将训练数组输入 tensorflow 时出错答案

【问题标题】：Errors feeding training arrays into tensorflow将训练数组输入 tensorflow 时出错
【发布时间】：2020-09-10 17:06:20
【问题描述】：

我正在尝试训练神经网络，但我不断收到输入错误。

在我的问题中，X 被定义为 286784 个 268d 数组。每个数据点都有一个 int 标签，我将其转换为分类标签。我正在尝试分类 14 个类。请参阅下面的详细信息：

X_train.shape = (286784,)
X_train[0].shape = (268,)
y_train.shape = (286784, 14)
y_train[0].shape = (14,)

X_test.shape = (71696,)
X_test[0].shape = (268,)
y_test.shape = (71696, 14)
y_test[0].shape = (14,)

要创建数据并训练我的模型，这是我的代码：

from ast import literal_eval

df = pd.read_csv('data.csv')
# convert string representation to list
df['X'] = df.X.apply(lambda x: literal_eval(str(x)))

# convert each item (list) to numpy array
def convert_to_numpy(df):
  return np.asarray(df.X)
df['X'] = df.apply(convert_to_numpy, axis=1)

# train test split
X_train, X_test, y_train, y_test = train_test_split(df.X.values, df.y.values, test_size = 0.2, random_state = 42)

# convert y to categorical
y_train, y_test = to_categorical(y_train, dtype='float32'), to_categorical(y_test, dtype='float32')


# build model
def compile_and_fit(X_train, y_train, X_val, y_val):


    model = Sequential()
    model.add(Dense(200, activation='relu'))
    model.add(Dense(1000, activation='relu'))
    model.add(Dropout(0.3))
    model.add(Dense(3, activation='softmax'))

    opt = keras.optimizers.Adam(learning_rate=0.005)
    

    model.compile(optimizer=opt,
                  loss='categorical_crossentropy',
                  metrics=['accuracy'])
    model_history = model.fit(X_train,
              y_train,
              epochs=1,
              batch_size=32,
              validation_data=(X_val, y_val))
    print(model.summary())
  
    return model_history, model

# run model
hist, model = compile_and_fit(X_train,
                          y_train,
                          X_test,
                          y_test)

然后我得到这个错误：

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-16-c6305c0e6184> in <module>()
      3                           y_train,
      4                           X_test,
----> 5                           y_test)
      6 
      7 # visualize training

14 frames
<ipython-input-15-c22b1f6df674> in compile_and_fit(X_train, y_train, X_val, y_val)
     18               epochs=1,
     19               batch_size=32,
---> 20               validation_data=(X_val, y_val))
     21     print(model.summary())
     22 

/usr/local/lib/python3.6/dist-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.

/usr/local/lib/python3.6/dist-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.

/usr/local/lib/python3.6/dist-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()

/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/data_adapter.py in __init__(self, x, y, sample_weights, sample_weight_modes, batch_size, epochs, steps, shuffle, **kwargs)
    263                **kwargs):
    264     super(TensorLikeDataAdapter, self).__init__(x, y, **kwargs)
--> 265     x, y, sample_weights = _process_tensorlike((x, y, sample_weights))
    266     sample_weight_modes = broadcast_sample_weight_modes(
    267         sample_weights, sample_weight_modes)

/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/data_adapter.py in _process_tensorlike(inputs)
   1019     return x
   1020 
-> 1021   inputs = nest.map_structure(_convert_numpy_and_scipy, inputs)
   1022   return nest.list_to_tuple(inputs)
   1023 

/usr/local/lib/python3.6/dist-packages/tensorflow/python/util/nest.py in map_structure(func, *structure, **kwargs)
    633 
    634   return pack_sequence_as(
--> 635       structure[0], [func(*x) for x in entries],
    636       expand_composites=expand_composites)
    637 

/usr/local/lib/python3.6/dist-packages/tensorflow/python/util/nest.py in <listcomp>(.0)
    633 
    634   return pack_sequence_as(
--> 635       structure[0], [func(*x) for x in entries],
    636       expand_composites=expand_composites)
    637 

/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/data_adapter.py in _convert_numpy_and_scipy(x)
   1014       if issubclass(x.dtype.type, np.floating):
   1015         dtype = backend.floatx()
-> 1016       return ops.convert_to_tensor(x, dtype=dtype)
   1017     elif scipy_sparse and scipy_sparse.issparse(x):
   1018       return _scipy_sparse_to_sparse_tensor(x)

/usr/local/lib/python3.6/dist-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:

/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/tensor_conversion_registry.py in _default_conversion_function(***failed resolving arguments***)
     50 def _default_conversion_function(value, dtype, name, as_ref):
     51   del as_ref  # Unused.
---> 52   return constant_op.constant(value, dtype, name=name)
     53 
     54 

/usr/local/lib/python3.6/dist-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 

/usr/local/lib/python3.6/dist-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()

/usr/local/lib/python3.6/dist-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

/usr/local/lib/python3.6/dist-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 numpy.ndarray).

如果这有什么不同的话，我正在使用 colab 进行培训。知道我在这里做错了什么吗？

【问题讨论】：

你试过用tf.convert_to_tensor转换你的输入吗？
@runDOSrun 我试过了，我得到了这个错误：ValueError: setting an array element with a sequence.

标签： python tensorflow keras deep-learning neural-network

【解决方案1】：

你能把参差不齐的张量转换成张量并用于进一步处理吗？

例如：

import tensorflow as tf

mylist = [
    [[1, 2, 3], [5, 4, 6]], 
    [[2, 8, 9]]
]

rt = tf.ragged.constant(mylist)

features_set = rt.to_tensor()

features_set

输出：

<tf.Tensor: shape=(2, 2, 3), dtype=int32, numpy=
array([[[1, 2, 3],
        [5, 4, 6]],

       [[2, 8, 9],
        [0, 0, 0]]], dtype=int32)>

我能够使用以下虚拟代码在 Tensorflow Version 2.x 中重现您的错误 -

重现错误的代码 -

%tensorflow_version 2.x
import tensorflow as tf
print(tf.__version__)
import numpy as np
from numpy import loadtxt
from tensorflow.keras.models import Model
from tensorflow.keras.layers import Dense, Input, Concatenate

input1 = Input(shape=(3,))

# define model
x = Dense(12, input_shape = (2,), activation='relu')(input1)
x = Dense(8, activation='relu')(x)
x = Dense(1, activation='sigmoid')(x)

model = Model(inputs=input1, outputs=x)

# compile model
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])

# Model Summary
model.summary()

features_set = [
    [[1, 2, 3], [5, 4, 6]], 
    [[2, 8, 9]]
]

labels = [5, 8]

# Fit the model
model.fit(x=features_set, y=labels, epochs=150, batch_size=10, verbose=0)

输出 -

2.3.0
Model: "functional_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_1 (InputLayer)         [(None, 3)]               0         
_________________________________________________________________
dense (Dense)                (None, 12)                48        
_________________________________________________________________
dense_1 (Dense)              (None, 8)                 104       
_________________________________________________________________
dense_2 (Dense)              (None, 1)                 9         
=================================================================
Total params: 161
Trainable params: 161
Non-trainable params: 0
_________________________________________________________________
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-1-ecade355bf68> in <module>()
     35 
     36 # Fit the model
---> 37 model.fit(x=features_set, y=labels, epochs=150, batch_size=10, verbose=0)

14 frames
/usr/local/lib/python3.6/dist-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 list).

要修复此错误，请使用 tf.ragged.constant 将 list 转换为不规则张量。之后将参差不齐的张量转换为张量并在模型中使用。

固定代码 -

%tensorflow_version 2.x
import tensorflow as tf
print(tf.__version__)
import numpy as np
from numpy import loadtxt
from tensorflow.keras.models import Model
from tensorflow.keras.layers import Dense, Input, Concatenate

input1 = Input(shape=(3,))

# define model
x = Dense(12, input_shape = (2,), activation='relu')(input1)
x = Dense(8, activation='relu')(x)
x = Dense(1, activation='sigmoid')(x)

model = Model(inputs=input1, outputs=x)

# compile model
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])

# Model Summary
model.summary()


rt = tf.ragged.constant([
    [[1, 2, 3], [5, 4, 6]], 
    [[2, 8, 9]]
])

features_set = rt.to_tensor()

labels = np.asarray([5, 8])

# Fit the model
model.fit(x=features_set, y=labels, epochs=150, batch_size=10, verbose=0)

输出 -

2.3.0
Model: "functional_35"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_18 (InputLayer)        [(None, 3)]               0         
_________________________________________________________________
dense_51 (Dense)             (None, 12)                48        
_________________________________________________________________
dense_52 (Dense)             (None, 8)                 104       
_________________________________________________________________
dense_53 (Dense)             (None, 1)                 9         
=================================================================
Total params: 161
Trainable params: 161
Non-trainable params: 0
_________________________________________________________________
WARNING:tensorflow:Model was constructed with shape (None, 3) for input Tensor("input_18:0", shape=(None, 3), dtype=float32), but it was called on an input with incompatible shape (None, 2, 3).
WARNING:tensorflow:Model was constructed with shape (None, 3) for input Tensor("input_18:0", shape=(None, 3), dtype=float32), but it was called on an input with incompatible shape (None, 2, 3).
<tensorflow.python.keras.callbacks.History at 0x7f181ad9e400>

【讨论】：