tensorflow 一些常用函数的功能

tf.train.slice_input_producer(tensor_list, num_epochs=None, shuffle=True, seed=None, capacity=32, shared_name=None, name=None)

Produces a slice of each Tensor in tensor_list.

Implemented using a Queue -- a QueueRunner for the Queue is added to the current Graph's QUEUE_RUNNERcollection.

Args:

• tensor_list: A list of Tensor objects. Every Tensor in tensor_list must have the same size in the first dimension.
• num_epochs: An integer (optional). If specified, slice_input_producer produces each slice num_epochs times before generating an OutOfRange error. If not specified, slice_input_producer can cycle through the slices an unlimited number of times.
• shuffle: Boolean. If true, the integers are randomly shuffled within each epoch.
• seed: An integer (optional). Seed used if shuffle == True.
• capacity: An integer. Sets the queue capacity.
• shared_name: (optional). If set, this queue will be shared under the given name across multiple sessions.
• name: A name for the operations (optional).

Returns:

A list of tensors, one for each element of tensor_list. If the tensor in tensor_list has shape [N, a, b, .., z], then the corresponding output tensor will have shape [a, b, ..., z].

tf.train.shuffle_batch(tensors, batch_size, capacity, min_after_dequeue, num_threads=1, seed=None, enqueue_many=False, shapes=None, allow_smaller_final_batch=False, shared_name=None, name=None)

Creates batches by randomly shuffling tensors.

This function adds the following to the current Graph:

• A shuffling queue into which tensors from tensors are enqueued.
• A dequeue_many operation to create batches from the queue.
• A QueueRunner to QUEUE_RUNNER collection, to enqueue the tensors from tensors.

If enqueue_many is False, tensors is assumed to represent a single example. An input tensor with shape [x, y, z] will be output as a tensor with shape [batch_size, x, y, z].

If enqueue_many is True, tensors is assumed to represent a batch of examples, where the first dimension is indexed by example, and all members of tensors should have the same size in the first dimension. If an input tensor has shape [*, x, y, z], the output will have shape [batch_size, x, y, z].

The capacity argument controls the how long the prefetching is allowed to grow the queues.

The returned operation is a dequeue operation and will throw tf.errors.OutOfRangeError if the input queue is exhausted. If this operation is feeding another input queue, its queue runner will catch this exception, however, if this operation is used in your main thread you are responsible for catching this yourself.

For example:

   
   

    
    

    
    

   
   
# Creates batches of 32 images and 32 labels.
image_batch, label_batch = tf.train.shuffle_batch(
      [single_image, single_label],
      batch_size=32,
      num_threads=4,
      capacity=50000,
      min_after_dequeue=10000)

N.B.: You must ensure that either (i) the shapes argument is passed, or (ii) all of the tensors in tensors must have fully-defined shapes. ValueError will be raised if neither of these conditions holds.