Function Differences with tf.data.TFRecordDataset

View Source On Gitee

tf.data.TFRecordDataset

class tf.data.TFRecordDataset(
    filenames,
    compression_type=None,
    buffer_size=None,
    num_parallel_reads=None
)

For more information, see tf.data.TFRecordDataset.

mindspore.dataset.TFRecordDataset

class mindspore.dataset.TFRecordDataset(
    dataset_files,
    schema=None,
    columns_list=None,
    num_samples=None,
    num_parallel_workers=None,
    shuffle=Shuffle.GLOBAL,
    num_shards=None,
    shard_id=None,
    shard_equal_rows=False,
    cache=None
)

For more information, see mindspore.dataset.TFRecordDataset.

Differences

TensorFlow: Create Dataset from a list of TFRecord files. It supports decompression operations and can set the cache size.

MindSpore: Create Dataset from a list of TFRecord files. It supports setting the number of samples and the schema of data.

Code Example

# The following implements TFRecordDataset with MindSpore.
import mindspore.dataset as ds

dataset_files = ['/tmp/example0.tfrecord',
                 '/tmp/example1.tfrecord']
dataset = ds.TFRecordDataset(dataset_files)

# The following implements TFRecordDataset with TensorFlow.
import tensorflow as tf

filenames = ['/tmp/example0.tfrecord',
             '/tmp/example1.tfrecord']
dataset = tf.data.TFRecordDataset(filenames)