Tensorflow TF 2.0 Keras拟合_生成器：数据_生成器输出错误形状

我正在尝试使用TensorFlow2.0和tf.kerasAPI来训练图像字幕模型。我使用的数据集是Flickr 8K数据集，虽然我的计算机可以在RAM中保存整个数据集，但我希望使用fit_generator和data_generator来批量加载和准备数据（因为一旦我可以使用此数据集，我将尝试使用更大的数据集来训练此模型）

我对数据和模型定义进行预处理的方式还可以。我可以对手动生成的批处理执行model.predict（），模型输出预期的数据形状，并且没有错误。我还可以手动使用data_generator准备完整的数据集，并使用model.fit（）使用整个数据。它工作正常，模型训练没有错误

当我尝试使用fit_generator进行训练时出现问题，该生成器将输出此错误（文章末尾的全长错误输出）：

如果我单独调用generator函数来检查生成批次的类型和形状，在我看来一切正常：

generator = data_generator(train_descriptions, train_features, wordtoix, max_length, number_pics_per_bath)
data = next(generator)

print("Total items in data: ", len(data))

# Data[1] is the encoded Y
print("Encodded Y shape: ", data[1].shape)
print("Example Y: ", data[1][0])

# Data[0] is a list of [image_feature, encoded_caption]
print("X1 shape (image feature): ", data[0][0].shape)
print("X2 shape (image caption): ", data[0][1].shape)

Outputs:
    -----------------------------------
    Total items in data:  2
    Encodded Y shape:  (168, 1652)
    Example Y:  [0. 0. 1. ... 0. 0. 0.]
    X1 shape (image feature):  (168, 2048)
    X2 shape (image caption):  (168, 34)

这是数据发生器功能的代码：

# data generator, intended to be used in a call to model.fit_generator()
# $descriptions: a dictionary containing <image_id> -> [ text_captions_list ]
# photos: list of numpy arrays representing image features
# wordtoix: a dictionary to convert words to word_codes (integers)
# max_length: maximum word count for a caption
def data_generator(descriptions, photos, wordtoix, max_length, num_photos_per_batch):
    X1, X2, y = list(), list(), list()
    n=0
    # loop for ever over images
    while 1:
        for key, desc_list in descriptions.items():
            n+=1
            # retrieve the photo feature
            photo = photos[key+'.jpg']
            for desc in desc_list:
                # encode the sequence
                seq = [wordtoix[word] for word in desc.split(' ') if word in wordtoix]
                # split one sequence into multiple X, y pairs
                for i in range(1, len(seq)):
                    # split into input and output pair
                    in_seq, out_seq = seq[:i], seq[i]
                    # pad input sequence
                    in_seq = pad_sequences([in_seq], maxlen=max_length)[0]
                    # encode output sequence
                    out_seq = to_categorical([out_seq], num_classes=vocab_size)[0]
                    # store
                    X1.append(photo)
                    X2.append(in_seq)
                    y.append(out_seq)
            # yield the batch data
            if n==num_photos_per_batch:
                yield [[array(X1), array(X2)], array(y)]

                X1, X2, y = list(), list(), list()
                n=0

我使用带有imagenet预训练权重的inceptionv3模型从图像生成特征（然后保存到磁盘）

然后我使用这个需要“两个输入”的模型组件：图像特征数组和编码图像标题：

fit_generator的全长误差如下：

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-73-10ea3905954d> in <module>
      1 for i in range(epochs):
      2     generator = data_generator(train_descriptions, train_features, wordtoix, max_length, number_pics_per_bath)
----> 3     model.fit_generator(generator, epochs=1, steps_per_epoch=steps, verbose=1)
      4     model.save('./saved/model_' + str(i) + '.h5')

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training.py in fit_generator(self, generator, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, validation_freq, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch)
   1295         shuffle=shuffle,
   1296         initial_epoch=initial_epoch,
-> 1297         steps_name='steps_per_epoch')
   1298 
   1299   def evaluate_generator(self,

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training_generator.py in model_iteration(model, data, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, validation_freq, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch, mode, batch_size, steps_name, **kwargs)
    263 
    264       is_deferred = not model._is_compiled
--> 265       batch_outs = batch_function(*batch_data)
    266       if not isinstance(batch_outs, list):
    267         batch_outs = [batch_outs]

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training.py in train_on_batch(self, x, y, sample_weight, class_weight, reset_metrics)
    971       outputs = training_v2_utils.train_on_batch(
    972           self, x, y=y, sample_weight=sample_weight,
--> 973           class_weight=class_weight, reset_metrics=reset_metrics)
    974       outputs = (outputs['total_loss'] + outputs['output_losses'] +
    975                  outputs['metrics'])

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training_v2_utils.py in train_on_batch(model, x, y, sample_weight, class_weight, reset_metrics)
    251   x, y, sample_weights = model._standardize_user_data(
    252       x, y, sample_weight=sample_weight, class_weight=class_weight,
--> 253       extract_tensors_from_dataset=True)
    254   batch_size = array_ops.shape(nest.flatten(x, expand_composites=True)[0])[0]
    255   # If `model._distribution_strategy` is True, then we are in a replica context

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training.py in _standardize_user_data(self, x, y, sample_weight, class_weight, batch_size, check_steps, steps_name, steps, validation_split, shuffle, extract_tensors_from_dataset)
   2470           feed_input_shapes,
   2471           check_batch_axis=False,  # Don't enforce the batch size.
-> 2472           exception_prefix='input')
   2473 
   2474     # Get typespecs for the input data and sanitize it if necessary.

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training_utils.py in standardize_input_data(data, names, shapes, check_batch_axis, exception_prefix)
    504   elif isinstance(data, (list, tuple)):
    505     if isinstance(data[0], (list, tuple)):
--> 506       data = [np.asarray(d) for d in data]
    507     elif len(names) == 1 and isinstance(data[0], (float, int)):
    508       data = [np.asarray(data)]

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training_utils.py in <listcomp>(.0)
    504   elif isinstance(data, (list, tuple)):
    505     if isinstance(data[0], (list, tuple)):
--> 506       data = [np.asarray(d) for d in data]
    507     elif len(names) == 1 and isinstance(data[0], (float, int)):
    508       data = [np.asarray(data)]

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/numpy/core/_asarray.py in asarray(a, dtype, order)
     83 
     84     """
---> 85     return array(a, dtype, copy=False, order=order)
     86 
     87 

ValueError: could not broadcast input array from shape (168,2048) into shape (168)

---------------------------------------------------------------------------
ValueError回溯（最近一次调用上次）
在里面
1表示范围内的i（历元）：
2发生器=数据发生器（序列描述、序列特征、文字、最大长度、每个浴池的图片数量）
---->3.模型拟合生成器（生成器，历元=1，每历元步长=1，详细度=1）
4型号保存（'./已保存/型号_'+str（i）+'.h5'）
~/anaconda3/envs/tf2 gpu/lib/python3.6/site-packages/tensorflow\u core/python/keras/engine/training.py in-fit\u生成器（self、生成器、每个历元的步骤、历元、冗余、回调、验证数据、验证步骤、验证频率、类权重、最大队列大小、工作者、使用多处理、无序、初始历元）
1295洗牌=洗牌，
1296初始纪元=初始纪元，
->1297个步骤（每个时代的步骤）
1298
1299 def评估_发生器（自，
在模型迭代中~/anaconda3/envs/tf2 gpu/lib/python3.6/site-packages/tensorflow\u core/python/keras/engine/training\u generator.py（模型、数据、每个历元的步骤、历元、冗余、回调、验证数据、验证步骤、验证频率、类权重、最大队列大小、工人、使用多处理、随机、初始历元、模式、批大小、步骤名称、**kwargs）
263
264是延迟的=不是模型。\是编译的
-->265批处理输出=批处理功能（*批处理数据）
266如果不存在（批次，列表）：
267批次输出=[批次输出]
~/anaconda3/envs/tf2 gpu/lib/python3.6/site-packages/tensorflow\u core/python/keras/engine/training.py in-train\u-on\u批次（self，x，y，sample\u-weight，class\u-weight，reset\u-metrics）
971输出=批次上的培训(
972自身，x，y=y，样本重量=样本重量，
-->973类权重=类权重，重置度量=重置度量）
974输出=（输出['总损耗]+输出['输出损耗']+
975个输出[“指标]）
~/anaconda3/envs/tf2 gpu/lib/python3.6/site-packages/tensorflow\u core/python/keras/engine/training\u v2\u utils.py in train\u on\u批次（型号、x、y、样本权重、类权重、重置度量）
251 x，y，样本权重=模型。标准化用户数据(
252 x，y，样本重量=样本重量，类别重量=类别重量，
-->253从_数据集中提取_张量_=True）
254批次大小=数组操作形状（嵌套.展平（x，展开组合=真）[0]）[0]
255#如果“model._distribution_strategy”为真，则我们处于复制上下文中
~/anaconda3/envs/tf2 gpu/lib/python3.6/site-packages/tensorflow\u core/python/keras/engine/training.py in\u-standard\u-user\u数据（self、x、y、sample\u-weight、class\u-weight、batch\u-size、check\u-steps、steps\u-name、steps、validation\u-split、shuffle\u、shufflue、extract\u-tensors\u-from\u-data
2470个进纸输入形状，
2471检查_batch_axis=False，#不强制执行批大小。
->2472异常（前缀为“输入”）
2473
2474#获取输入数据的类型规范，必要时对其进行清理。
标准化输入数据中的~/anaconda3/envs/tf2 gpu/lib/python3.6/site-packages/tensorflow\u core/python/keras/engine/training\u utils.py（数据、名称、形状、检查批处理轴、异常前缀）
504 elif isinstance（数据，（列表，元组））：
505如果isinstance（数据[0]，（列表，元组））：
-->506数据=[np.asarray（d）表示数据中的d]
507 elif len（名称）==1和isinstance（数据[0]，（浮点，int））：
508数据=[np.asarray（数据）]
~/anaconda3/envs/tf2 gpu/lib/python3.6/site-packages/tensorflow\u core/python/keras/engine/training\u utils.py in（.0）
504 elif isinstance（数据，（列表，元组））：
505如果isinstance（数据[0]，（列表，元组））：
-->506数据=[np.asarray（d）表示数据中的d]
507 elif len（名称）==1和isinstance（数据[0]，（浮点，int））：
508数据=[np.asarray（数据）]
asarray中的~/anaconda3/envs/tf2 gpu/lib/python3.6/site-packages/numpy/core//u asarray.py（a，数据类型，顺序）
83
84     """
--->85返回数组（a，数据类型，副本=False，顺序=order）
86
87
ValueError:无法将输入数组从形状（1682048）广播到形状（168）

---

提前感谢您的帮助！

def data_generator应该有一个返回语句。

有同样的问题，您最终解决了吗？还没有，我从Keras方法转向了更纯粹的TF 2.0方法（使用TF.data.Dataset）

epochs = 20
steps = len(train_descriptions)
for i in range(epochs):
    generator = data_generator(train_descriptions, train_features, wordtoix, max_length)
    model.fit_generator(generator, epochs=1, steps_per_epoch=steps, verbose=1)
    model.save('./saved/model_' + str(i) + '.h5')
for i in range(epochs):
    generator = data_generator(train_descriptions, train_features, wordtoix, max_length, number_pics_per_bath)
    model.fit_generator(generator, epochs=1, steps_per_epoch=steps, verbose=1)
    model.save('./saved/model_' + str(i) + '.h5')

inputs1 = tf.keras.Input(shape=(2048,))
fe1 = tf.keras.layers.Dropout(0.5)(inputs1)
fe2 = tf.keras.layers.Dense(256, activation='relu')(fe1)

inputs2 = tf.keras.Input(shape=(max_length,))
se1 = tf.keras.layers.Embedding(vocab_size, embedding_dim, mask_zero=True)(inputs2)
se2 = tf.keras.layers.Dropout(0.5)(se1)
se3 = tf.keras.layers.LSTM(256)(se2)

decoder1 = tf.keras.layers.concatenate([fe2, se3])
decoder2 = tf.keras.layers.Dense(256, activation='relu')(decoder1)

outputs = tf.keras.layers.Dense(vocab_size, activation='softmax')(decoder2)

model = Model(inputs=[inputs1, inputs2], outputs=outputs)

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-73-10ea3905954d> in <module>
      1 for i in range(epochs):
      2     generator = data_generator(train_descriptions, train_features, wordtoix, max_length, number_pics_per_bath)
----> 3     model.fit_generator(generator, epochs=1, steps_per_epoch=steps, verbose=1)
      4     model.save('./saved/model_' + str(i) + '.h5')

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training.py in fit_generator(self, generator, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, validation_freq, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch)
   1295         shuffle=shuffle,
   1296         initial_epoch=initial_epoch,
-> 1297         steps_name='steps_per_epoch')
   1298 
   1299   def evaluate_generator(self,

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training_generator.py in model_iteration(model, data, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, validation_freq, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch, mode, batch_size, steps_name, **kwargs)
    263 
    264       is_deferred = not model._is_compiled
--> 265       batch_outs = batch_function(*batch_data)
    266       if not isinstance(batch_outs, list):
    267         batch_outs = [batch_outs]

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training.py in train_on_batch(self, x, y, sample_weight, class_weight, reset_metrics)
    971       outputs = training_v2_utils.train_on_batch(
    972           self, x, y=y, sample_weight=sample_weight,
--> 973           class_weight=class_weight, reset_metrics=reset_metrics)
    974       outputs = (outputs['total_loss'] + outputs['output_losses'] +
    975                  outputs['metrics'])

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training_v2_utils.py in train_on_batch(model, x, y, sample_weight, class_weight, reset_metrics)
    251   x, y, sample_weights = model._standardize_user_data(
    252       x, y, sample_weight=sample_weight, class_weight=class_weight,
--> 253       extract_tensors_from_dataset=True)
    254   batch_size = array_ops.shape(nest.flatten(x, expand_composites=True)[0])[0]
    255   # If `model._distribution_strategy` is True, then we are in a replica context

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training.py in _standardize_user_data(self, x, y, sample_weight, class_weight, batch_size, check_steps, steps_name, steps, validation_split, shuffle, extract_tensors_from_dataset)
   2470           feed_input_shapes,
   2471           check_batch_axis=False,  # Don't enforce the batch size.
-> 2472           exception_prefix='input')
   2473 
   2474     # Get typespecs for the input data and sanitize it if necessary.

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training_utils.py in standardize_input_data(data, names, shapes, check_batch_axis, exception_prefix)
    504   elif isinstance(data, (list, tuple)):
    505     if isinstance(data[0], (list, tuple)):
--> 506       data = [np.asarray(d) for d in data]
    507     elif len(names) == 1 and isinstance(data[0], (float, int)):
    508       data = [np.asarray(data)]

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/tensorflow_core/python/keras/engine/training_utils.py in <listcomp>(.0)
    504   elif isinstance(data, (list, tuple)):
    505     if isinstance(data[0], (list, tuple)):
--> 506       data = [np.asarray(d) for d in data]
    507     elif len(names) == 1 and isinstance(data[0], (float, int)):
    508       data = [np.asarray(data)]

~/anaconda3/envs/tf2-gpu/lib/python3.6/site-packages/numpy/core/_asarray.py in asarray(a, dtype, order)
     83 
     84     """
---> 85     return array(a, dtype, copy=False, order=order)
     86 
     87 

ValueError: could not broadcast input array from shape (168,2048) into shape (168)