Question

我正在尝试在一些数据上训练LSTM网络，不幸的是，我一直遇到以下错误：
InvalidArgumentError: indices[] = is not in [0, 4704)

Train on 180596 samples, validate on 45149 samples
Epoch 1/1
---------------------------------------------------------------------------
InvalidArgumentError                      Traceback (most recent call last)
<ipython-input-52-a434c3619685> in <module>()
     14                     epochs=1,
     15                     batch_size=128,
---> 16                     validation_split=0.2)

c:\program files\python3x64\lib\site-packages\keras\engine\training.py in fit(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, **kwargs)
   1040                                         initial_epoch=initial_epoch,
   1041                                         steps_per_epoch=steps_per_epoch,
-> 1042                                         validation_steps=validation_steps)
   1043 
   1044     def evaluate(self, x=None, y=None,

c:\program files\python3x64\lib\site-packages\keras\engine\training_arrays.py in fit_loop(model, f, ins, out_labels, batch_size, epochs, verbose, callbacks, val_f, val_ins, shuffle, callback_metrics, initial_epoch, steps_per_epoch, validation_steps)
    197                     ins_batch[i] = ins_batch[i].toarray()
    198 
--> 199                 outs = f(ins_batch)
    200                 if not isinstance(outs, list):
    201                     outs = [outs]

c:\program files\python3x64\lib\site-packages\keras\backend\tensorflow_backend.py in __call__(self, inputs)
   2659                 return self._legacy_call(inputs)
   2660 
-> 2661             return self._call(inputs)
   2662         else:
   2663             if py_any(is_tensor(x) for x in inputs):

c:\program files\python3x64\lib\site-packages\keras\backend\tensorflow_backend.py in _call(self, inputs)
   2629                                 symbol_vals,
   2630                                 session)
-> 2631         fetched = self._callable_fn(*array_vals)
   2632         return fetched[:len(self.outputs)]
   2633 

c:\program files\python3x64\lib\site-packages\tensorflow\python\client\session.py in __call__(self, *args)
   1452         else:
   1453           return tf_session.TF_DeprecatedSessionRunCallable(
-> 1454               self._session._session, self._handle, args, status, None)
   1455 
   1456     def __del__(self):

c:\program files\python3x64\lib\site-packages\tensorflow\python\framework\errors_impl.py in __exit__(self, type_arg, value_arg, traceback_arg)
    517             None, None,
    518             compat.as_text(c_api.TF_Message(self.status.status)),
--> 519             c_api.TF_GetCode(self.status.status))
    520     # Delete the underlying status object from memory otherwise it stays alive
    521     # as there is a reference to status from this from the traceback due to

InvalidArgumentError: indices[62,0] = 15757 is not in [0, 4704)
     [[Node: embedding_15/embedding_lookup = GatherV2[Taxis=DT_INT32, Tindices=DT_INT32, Tparams=DT_FLOAT, _class=["loc:@training_14/RMSprop/Assign_1"], _device="/job:localhost/replica:0/task:0/device:CPU:0"](embedding_15/embeddings/read, embedding_15/Cast, training_14/RMSprop/gradients/embedding_15/embedding_lookup_grad/concat/axis)]]

我试图找到解决方案，但无济于事。
我正在使用的数据是“一次热编码”和标准化的网络流。
至少就NaN和无穷大而言，我至少应该是很好的选择。看一下输入数据：

In [14]:
print(flows_nd.shape)
print(type(flows_nd))
print(type(flows_nd[0]))

(225745, 4704)
<class 'numpy.ndarray'>
<class 'numpy.ndarray'>

这是网络本身：

from keras.models import Sequential
from keras.layers import Dense, Embedding
from keras.layers import LSTM

model = Sequential()
model.add(Embedding(flows_nd.shape[-1], 32))
model.add(LSTM(32))
model.add(Dense(1, activation='sigmoid'))

model.compile(optimizer='rmsprop',
              loss='binary_crossentropy',
              metrics=['acc'])
history = model.fit(flows_nd, enc_labels,
                    epochs=1,
                    batch_size=128,
                    validation_split=0.2)

根据我的阅读，张量流中的vocab大小可能有问题，但是我不确定我可以在Keras层中更改它。

如果有人想近距离观察，我已经组装了Jupyter Notebook with my current code。

任何帮助将不胜感激。预先感谢！

解决方案：
正如Mitiku所指出的，我是将要素的数量而不是词汇量（Keras Embedding Layer Documentation）传递给嵌入层。这就是我将代码更新为有效示例的方式：

#find the maximum vocabulary size
voc_size = (flows_scaled.max()+1).astype('int64')
print(voc_size)

# build the model
from keras.models import Sequential
from keras.layers import Dense, Embedding
from keras.layers import LSTM

model = Sequential()
model.add(Embedding(voc_size, 32)) 
model.add(LSTM(32))
model.add(Dense(1, activation='sigmoid'))

model.compile(optimizer='rmsprop',
              loss='binary_crossentropy',
              metrics=['acc'])

history = model.fit(flows_scaled, enc_labels,
                    epochs=1,
                    batch_size=128,
                    validation_split=0.2)

Answer 1

嵌入层期望第一个参数是词汇量。这意味着它是[maximum integer index + 1][1]

但是您要通过的特征数是词汇量。这是不正确的。您实际上可以做的是找到最大数量flow_nd并传递该数字。

voc_size = flows_nd.max()+1

model = Sequential()
model.add(Embedding(voc_size, 32))

使用Keras训练时的Tensorflow InvalidArgumentError（索引）

1 个答案: