f2073333b710a340403843763ba60eb1e6699916,examples/data_process/tutorial_tfrecord.py,,read_and_decode,#Any#,74

Before Change

    // generate a queue with a given file name
    filename_queue = tf.train.string_input_producer([filename])
    reader = tf.TFRecordReader()
    _, serialized_example = reader.read(filename_queue)  // return the file and the name of file
    features = tf.parse_single_example(
        serialized_example,  // see parse_single_sequence_example for sequence example
        features={

After Change

// use shuffle and batch
def read_and_decode(filename):
    // generate a queue with a given file name
    raw_dataset = tf.data.TFRecordDataset([filename]).shuffle().batch(4)
    for serialized_example in raw_dataset:
        features = tf.io.parse_example(
            serialized_example,
            features={
                "label": tf.io.FixedLenFeature([], tf.int64),
                "img_raw": tf.io.FixedLenFeature([], tf.string),
            }
        )
        // You can do more image distortion here for training data
        img_batch = tf.io.decode_raw(features["img_raw"], tf.uint8)
        img_batch = tf.reshape(img_batch, [4, 224, 224, 3])
        // img = tf.cast(img, tf.float32) * (1. / 255) - 0.5
        label_batch = tf.cast(features["label"], tf.int32)
        yield img_batch, label_batch


img_batch, label_batch = next(read_and_decode("train.tfrecords"))
print("img_batch   : %s" % img_batch.shape)
print("label_batch : %s" % label_batch.shape)
tl.visualize.images2d(img_batch, second=1, saveable=False, name="batch", dtype=None, fig_idx=2020121)

Instances