python 在非Eager Execution模式下，可以使用make_one_shot_iterator（）迭代数据集

Posted 2021-05-09

import numpy as np
import tensorflow as tf

data_np = np.random.randint(0, 100, (100, 5))
dataset = tf.data.Dataset.from_tensor_slices((data_np)).shuffle(100).batch(20)
iterator = dataset.make_one_shot_iterator()
next_batch = iterator.get_next()

sess = tf.Session()                
while True:                      
    try:                        
        print(sess.run(next_batch))
    except tf.errors.OutOfRangeError:
        break