我的问题是如何从多个(或切分的)tfrecords中获取批量输入。我读过这个例子。基本流程是,以训练集为例,(1)首先从这些文件名生成一系列tfrecords(例如,train-000-of-005train-001-of-005、…),(2)生成一个列表并将其输入到tf.train.string_input_producer中以获取队列,(3)同时生成一个tf.RandomShuffleQueue以执行其他操作,(4)使用tf.train.batch_join生成批处理iNETP.
我认为这很复杂,我不确定这个过程的逻辑。在我的例子中,我有一个.npy文件的列表,我想生成切分的tfrecords(多个单独的tfrecords,而不仅仅是一个大文件)。每个.npy文件包含不同数量的阳性和阴性样本(2类)。一种基本的方法是生成一个大的tfrecord文件。但文件太大(~20Gb)。所以我求助于碎片的TFrecords。有没有更简单的方法?谢谢。

最佳答案

使用Dataset API简化整个过程。这两个部分都是:(1): Convert numpy array to tfrecords(2,3,4): read the tfrecords to generate batches
1。从numpy数组创建tfrecords:

    def npy_to_tfrecords(...):
       # write records to a tfrecords file
       writer = tf.python_io.TFRecordWriter(output_file)

       # Loop through all the features you want to write
       for ... :
          let say X is of np.array([[...][...]])
          let say y is of np.array[[0/1]]

         # Feature contains a map of string to feature proto objects
         feature = {}
         feature['X'] = tf.train.Feature(float_list=tf.train.FloatList(value=X.flatten()))
         feature['y'] = tf.train.Feature(int64_list=tf.train.Int64List(value=y))

         # Construct the Example proto object
         example = tf.train.Example(features=tf.train.Features(feature=feature))

         # Serialize the example to a string
         serialized = example.SerializeToString()

         # write the serialized objec to the disk
         writer.write(serialized)
      writer.close()

2。使用数据集API读取tfrecords(tensorflow>=1.2):
    # Creates a dataset that reads all of the examples from filenames.
    filenames = ["file1.tfrecord", "file2.tfrecord", ..."fileN.tfrecord"]
    dataset = tf.contrib.data.TFRecordDataset(filenames)
    # for version 1.5 and above use tf.data.TFRecordDataset

    # example proto decode
    def _parse_function(example_proto):
      keys_to_features = {'X':tf.FixedLenFeature((shape_of_npy_array), tf.float32),
                          'y': tf.FixedLenFeature((), tf.int64, default_value=0)}
      parsed_features = tf.parse_single_example(example_proto, keys_to_features)
     return parsed_features['X'], parsed_features['y']

    # Parse the record into tensors.
    dataset = dataset.map(_parse_function)

    # Shuffle the dataset
    dataset = dataset.shuffle(buffer_size=10000)

    # Repeat the input indefinitly
    dataset = dataset.repeat()

    # Generate batches
    dataset = dataset.batch(batch_size)

    # Create a one-shot iterator
    iterator = dataset.make_one_shot_iterator()

    # Get batch X and y
    X, y = iterator.get_next()

09-25 21:08