来自Keras ImageDataGenerator的数据批量的Feed Slim训练循环

Question

我有来自Keras的ImageDataGenerator对象，它无限期地产生批次。我怎样才能将n批次的每个尺寸送到张量流量变细的训练循环中。

from keras.preprocessing.image import ImageDataGenerator
import tensorflow.contrib.slim as slim

datagen = ImageDataGenerator(rotation_range=20,
  width_shift_range=0.2,
  height_shift_range=0.2,
  horizontal_flip=True)
for X_batch, Y_batch in datagen.flow(X_train, Y_train, batch_size=s):
    # here I want to feed X_batch and Y_batch to slim.learning.train to run train_op in a loop

    batches += 1
    if batches >= n:
        # we need to break the loop by hand because
        # the generator loops indefinitely
        break

请注意，我想使用由tensor-slim提供的训练循环，而不是使用原生张量流来编写我自己的例程进行训练。