Shuffle (1) 썸네일형 리스트형 mnist your own dataset - tensorflow.data.Dataset 2 Update 2019.8.9: Posting def data_slice_and_batch(DATA_PATH_LIST, labels, bufferSize, batch_size): dataset = tf.data.Dataset.from_tensor_slices((path_list, labels)) dataset = dataset.map(lambda path_list, labels: tuple(tf.py_func(_read_py_function, [path_list, labels], [tf.int32, tf.uint8]))) dataset = dataset.shuffle(buffer_size=(bufferSize)) dataset = dataset.repeat() dataset = dataset.batch(b.. 이전 1 다음