids.append(1)
# 创新点2:随机掩蔽
(tokens, masked_lm_positions,
masked_lm_labels) = create_masked_lm_predictions(
tokens, masked_lm_prob, max_predictions_per_seq, vocab_words, rng)
instance = TrainingInstance(
tokens=tokens,
segment_ids=segment_ids,
is_random_next=is_random_next,
masked_lm_positions=masked_lm_positions,
masked_lm_labels=masked_lm_labels)
instances.append(instance)
current_chunk = []
current_length = 0
i += 1
return instances
def create_masked_lm_predictions(tokens, masked_lm_prob,
max_predictions_per_seq, vocab_words, rng):
"""Creates the predictions for the masked LM objective."""
cand_indexes = []
for (i, token) in enumerate(tokens):
if token == "[CLS]" or token == "[SEP]":
continue
cand_indexes.append(i)
rng.shuffle(cand_indexes) # 随机挑选词进行掩蔽
output_tokens = list(tokens)
# 随机掩蔽的词数量masked_lm_prob(15%)
num_to_predict = min(max_predictions_per_seq,
max(1, int(round(len(tokens) * masked_lm_prob))))
masked_lms = []
covered_indexes = set()
for index in cand_indexes:
if len(masked_lms) >= num_to_predict:
break
if index in covered_indexes:
continue
covered_indexes.add(index)
masked_token = None
# 80% of the time, replace with [MASK]
# 80%替换为[MASK]
if rng.random() < 0.8:
masked_token = "[MASK]"
else:
# 10% of the time, keep original
# 10%保持原样(0.2 x 0.5)
if rng.random() < 0.5:
masked_token = tokens[index]
# 10% of the time, replace with random word
# 10%随机使用字典库里面的词进行替换(可能是原样的词)
else:
masked_token = vocab_words[rng.randint(0, len(vocab_words) - 1)]
output_tokens[index] = masked_token
masked_lms.append(MaskedLmInstance(index=index, label=tokens[index]))
masked_lms = sorted(masked_lms, key=lambda x: x.index)
# 随机掩蔽的词的位置和真实的label
masked_lm_positions = []
masked_lm_labels = []
for p in masked_lms:
masked_lm_positions.append(p.index)
masked_lm_labels.append(p.label)
return (output_tokens, masked_lm_positions, masked_lm_labels)
创新点有:
Next Sentence Prediction
通过将句子拆分成A(当前句)、B句(下一句),基于A句选择B句的策略为:50%几率A、B句是真实的连续句,50%几率A、B句是真实的不连续句。
Masked LM
使用随机15%的token词作mask,mask的策略为 :
(1) 80%替换为[MASK]
(2) 10%保持原词
(3) 10%替换为随机词
这样mask策略的目的:
作者认为: Intuitively, it is reasonable to believe that a deep bidirectional model is strictly more powerful than either a left-to-right model or the shallow concatenation of a left-toright and a right-to-left model. Unfortunately, standard conditional language models can only be trained left-to-right or right-to-left, since bidirectional conditioning would allow each word to indirectly “see itself”, and the model could trivially predict the target word in a multi-layered context.
传统的bidirectional不够deep bidirectional。作者提出了Masked LM。随机选择15%的token来进行mask。选择上述MASK策略的原因:
随机的词带来的负面影响可以忽略不计,因为15%*10%=1.5%的概率很小。
具体实现见代码里面的注释。
def write_instance_to_example_files(instances, tokenizer, max_seq_length,
max_predictions_per_seq, output_files):
"""Create TF example files from `TrainingInstance`s."""
writers = []
for output_file in output_files:
writers.append(tf.python_io.TFRecordWriter(output_file))
writer_index = 0
total_written = 0
for (inst_index, instance) in enumerate(instances):
input_ids = tokenizer.convert_tokens_to_ids(instance.tokens)
input_mask = [1] * len(input_ids)
segment_ids = list(instance.segment_ids)
assert len(input_ids) <= max_seq_length
while len(input_ids) < max_seq_length:
input_ids.append(0)
input_mask.append(0)
segment_ids.append(0)
assert len(input_ids) == max_seq_length
assert len(input_mask) == max_seq_length
assert len(segment_ids) == max_seq_length
masked_lm_positions = list(instance.masked_lm_positions)
masked_lm_ids = tokenizer.convert_tokens_to_ids(instance.masked_lm_labels)
masked_lm_weights = [1.0] * len(masked_lm_ids)
# 对于不足max_predictions_per_seq的部分position, ids, weights补零,以供train/eval使用
while len(masked_lm_positions) < max_predictions_per_seq:
masked_lm_positions.append(0)
masked_lm_ids.append(0)
masked_lm_weights.append(0.0)
next_sentence_label = 1 if instance.is_random_next else 0
features = collections.OrderedDict()
features["input_ids"] = create_int_feature(input_ids)
features["input_mask"] = create_int_feature(input_mask)
features["segment_ids"] = create_int_feature(segment_ids)
features["masked_lm_positions"] = create_int_feature(masked_lm_positions)
features["masked_lm_ids"] = create_int_feature(masked_lm_ids)
features["masked_lm_weights"] = create_float_feature(masked_lm_weights)
features["next_sentence_labels"] = create_int_feature([next_sentence_label])
tf_example = tf.train.Example(features=tf.train.Features(feature=features))
writers[writer_index].write(tf_example.SerializeToString())
writer_index = (writer_index + 1) % len(writers)
total_written += 1
if inst_index < 20:
tf.logging.info("*** Example ***")
tf.logging.info("tokens: %s" % " ".join(
[tokenization.printable_text(x) for x in instance.tokens]))
for feature_name in features.keys():
feature = features[feature_name]
values = []
if feature.int64_list.value:
values = feature.int64_list.value
elif feature.float_list.value:
values = feature.float_list.value
tf.logging.info(
"%s: %s" % (feature_name, " ".join([str(x) for x in values])))
for writer in writers:
writer.close()
tf.logging.info("Wrote %d total instances", total_written)
python run_pretraining.py
–input_file=/tmp/tf_examples.tfrecord
–output_dir=/tmp/pretraining_output
–do_train=True
–do_eval=True
–bert_config_file=KaTeX parse error: Undefined control sequence: \ at position 32: …rt_config.json \̲ ̲ --init_checkp…BERT_BASE_DIR/bert_model.ckpt
–train_batch_size=32
–max_seq_length=128
–max_predictions_per_seq=20
–num_train_steps=20
–num_warmup_steps=10
–learning_rate=2e-5
def main(_):
tf.logging.set_verbosity(tf.logging.INFO)
if not FLAGS.do_train and not FLAGS.do_eval:
raise ValueError("At least one of `do_train` or `do_eval` must be True.")
bert_config = modeling.BertConfig.from_json_file(FLAGS.bert_config_file)
tf.gfile.MakeDirs(FLAGS.output_dir)
input_files = []
for input_pattern in FLAGS.input_file.split(","):
input_files.extend(tf.gfile.Glob(input_pattern))
tf.logging.info("*** Input Files ***")
for input_file in input_files:
tf.logging.info(" %s" % input_file)
tpu_cluster_resolver = None
if FLAGS.use_tpu and FLAGS.tpu_name:
tpu_cluster_resolver = tf.contrib.cluster_resolver.TPUClusterResolver(
FLAGS.tpu_name, zone=FLAGS.tpu_zone, project=FLAGS.gcp_project)
is_per_host = tf.contrib.tpu.InputPipelineConfig.PER_HOST_V2
run_config = tf.contrib.tpu.RunConfig(
cluster=tpu_cluster_resolver,
master=FLAGS.master,
model_dir=FLAGS.output_dir,
save_checkpoints_steps=FLAGS.save_checkpoints_steps,
tpu_config=tf.contrib.tpu.TPUConfig(
iterations_per_loop=FLAGS.iterations_per_loop,
num_shards=FLAGS.num_tpu_cores,
per_host_input_for_training=is_per_host))
model_fn = model_fn_builder(
bert_config=bert_config,
init_checkpoint=FLAGS.init_checkpoint,
learning_rate=FLAGS.learning_rate,
num_train_steps=FLAGS.num_train_steps,
num_warmup_steps=FLAGS.num_warmup_steps,
use_tpu=FLAGS.use_tpu,
use_one_hot_embeddings=FLAGS.use_tpu)
# If TPU is not available, this will fall back to normal Estimator on CPU
# or GPU.
estimator = tf.contrib.tpu.TPUEstimator(
use_tpu=FLAGS.use_tpu,
model_fn=model_fn,
config=run_config,
train_batch_size=FLAGS.train_batch_size,
eval_batch_size=FLAGS.eval_batch_size)
if FLAGS.do_train:
tf.logging.info("***** Running training *****")
tf.logging.info(" Batch size = %d", FLAGS.train_batch_size)
train_input_fn = input_fn_builder(
input_files=input_files,
max_seq_length=FLAGS.max_seq_length,
max_predictions_per_seq=FLAGS.max_predictions_per_seq,
is_training=True)
estimator.train(input_fn=train_input_fn, max_steps=FLAGS.num_train_steps)
if FLAGS.do_eval:
tf.logging.info("***** Running evaluation *****")
tf.logging.info(" Batch size = %d", FLAGS.eval_batch_size)
eval_input_fn = input_fn_builder(
input_files=input_files,
max_seq_length=FLAGS.max_seq_length,
max_predictions_per_seq=FLAGS.max_predictions_per_seq,
is_training=False)
result = estimator.evaluate(
input_fn=eval_input_fn, steps=FLAGS.max_eval_steps)
output_eval_file = os.path.join(FLAGS.output_dir, "eval_results.txt")
with tf.gfile.GFile(output_eval_file, "w") as writer:
tf.logging.info("***** Eval results *****")
for key in sorted(result.keys()):
tf.logging.info(" %s = %s", key, str(result[key]))
writer.write("%s = %s\n" % (key, str(result[key])))
代码执行如下步骤:
model_fn的代码及注释:
def model_fn_builder(bert_config, init_checkpoint, learning_rate,
num_train_steps, num_warmup_steps, use_tpu,
use_one_hot_embeddings):
"""Returns `model_fn` closure for TPUEstimator."""
def model_fn(features, labels, mode, params): # pylint: disable=unused-argument
"""The `model_fn` for TPUEstimator."""
tf.logging.info("*** Features ***")
for name in sorted(features.keys()):
tf.logging.info(" name = %s, shape = %s" % (name, features[name].shape))
input_ids = features["input_ids"]
input_mask = features["input_mask"]
segment_ids = features["segment_ids"]
masked_lm_positions = features["masked_lm_positions"]
masked_lm_ids = features["masked_lm_ids"]
masked_lm_weights = features["masked_lm_weights"]
next_sentence_labels = features["next_sentence_labels"]
is_training = (mode == tf.estimator.ModeKeys.TRAIN)
# Bert模型,下一章再详细介绍
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)
# get_sequence_output:最后一层隐藏层[batch_size, seq_length, hidden_size]
# get_embedding_table:embedding层
# get_pooled_output: 最后一层隐藏层第一个token即[CLS]通过一个dense层的输出[batch_size, hidden_size],该token包含了整个句子的信息,适用于segment-level的分类。
(masked_lm_loss,
masked_lm_example_loss, masked_lm_log_probs) = get_masked_lm_output(
bert_config, model.get_sequence_output(), model.get_embedding_table(),
masked_lm_positions, masked_lm_ids, masked_lm_weights)
(next_sentence_loss, next_sentence_example_loss,
next_sentence_log_probs) = get_next_sentence_output(
bert_config, model.get_pooled_output(), next_sentence_labels)
total_loss = masked_lm_loss + next_sentence_loss
tvars = tf.trainable_variables()
initialized_variable_names = {}
scaffold_fn = None
if init_checkpoint:
(assignment_map, initialized_variable_names
) = modeling.get_assignment_map_from_checkpoint(tvars, init_checkpoint)
if use_tpu:
def tpu_scaffold():
tf.train.init_from_checkpoint(init_checkpoint, assignment_map)
return tf.train.Scaffold()
scaffold_fn = tpu_scaffold
else:
tf.train.init_from_checkpoint(init_checkpoint, assignment_map)
tf.logging.info("**** Trainable Variables ****")
for var in tvars:
init_string = ""
if var.name in initialized_variable_names:
init_string = ", *INIT_FROM_CKPT*"
tf.logging.info(" name = %s, shape = %s%s", var.name, var.shape,
init_string)
output_spec = None
if mode == tf.estimator.ModeKeys.TRAIN:
# 使用AdamWeightDecayOptimizer最优化
train_op = optimization.create_optimizer(
total_loss, learning_rate, num_train_steps, num_warmup_steps, use_tpu)
output_spec = tf.contrib.tpu.TPUEstimatorSpec(
mode=mode,
loss=total_loss,
train_op=train_op,
scaffold_fn=scaffold_fn)
elif mode == tf.estimator.ModeKeys.EVAL:
def metric_fn(masked_lm_example_loss, masked_lm_log_probs, masked_lm_ids,
masked_lm_weights, next_sentence_example_loss,
next_sentence_log_probs, next_sentence_labels):
"""Computes the loss and accuracy of the model."""
# 计算masked_lm和next_sentence预测准确度和平均损失
masked_lm_log_probs = tf.reshape(masked_lm_log_probs,
[-1, masked_lm_log_probs.shape[-1]])
masked_lm_predictions = tf.argmax(
masked_lm_log_probs, axis=-1, output_type=tf.int32)
masked_lm_example_loss = tf.reshape(masked_lm_example_loss, [-1])
masked_lm_ids = tf.reshape(masked_lm_ids, [-1])
masked_lm_weights = tf.reshape(masked_lm_weights, [-1])
masked_lm_accuracy = tf.metrics.accuracy(
labels=masked_lm_ids,
predictions=masked_lm_predictions,
weights=masked_lm_weights)
masked_lm_mean_loss = tf.metrics.mean(
values=masked_lm_example_loss, weights=masked_lm_weights)
next_sentence_log_probs = tf.reshape(
next_sentence_log_probs, [-1, next_sentence_log_probs.shape[-1]])
next_sentence_predictions = tf.argmax(
next_sentence_log_probs, axis=-1, output_type=tf.int32)
next_sentence_labels = tf.reshape(next_sentence_labels, [-1])
next_sentence_accuracy = tf.metrics.accuracy(
labels=next_sentence_labels, predictions=next_sentence_predictions)
next_sentence_mean_loss = tf.metrics.mean(
values=next_sentence_example_loss)
return {
"masked_lm_accuracy": masked_lm_accuracy,
"masked_lm_loss": masked_lm_mean_loss,
"next_sentence_accuracy": next_sentence_accuracy,
"next_sentence_loss": next_sentence_mean_loss,
}
eval_metrics = (metric_fn, [
masked_lm_example_loss, masked_lm_log_probs, masked_lm_ids,
masked_lm_weights, next_sentence_example_loss,
next_sentence_log_probs, next_sentence_labels
])
output_spec = tf.contrib.tpu.TPUEstimatorSpec(
mode=mode,
loss=total_loss,
eval_metrics=eval_metrics,
scaffold_fn=scaffold_fn)
else:
raise ValueError("Only TRAIN and EVAL modes are supported: %s" % (mode))
return output_spec
return model_fn
def get_masked_lm_output(bert_config, input_tensor, output_weights, positions,
label_ids, label_weights):
"""Get loss and log probs for the masked LM."""
# 将position进行flatten,由[batch_size, max_predictions_per_seq]转换成[batch_size * max_predictions_per_seq];
# 然后再从input_tensor中查找到positions对应位置的tensor,[batch_size * max_predictions_per_seq, hidden_size]
input_tensor = gather_indexes(input_tensor, positions)
# 非线性转换,加入LN
with tf.variable_scope("cls/predictions"):
# We apply one more non-linear transformation before the output layer.
# This matrix is not used after pre-training.
with tf.variable_scope("transform"):
input_tensor = tf.layers.dense(
input_tensor,
units=bert_config.hidden_size, # 隐藏层大小
activation=modeling.get_activation(bert_config.hidden_act), # 激活函数
kernel_initializer=modeling.create_initializer(
bert_config.initializer_range)) # 截断正太分布标准差
input_tensor = modeling.layer_norm(input_tensor)
# The output weights are the same as the input embeddings, but there is
# an output-only bias for each token.
output_bias = tf.get_variable(
"output_bias",
shape=[bert_config.vocab_size],
initializer=tf.zeros_initializer())
# softmax(wx+b)
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)
# 将label_ids进行flatten,[batch_size * max_predictions_per_seq]
label_ids = tf.reshape(label_ids, [-1])
label_weights = tf.reshape(label_weights, [-1])
# 将label_ids进行onehot。[batch_size * max_predictions_per_seq, vocab_size]
one_hot_labels = tf.one_hot(
label_ids, depth=bert_config.vocab_size, dtype=tf.float32)
# The `positions` tensor might be zero-padded (if the sequence is too
# short to have the maximum number of predictions). The `label_weights`
# tensor has a value of 1.0 for every real prediction and 0.0 for the
# padding predictions.
# 每个sample的softmax损失
per_example_loss = -tf.reduce_sum(log_probs * one_hot_labels, axis=[-1])
numerator = tf.reduce_sum(label_weights * per_example_loss)
denominator = tf.reduce_sum(label_weights) + 1e-5
# 用masked_lm_weights来对每个sample的loss加权求平均损失
# 对于position小于max_predictions_per_seq的masked_lm_weights权重填充为0,其余为1
loss = numerator / denominator
return (loss, per_example_loss, log_probs)
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())
# softmax(wx+b), 使用bert的pooled_output作为输入来作next sentence分类
# (0: real next sentence, 1: random sentence)
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)
主要定义masked_lm和next_sentence两个任务的loss,bert模型 ,adam最优化以及相应的metric指标。
注:TPUEstimator当TPU不可见的时候,会自动fall back到CPU或者GPU
def input_fn_builder(input_files,
max_seq_length,
max_predictions_per_seq,
is_training,
num_cpu_threads=4):
"""Creates an `input_fn` closure to be passed to TPUEstimator."""
def input_fn(params):
"""The actual input function."""
batch_size = params["batch_size"]
#
name_to_features = {
"input_ids":
tf.FixedLenFeature([max_seq_length], tf.int64),
"input_mask":
tf.FixedLenFeature([max_seq_length], tf.int64),
"segment_ids":
tf.FixedLenFeature([max_seq_length], tf.int64),
"masked_lm_positions":
tf.FixedLenFeature([max_predictions_per_seq], tf.int64),
"masked_lm_ids":
tf.FixedLenFeature([max_predictions_per_seq], tf.int64),
"masked_lm_weights":
tf.FixedLenFeature([max_predictions_per_seq], tf.float32),
"next_sentence_labels":
tf.FixedLenFeature([1], tf.int64),
}
# For training, we want a lot of parallel reading and shuffling.
# For eval, we want no shuffling and parallel reading doesn't matter.
if is_training:
d = tf.data.Dataset.from_tensor_slices(tf.constant(input_files))
d = d.repeat()
d = d.shuffle(buffer_size=len(input_files))
# `cycle_length` is the number of parallel files that get read.
cycle_length = min(num_cpu_threads, len(input_files))
# `sloppy` mode means that the interleaving is not exact. This adds
# even more randomness to the training pipeline.
# 并行化处理
d = d.apply(
tf.contrib.data.parallel_interleave(
tf.data.TFRecordDataset,
sloppy=is_training,
cycle_length=cycle_length))
d = d.shuffle(buffer_size=100)
else:
d = tf.data.TFRecordDataset(input_files)
# Since we evaluate for a fixed number of steps we don't want to encounter
# out-of-range exceptions.
d = d.repeat()
# We must `drop_remainder` on training because the TPU requires fixed
# size dimensions. For eval, we assume we are evaluating on the CPU or GPU
# and we *don't* want to drop the remainder, otherwise we wont cover
# every sample.
d = d.apply(
tf.contrib.data.map_and_batch(
lambda record: _decode_record(record, name_to_features),
batch_size=batch_size,
num_parallel_batches=num_cpu_threads,
drop_remainder=True))
return d
return input_fn