Tensorflow 2中的fit方法中使用Dataset和ndarray有什么区别？

debugcn 发表于 Dev

Zhongzheng_11

作为TF的新手，我对在训练模型时使用BatchDataset感到有些困惑。

让我们以MNIST为例。在此分类任务中，我们可以加载数据并将x_trian，y_train的ndarray直接输入模型中。

mnist = tf.keras.datasets.mnist

(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0

model = tf.keras.models.Sequential([
  tf.keras.layers.Flatten(input_shape=(28, 28)),
  tf.keras.layers.Dense(128, activation='relu'),
  tf.keras.layers.Dropout(0.2),
  tf.keras.layers.Dense(10, activation='softmax')
])

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

model.fit(x_train,y_train, epochs=5)

培训结果是：

Epoch 1/5
2021-02-17 15:43:02.621749: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cublas64_10.dll
   1/1875 [..............................] - ETA: 0s - loss: 2.2977 - accuracy: 0.0938WARNING:tensorflow:Callbacks method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0000s vs `on_train_batch_end` time: 0.0010s). Check your callbacks.
1875/1875 [==============================] - 2s 1ms/step - loss: 0.3047 - accuracy: 0.9117
Epoch 2/5
1875/1875 [==============================] - 2s 1ms/step - loss: 0.1473 - accuracy: 0.9569
Epoch 3/5
1875/1875 [==============================] - 2s 1ms/step - loss: 0.1097 - accuracy: 0.9673
Epoch 4/5
1875/1875 [==============================] - 2s 1ms/step - loss: 0.0905 - accuracy: 0.9724
Epoch 5/5
1875/1875 [==============================] - 2s 1ms/step - loss: 0.0759 - accuracy: 0.9764

我们还可以使用tf.data.Dataset.from_tensor_slices生成BatchDataset并将其输入以适合函数。

mnist = tf.keras.datasets.mnist

(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0

train_ds = tf.data.Dataset.from_tensor_slices(
    (x_train, y_train)).shuffle(10000).batch(32)

test_ds = tf.data.Dataset.from_tensor_slices((x_test, y_test)).batch(32)

model = tf.keras.models.Sequential([
  tf.keras.layers.Flatten(input_shape=(28, 28)),
  tf.keras.layers.Dense(128, activation='relu'),
  tf.keras.layers.Dropout(0.2),
  tf.keras.layers.Dense(10, activation='softmax')
])

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

model.fit(train_ds, epochs=5)

训练过程的结果如下。

Epoch 1/5
2021-02-17 15:30:34.698718: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cublas64_10.dll
1875/1875 [==============================] - 3s 1ms/step - loss: 0.2969 - accuracy: 0.9140
Epoch 2/5
1875/1875 [==============================] - 3s 1ms/step - loss: 0.1462 - accuracy: 0.9566
Epoch 3/5
1875/1875 [==============================] - 3s 1ms/step - loss: 0.1087 - accuracy: 0.9669
Epoch 4/5
1875/1875 [==============================] - 3s 1ms/step - loss: 0.0881 - accuracy: 0.9730
Epoch 5/5
1875/1875 [==============================] - 3s 1ms/step - loss: 0.0765 - accuracy: 0.9759

可以使用两种方法成功训练模型，但是两者之间有什么区别吗？使用数据集进行培训是否还有其他优势？如果在这种情况下这两种方法之间没有区别，那么生成用于训练的数据集的典型用法是什么？何时应使用此方法？

谢谢你。

因纳特

使用时Model.fit(x=None, y=None, ...，我们可以将训练对参数传递为纯numpy数组或keras.utils.Sequence或tf.data。

当我们按如下方式使用时，我们将每个训练对（x和y）作为直接的numpy数组分别传递给该fit函数。

# data 
(x_train, y_train), (_, _) = tf.keras.datasets.mnist.load_data()

# fit
model.fit(x = x_train, y = y_train, ... 

# check
print(x_train.shape, y_train.shape)
print(type(x_train), type(y_train))

# (60000, 28, 28) (60000,)
# <class 'numpy.ndarray'> <class 'numpy.ndarray'>

在另一方面tf.data和Sequence我们通过培训对作为元组的形状，目前仍是数据类型是ndarray。根据文档，

一个tf.data数据集。应该返回的任何一个元组（inputs，targets）
生成器或keras.utils.Sequence返回（inputs，targets）

# data
train_ds = tf.data.Dataset.from_tensor_slices((x_train, y_train)).shuffle(10000).batch(2)

# check
next(iter(train_ds))

(<tf.Tensor: shape=(2, 28, 28), dtype=uint8, numpy= array([[[...], [[...]]], dtype=uint8)>,
 <tf.Tensor: shape=(2,), dtype=uint8, numpy=array([7, 8], dtype=uint8)>)

这就是为什么，如果x是tf.data，generator或keras.utils.Sequence实例，y不应该被指定（因为目标将从中获得x）。