Tensorflow:无法为生成TypeSpec

Question

我建立了一个模型，试图在-1到1的区间内预测句子的情绪。我想将标量(1个神经元)输出更改为3个神经元%的输出( 0%负，100%中性，0%正)。为此，我通过以下方式更改了我的Y值：

def encodeSentiment(sentiment):
    if sentiment == "negative":
        return tf.one_hot(0, 3)
    if sentiment == "neutral":
        return tf.one_hot(1, 3)
    if sentiment == "positive":
        return tf.one_hot(2, 3)
    assert False

这样做之后，我的Y看起来像这样：

​

我刚刚将输出图层从tf.keras.layers.Dense(1, activation='tanh')更改为tf.keras.layers.Dense(3, activation='softmax')

model = tf.keras.Sequential([
    encoder,
    tf.keras.layers.Embedding(len(encoder.get_vocabulary()), 128, mask_zero=True),
    tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(128,  return_sequences=True)),
    tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(64)),
    tf.keras.layers.Dense(128, activation='relu'),
    tf.keras.layers.Dropout(0.5),
    tf.keras.layers.Dense(3, activation='softmax')
])

但是当我尝试训练网络时，我得到了以下错误：

TypeError: Could not build a TypeSpec for 0        (tf.Tensor(0.0, shape=(), dtype=float32), tf.T...
1        (tf.Tensor(1.0, shape=(), dtype=float32), tf.T...
2        (tf.Tensor(1.0, shape=(), dtype=float32), tf.T...
3        (tf.Tensor(1.0, shape=(), dtype=float32), tf.T...
4        (tf.Tensor(1.0, shape=(), dtype=float32), tf.T...
                               ...                        
27476    (tf.Tensor(1.0, shape=(), dtype=float32), tf.T...
27477    (tf.Tensor(1.0, shape=(), dtype=float32), tf.T...
27478    (tf.Tensor(0.0, shape=(), dtype=float32), tf.T...
27479    (tf.Tensor(0.0, shape=(), dtype=float32), tf.T...
27480    (tf.Tensor(0.0, shape=(), dtype=float32), tf.T...
Name: sentiment_encode, Length: 27481, dtype: object with type Series

During handling of the above exception, another exception occurred:

ValueError                                Traceback (most recent call last)
<ipython-input-105-b16132770690> in <module>
     19                     validation_steps=30,
     20                     verbose=2, #output 1x per epoch
---> 21                     callbacks=[earlyStopping])

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\keras\engine\training.py in _method_wrapper(self, *args, **kwargs)
    106   def _method_wrapper(self, *args, **kwargs):
    107     if not self._in_multi_worker_mode():  # pylint: disable=protected-access
--> 108       return method(self, *args, **kwargs)
    109 
    110     # Running inside `run_distribute_coordinator` already.

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\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, validation_batch_size, validation_freq, max_queue_size, workers, use_multiprocessing)
   1061           use_multiprocessing=use_multiprocessing,
   1062           model=self,
-> 1063           steps_per_execution=self._steps_per_execution)
   1064 
   1065       # Container that configures and calls `tf.keras.Callback`s.

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\keras\engine\data_adapter.py in __init__(self, x, y, sample_weight, batch_size, steps_per_epoch, initial_epoch, epochs, shuffle, class_weight, max_queue_size, workers, use_multiprocessing, model, steps_per_execution)
   1115         use_multiprocessing=use_multiprocessing,
   1116         distribution_strategy=ds_context.get_strategy(),
-> 1117         model=model)
   1118 
   1119     strategy = ds_context.get_strategy()

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\keras\engine\data_adapter.py in __init__(self, x, y, sample_weights, sample_weight_modes, batch_size, epochs, steps, shuffle, **kwargs)
    362     indices_dataset = indices_dataset.flat_map(slice_batch_indices)
    363 
--> 364     dataset = self.slice_inputs(indices_dataset, inputs)
    365 
    366     if shuffle == "batch":

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\keras\engine\data_adapter.py in slice_inputs(self, indices_dataset, inputs)
    388     dataset = dataset_ops.DatasetV2.zip((
    389         indices_dataset,
--> 390         dataset_ops.DatasetV2.from_tensors(inputs).repeat()
    391     ))
    392 

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\data\ops\dataset_ops.py in from_tensors(tensors)
    602       Dataset: A `Dataset`.
    603     """
--> 604     return TensorDataset(tensors)
    605 
    606   @staticmethod

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\data\ops\dataset_ops.py in __init__(self, element)
   2980   def __init__(self, element):
   2981     """See `Dataset.from_tensors()` for details."""
-> 2982     element = structure.normalize_element(element)
   2983     self._structure = structure.type_spec_from_value(element)
   2984     self._tensors = structure.to_tensor_list(self._structure, element)

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\data\util\structure.py in normalize_element(element)
     96         # the value. As a fallback try converting the value to a tensor.
     97         normalized_components.append(
---> 98             ops.convert_to_tensor(t, name="component_%d" % i))
     99       else:
    100         if isinstance(spec, sparse_tensor.SparseTensorSpec):

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\framework\ops.py in convert_to_tensor(value, dtype, name, as_ref, preferred_dtype, dtype_hint, ctx, accepted_result_types)
   1497 
   1498     if ret is None:
-> 1499       ret = conversion_func(value, dtype=dtype, name=name, as_ref=as_ref)
   1500 
   1501     if ret is NotImplemented:

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\framework\constant_op.py in _constant_tensor_conversion_function(v, dtype, name, as_ref)
    336                                          as_ref=False):
    337   _ = as_ref
--> 338   return constant(v, dtype=dtype, name=name)
    339 
    340 

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\framework\constant_op.py in constant(value, dtype, shape, name)
    262   """
    263   return _constant_impl(value, dtype, shape, name, verify_shape=False,
--> 264                         allow_broadcast=True)
    265 
    266 

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\framework\constant_op.py in _constant_impl(value, dtype, shape, name, verify_shape, allow_broadcast)
    273       with trace.Trace("tf.constant"):
    274         return _constant_eager_impl(ctx, value, dtype, shape, verify_shape)
--> 275     return _constant_eager_impl(ctx, value, dtype, shape, verify_shape)
    276 
    277   g = ops.get_default_graph()

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\framework\constant_op.py in _constant_eager_impl(ctx, value, dtype, shape, verify_shape)
    298 def _constant_eager_impl(ctx, value, dtype, shape, verify_shape):
    299   """Implementation of eager constant."""
--> 300   t = convert_to_eager_tensor(value, ctx, dtype)
    301   if shape is None:
    302     return t

~\AppData\Roaming\Python\Python37\site-packages\tensorflow\python\framework\constant_op.py in convert_to_eager_tensor(value, ctx, dtype)
     96       dtype = dtypes.as_dtype(dtype).as_datatype_enum
     97   ctx.ensure_initialized()
---> 98   return ops.EagerTensor(value, ctx.device_name, dtype)
     99 
    100 

ValueError: Failed to convert a NumPy array to a Tensor (Unsupported object type tensorflow.python.framework.ops.EagerTensor).

model.compile(loss=tf.keras.losses.CategoricalCrossentropy(), 
            optimizer=tf.keras.optimizers.Adam(1e-4),
            metrics=['accuracy'])

history = model.fit(x=trainX, y=trainY,
                    validation_data=(valX, valY),
                    epochs=100, #1k
                    validation_steps=30,
                    verbose=2, #output 1x per epoch
                    callbacks=[earlyStopping])

我是Tensorflow的新手，你们能帮我找到解决这个问题的方法吗？

高级中的Ty

Answer 1

看起来TrainY是一个列表。您的目标可以是Numpy数组或TensorFlow张量。它应该与x一致(你不能有Numpy输入和张量目标，或者相反)