sparse_categorical_crossentropy() 缺少 2 个必需的位置参数：'y_true' 和 'y_pred'

Question

我想使用 Bert 语言模型来训练多 class 文本 class 化任务。以前我使用 LSTM 训练时没有任何错误，但 Bert 给了我这个错误。我收到以下错误，我真的不知道如何解决它，有人可以帮我吗？

不幸的是，keras 库中使用 Bert 的文档很少。

错误：

TypeError                                 Traceback (most recent call last)
<ipython-input-177-7b203e5e7f55> in <module>()
      3 
      4 model.compile(optimizer = tf.keras.optimizers.Adam(learning_rate= 2e-5),
----> 5               loss = tf.keras.losses.sparse_categorical_crossentropy(from_logits=True),
      6               metrics = [tf.keras.metrics.categorical_accuracy()])
      7 model.summary

/usr/local/lib/python3.7/dist-packages/tensorflow/python/util/dispatch.py in wrapper(*args, **kwargs)
    199     """Call target, and fall back on dispatchers if there is a TypeError."""
    200     try:
--> 201       return target(*args, **kwargs)
    202     except (TypeError, ValueError):
    203       # Note: convert_to_eager_tensor currently raises a ValueError, not a

TypeError: sparse_categorical_crossentropy() missing 2 required positional arguments: 'y_true' and 'y_pred'
SEARCH

为训练和评估创建tf.data.Datasets

X = data.text_cleaned
Y = data.label

train_df, remaining = train_test_split(data, train_size=0.8, random_state=42)
valid_df, _ = train_test_split(remaining, random_state=42)

print(train_df.shape)
print(valid_df.shape)


with tf.device('/cpu:0'):
  train_data = tf.data.Dataset.from_tensor_slices((train_df["text_cleaned"].values,
                                            train_df["label"].values ))
  valid_data = tf.data.Dataset.from_tensor_slices((valid_df["text_cleaned"].values,
                                            valid_df["label"].values ))
  for text, label in train_data.take(2):
    print(text)
    print(label)

从 TensorFlow Hub 下载预训练的 BERT 模型


label_list = [0.0, 1.0, 2.0] # Label categories
max_seq_length = 64  # maximum length of (token) input sequences
train_batch_size = 32



# Get BERT layer and tokenizer:
# More details here: https://tfhub.dev/tensorflow/bert_en_uncased_L-12_H-768_A-12/2

bert_layer = hub.KerasLayer("https://tfhub.dev/tensorflow/bert_en_uncased_L-12_H-768_A-12/2",
                            trainable= True)
vocab_file = bert_layer.resolved_object.vocab_file.asset_path.numpy()
do_lower_case = bert_layer.resolved_object.do_lower_case.numpy()
tokenizer = tokenization.FullTokenizer(vocab_file, do_lower_case)

tokenizer.wordpiece_tokenizer.tokenize('Hi, how are you doing?')

为 BERT 对文本进行标记化和预处理

# This provides a function to convert row to input features and label

def to_feature(text, label, label_list=label_list, max_seq_length=max_seq_length, tokenizer=tokenizer):
  example = classifier_data_lib.InputExample(quit = None,
                                             text_a= text.numpy(),
                                             text_b= None,
                                             label= label.numpy())
  feature = classifier_data_lib.convert_single_example(0, example, label_list, max_seq_length,tokenizer)
  return (feature.input_ids, feature.input_mask, feature.segment_ids, feature.label_id )

将 Python 函数包装到 TensorFlow 操作中以实现急切执行

def to_feature_map(text, label):
  input_ids, input_mask, input_type_ids, label_id = tf.py_function(to_feature,
                                                                inp = [text, label],
                                                                Tout = [tf.int32,
                                                                        tf.int32,
                                                                        tf.int32,
                                                                        tf.int32])
  input_ids.set_shape([max_seq_length])
  input_mask.set_shape([max_seq_length])
  input_type_ids.set_shape([max_seq_length])
  label_id.set_shape([])

  x= {
      'input_word_ids':input_ids,
      'input_mask' : input_mask,
      'input_type_ids' : input_type_ids,

  }
  return (x, label_id)

使用 tf.data

创建 TensorFlow 输入管道

with tf.device('/cpu:0'):
  train_data = (train_data.map(to_feature_map,
                               num_parallel_calls = tf.data.experimental.AUTOTUNE)
  .shuffle(1000)
  .batch(32, drop_remainder = True)
  .prefetch(tf.data.experimental.AUTOTUNE))
 

 
  valid_data = (valid_data.map(to_feature_map,
                               num_parallel_calls = tf.data.experimental.AUTOTUNE)
  .shuffle(1000)
  .batch(32, drop_remainder = True)
  .prefetch(tf.data.experimental.AUTOTUNE))

向 BERT 层添加分类头

def create_model():
  input_word_ids = tf.keras.layers.Input(shape=(max_seq_length,), dtype=tf.int32,
                                       name="input_word_ids")
  input_mask = tf.keras.layers.Input(shape=(max_seq_length,), dtype=tf.int32,
                                   name="input_mask")
  input_type_ids = tf.keras.layers.Input(shape=(max_seq_length,), dtype=tf.int32,
                                    name="input_type_ids")
  
  pooled_output, sequence_output = bert_layer([input_word_ids, input_mask, input_type_ids])

  drop = tf.keras.layers.Dropout(0.5)(pooled_output)
  output = tf.keras.layers.Dense(3, activation='softmax',name = "output")(drop)

  model = tf.keras.Model(
      inputs = {
          'input_word_ids':input_word_ids,
          'input_mask' : input_mask,
          'input_type_ids' : input_type_ids},
           outputs = output)
  return model

为文本分类微调 BERT



model = create_model()

model.compile(optimizer = tf.keras.optimizers.Adam(learning_rate= 2e-5),
              loss = tf.keras.losses.sparse_categorical_crossentropy(),
              metrics = [tf.keras.metrics.categorical_accuracy()])
model.summary

Answer 1

使用:

解决

model = create_model()

model.compile(optimizer = tf.keras.optimizers.Adam(learning_rate= 2e-5),
              loss = tf.keras.losses.SparseCategoricalCrossentropy(),
              metrics = [tf.keras.metrics.Accuracy()])
model.summary()

sparse_categorical_crossentropy() 缺少 2 个必需的位置参数：'y_true' 和 'y_pred'

sparse_categorical_crossentropy() missing 2 required positional arguments: 'y_true' and 'y_pred'

python

nlp

deep-learning

lstm

bert-language-model