我正在尝试使用 bert 预训练模型进行意图分类。这是我在 jupyter notebok 中的代码。
class DataPreparation:
text_column = "text"
label_column = "intent"
def __init__(self, train, test, tokenizer: FullTokenizer, classes, max_seq_len=192):
self.tokenizer = tokenizer
self.max_seq_len = 0
self.classes = classes
((self.train_x, self.train_y), (self.test_x, self.test_y)) = map(self.prepare_data, [train, test])
print("max seq_len", self.max_seq_len)
self.max_seq_len = min(self.max_seq_len, max_seq_len)
self.train_x, self.test_x = map(self.data_padding, [self.train_x, self.test_x])
def prepare_data(self, df):
x, y = [], []
for _, row in tqdm(df.iterrows()):
text, label = row[DataPreparation.text_column], row[DataPreparation.label_column]
tokens = self.tokenizer.tokenize(text)
tokens = ["[CLS]"] + tokens + ["[SEP]"]
token_ids = self.tokenizer.convert_tokens_to_ids(tokens)
self.max_seq_len = max(self.max_seq_len, len(token_ids))
x.append(token_ids)
y.append(self.classes.index(label))
return np.array(x), np.array(y)
def data_padding(self, ids):
x = []
for input_ids in ids:
input_ids = input_ids[:min(len(input_ids), self.max_seq_len - 2)]
input_ids = input_ids + [0] * (self.max_seq_len - len(input_ids))
x.append(np.array(input_ids))
return np.array(x)
tokenizer = FullTokenizer(vocab_file=os.path.join(bert_ckpt_dir, "vocab.txt"))
def model_defination(max_seq_len, bert_ckpt_file):
with tf.io.gfile.GFile(bert_config_file, "r") as reader:
bc = StockBertConfig.from_json_string(reader.read())
bert_params = map_stock_config_to_params(bc)
bert_params.adapter_size = None
bert = BertModelLayer.from_params(bert_params, name="bert")
input_ids = keras.layers.Input(shape=(max_seq_len, ), dtype='int32',name="input_ids")
bert_output = bert(input_ids)
print("bert shape", bert_output.shape)
cls_out = keras.layers.Lambda(lambda seq: seq[:, 0, :])(bert_output)
cls_out = keras.layers.Dropout(0.5)(cls_out)
logits = keras.layers.Dense(units=768, activation="tanh")(cls_out)
logits = keras.layers.Dropout(0.5)(logits)
logits = keras.layers.Dense(units=len(classes), activation="softmax")(logits)
model = keras.Model(inputs=input_ids, outputs=logits)
model.build(input_shape=(None, max_seq_len))
load_stock_weights(bert, bert_ckpt_file)
return model
classes = train.intent.unique().tolist()
data = DataPreparation(train, test, tokenizer, classes, max_seq_len=128)
data.train_x.shape
data.train_y[0]
model = model_defination(data.max_seq_len, bert_ckpt_file)
现在，当我尝试调用该函数时，出现错误。参数值 max_seq_len = 55，bert_ckpt_file = bert 检查点文件。
当我创建模型时，我收到以下错误：
TypeError Traceback (most recent call last)
<ipython-input-17-af3e534b3882> in <module>
----> 1 model = model_defination(data.max_seq_len, bert_ckpt_file)
<ipython-input-16-a83a622dafe3> in model_defination(max_seq_len, bert_ckpt_file)
9 input_ids = keras.layers.Input(shape=(max_seq_len, ), dtype='int32',name="input_ids")
10 #input_spec = tf.keras.layers.InputSpec(ndim=3)
---> 11 bert_output = bert(input_ids)
12
13 print("bert shape", bert_output.shape)
~\Anaconda3\lib\site-packages\keras\engine\base_layer.py in __call__(self, *args, **kwargs)
974 # >> model = tf.keras.Model(inputs, outputs)
975 if _in_functional_construction_mode(self, inputs, args, kwargs, input_list):
--> 976 return self._functional_construction_call(inputs, args, kwargs,
977 input_list)
978
~\Anaconda3\lib\site-packages\keras\engine\base_layer.py in _functional_construction_call(self, inputs, args, kwargs, input_list)
1112 layer=self, inputs=inputs, build_graph=True, training=training_value):
1113 # Check input assumptions set after layer building, e.g. input shape.
-> 1114 outputs = self._keras_tensor_symbolic_call(
1115 inputs, input_masks, args, kwargs)
1116
~\Anaconda3\lib\site-packages\keras\engine\base_layer.py in _keras_tensor_symbolic_call(self, inputs, input_masks, args, kwargs)
846 return tf.nest.map_structure(keras_tensor.KerasTensor, output_signature)
847 else:
--> 848 return self._infer_output_signature(inputs, args, kwargs, input_masks)
849
850 def _infer_output_signature(self, inputs, args, kwargs, input_masks):
~\Anaconda3\lib\site-packages\keras\engine\base_layer.py in _infer_output_signature(self, inputs, args, kwargs, input_masks)
886 self._maybe_build(inputs)
887 inputs = self._maybe_cast_inputs(inputs)
--> 888 outputs = call_fn(inputs, *args, **kwargs)
889
890 self._handle_activity_regularization(inputs, outputs)
~\Anaconda3\lib\site-packages\tensorflow\python\autograph\impl\api.py in wrapper(*args, **kwargs)
693 except Exception as e: # pylint:disable=broad-except
694 if hasattr(e, 'ag_error_metadata'):
--> 695 raise e.ag_error_metadata.to_exception(e)
696 else:
697 raise
TypeError: in user code:
C:\Users\kamrul.moin\Anaconda3\lib\site-packages\bert\model.py:80 call *
output = self.encoders_layer(embedding_output, mask=mask, training=training)
C:\Users\kamrul.moin\Anaconda3\lib\site-packages\keras\engine\base_layer.py:1030 __call__ **
self._maybe_build(inputs)
C:\Users\kamrul.moin\Anaconda3\lib\site-packages\keras\engine\base_layer.py:2659 _maybe_build
self.build(input_shapes) # pylint:disable=not-callable
C:\Users\kamrul.moin\Anaconda3\lib\site-packages\bert\transformer.py:209 build
self.input_spec = keras.layers.InputSpec(shape=input_shape)
C:\Users\kamrul.moin\Anaconda3\lib\site-packages\keras\engine\base_layer.py:2777 __setattr__
super(tf.__internal__.tracking.AutoTrackable, self).__setattr__(name, value) # pylint: disable=bad-super-call
C:\Users\kamrul.moin\Anaconda3\lib\site-packages\tensorflow\python\training\tracking\base.py:530 _method_wrapper
result = method(self, *args, **kwargs)
C:\Users\kamrul.moin\Anaconda3\lib\site-packages\keras\engine\base_layer.py:1296 input_spec
raise TypeError('Layer input_spec must be an instance of InputSpec. '
TypeError: Layer input_spec must be an instance of InputSpec. Got: InputSpec(shape=(None, 55, 768), ndim=3)
我已经解决了这个错误。这是由于我的训练数据的形状造成的。我将索引添加为训练数据中的一列。重置训练数据中的索引列后，我得到了正确的结果。 下面的代码解决了错误：
train = train.reset_index(drop=True)
test = test.reset_index(drop=True)
这个问题你解决了吗？我也面临着同样的问题。