Tensorflow Dataset API

------------恢复内容开始------------

The tf.data.Dataset API supports writing descriptive and efficient input pipelines. Dataset usage follows a common pattern:

  1. Create a source dataset from your input data.
  2. Apply dataset transformations to preprocess the data.
  3. Iterate over the dataset and process the elements.

------------恢复内容结束------------

Methods:

 

1. batch 

1 batch(
2     batch_size, drop_remainder=False
3 )

Combines consecutive elements of this dataset into batches.

把当前的dataset分割成连续的batches

drop_remainder = true时把最后一组那些剩下的，除不尽的元素舍弃

1 dataset = tf.data.Dataset.range(8)
2 dataset = dataset.batch(3)
3 list(dataset.as_numpy_iterator())

输出：Array[[0, 1, 2], [3, 4, 5], [6, 7]]

 

1 dataset = tf.data.Dataset.range(8)
2 dataset = dataset.batch(3, drop_remainder=True)
3 list(dataset.as_numpy_iterator())

输出：Array[[0, 1, 2], [3, 4, 5]]

 

posted @ 2020-10-20 20:11  Noncoretime  阅读(192)  评论(0)  编辑  收藏  举报