# Copyright 2019-2023 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""
This file contains standard format dataset loading classes.
You can convert a dataset to a standard format using the following steps:
1. Use mindspore.mindrecord.FileWriter / tf.io.TFRecordWriter api to
convert dataset to MindRecord / TFRecord.
2. Use MindDataset / TFRecordDataset to load MindRecord / TFRecrod files.
After declaring the dataset object, you can further apply dataset operations
(e.g. filter, skip, concat, map, batch) on it.
"""
import platform
import numpy as np
import mindspore._c_dataengine as cde
from mindspore import log as logger
from .datasets import UnionBaseDataset, SourceDataset, MappableDataset, Shuffle, Schema, \
shuffle_to_shuffle_mode, shuffle_to_bool
from .datasets_user_defined import GeneratorDataset
from .obs.obs_mindrecord_dataset import MindRecordFromOBS
from .validators import check_csvdataset, check_minddataset, check_tfrecorddataset, check_obsminddataset
from ...mindrecord.config import _get_enc_key, _get_dec_mode, _get_hash_mode, decrypt, verify_file_hash
from ..core.validator_helpers import replace_none
from . import samplers
[docs]class CSVDataset(SourceDataset, UnionBaseDataset):
"""
A source dataset that reads and parses comma-separated values
`(CSV) <https://en.wikipedia.org/wiki/Comma-separated_values>`_ files as dataset.
The columns of generated dataset depend on the source CSV files.
Args:
dataset_files (Union[str, list[str]]): String or list of files to be read or glob strings to search
for a pattern of files. The list will be sorted in a lexicographical order.
field_delim (str, optional): A string that indicates the char delimiter to separate fields.
Default: ``','``.
column_defaults (list, optional): List of default values for the CSV field. Default: ``None``. Each item
in the list is either a valid type (float, int, or string). If this is not provided, treats all
columns as string type.
column_names (list[str], optional): List of column names of the dataset. Default: ``None``. If this
is not provided, infers the column_names from the first row of CSV file.
num_samples (int, optional): The number of samples to be included in the dataset.
Default: ``None``, will include all images.
num_parallel_workers (int, optional): Number of worker threads to read the data.
Default: ``None``, will use global default workers(8), it can be set
by :func:`mindspore.dataset.config.set_num_parallel_workers` .
shuffle (Union[bool, Shuffle], optional): Perform reshuffling of the data every epoch.
Default: ``Shuffle.GLOBAL`` . Bool type and Shuffle enum are both supported to pass in.
If `shuffle` is ``False`` , no shuffling will be performed.
If `shuffle` is ``True`` , performs global shuffle.
There are three levels of shuffling, desired shuffle enum defined by :class:`mindspore.dataset.Shuffle` .
- ``Shuffle.GLOBAL`` : Shuffle both the files and samples, same as setting shuffle to True.
- ``Shuffle.FILES`` : Shuffle files only.
num_shards (int, optional): Number of shards that the dataset will be divided into. Default: ``None`` .
When this argument is specified, `num_samples` reflects the maximum sample number of per shard.
shard_id (int, optional): The shard ID within `num_shards` . Default: ``None``. This
argument can only be specified when `num_shards` is also specified.
cache (DatasetCache, optional): Use tensor caching service to speed up dataset processing. More details:
`Single-Node Data Cache <https://www.mindspore.cn/tutorials/experts/en/r2.3.0/dataset/cache.html>`_ .
Default: ``None``, which means no cache is used.
Raises:
RuntimeError: If `dataset_files` are not valid or do not exist.
ValueError: If `field_delim` is invalid.
ValueError: If `num_parallel_workers` exceeds the max thread numbers.
RuntimeError: If `num_shards` is specified but `shard_id` is None.
RuntimeError: If `shard_id` is specified but `num_shards` is None.
ValueError: If `shard_id` is not in range of [0, `num_shards` ).
Examples:
>>> import mindspore.dataset as ds
>>> csv_dataset_dir = ["/path/to/csv_dataset_file"] # contains 1 or multiple csv files
>>> dataset = ds.CSVDataset(dataset_files=csv_dataset_dir, column_names=['col1', 'col2', 'col3', 'col4'])
"""
@check_csvdataset
def __init__(self, dataset_files, field_delim=',', column_defaults=None, column_names=None, num_samples=None,
num_parallel_workers=None, shuffle=Shuffle.GLOBAL, num_shards=None, shard_id=None, cache=None):
super().__init__(num_parallel_workers=num_parallel_workers, num_samples=num_samples, shuffle=shuffle,
num_shards=num_shards, shard_id=shard_id, cache=cache)
self.dataset_files = self._find_files(dataset_files)
self.dataset_files.sort()
self.field_delim = replace_none(field_delim, ',')
self.column_defaults = replace_none(column_defaults, [])
self.column_names = replace_none(column_names, [])
def parse(self, children=None):
return cde.CSVNode(self.dataset_files, self.field_delim, self.column_defaults, self.column_names,
self.num_samples, self.shuffle_flag, self.num_shards, self.shard_id)
[docs]class MindDataset(MappableDataset, UnionBaseDataset):
"""
A source dataset that reads and parses MindRecord dataset.
The columns of generated dataset depend on the source MindRecord files.
Args:
dataset_files (Union[str, list[str]]): If dataset_file is a str, it represents for
a file name of one component of a mindrecord source, other files with identical source
in the same path will be found and loaded automatically. If dataset_file is a list,
it represents for a list of dataset files to be read directly.
columns_list (list[str], optional): List of columns to be read. Default: ``None`` , read all columns.
num_parallel_workers (int, optional): Number of worker threads to read the data.
Default: ``None`` , will use global default workers(8), it can be set
by :func:`mindspore.dataset.config.set_num_parallel_workers` .
shuffle (Union[bool, Shuffle], optional): Perform reshuffling of the data every epoch.
Default: ``None``, performs `mindspore.dataset.Shuffle.GLOBAL`.
Bool type and Shuffle enum are both supported to pass in.
If `shuffle` is ``False`` , no shuffling will be performed.
If `shuffle` is ``True`` , performs global shuffle.
There are three levels of shuffling, desired shuffle enum defined by :class:`mindspore.dataset.Shuffle` .
- ``Shuffle.GLOBAL`` : Global shuffle of all rows of data in dataset, same as setting shuffle to True.
- ``Shuffle.FILES`` : Shuffle the file sequence but keep the order of data within each file.
Not supported when the number of samples in the dataset is greater than 100 million.
- ``Shuffle.INFILE`` : Keep the file sequence the same but shuffle the data within each file.
Not supported when the number of samples in the dataset is greater than 100 million.
num_shards (int, optional): Number of shards that the dataset will be divided into. Default: ``None`` .
When this argument is specified, `num_samples` reflects the maximum sample number of per shard.
shard_id (int, optional): The shard ID within `num_shards` . Default: ``None`` . This
argument can only be specified when `num_shards` is also specified.
sampler (Sampler, optional): Object used to choose samples from the
dataset. Default: ``None`` , sampler is exclusive
with shuffle and block_reader. Support list: :class:`mindspore.dataset.SubsetRandomSampler`,
:class:`mindspore.dataset.PKSampler`, :class:`mindspore.dataset.RandomSampler`,
:class:`mindspore.dataset.SequentialSampler`, :class:`mindspore.dataset.DistributedSampler`.
padded_sample (dict, optional): Samples will be appended to dataset, where
keys are the same as columns_list. Default: ``None``.
num_padded (int, optional): Number of padding samples. Dataset size
plus num_padded should be divisible by num_shards. Default: ``None``.
num_samples (int, optional): The number of samples to be included in the dataset.
Default: ``None`` , all samples.
cache (DatasetCache, optional): Use tensor caching service to speed up dataset processing. More details:
`Single-Node Data Cache <https://www.mindspore.cn/tutorials/experts/en/r2.3.0/dataset/cache.html>`_ .
Default: ``None`` , which means no cache is used.
Raises:
ValueError: If dataset_files are not valid or do not exist.
ValueError: If `num_parallel_workers` exceeds the max thread numbers.
RuntimeError: If `num_shards` is specified but `shard_id` is None.
RuntimeError: If `shard_id` is specified but `num_shards` is None.
ValueError: If `shard_id` is not in range of [0, `num_shards` ).
Note:
- The parameters `num_samples` , `shuffle` , `num_shards` , `shard_id` can be used to control the sampler
used in the dataset, and their effects when combined with parameter `sampler` are as follows.
.. include:: mindspore.dataset.sampler.txt
Examples:
>>> import mindspore.dataset as ds
>>> mindrecord_files = ["/path/to/mind_dataset_file"] # contains 1 or multiple MindRecord files
>>> dataset = ds.MindDataset(dataset_files=mindrecord_files)
"""
def parse(self, children=None):
return cde.MindDataNode(self.dataset_files, self.columns_list, self.sampler, self.new_padded_sample,
self.num_padded, shuffle_to_shuffle_mode(self.shuffle_option))
@check_minddataset
def __init__(self, dataset_files, columns_list=None, num_parallel_workers=None, shuffle=None, num_shards=None,
shard_id=None, sampler=None, padded_sample=None, num_padded=None, num_samples=None, cache=None):
super().__init__(num_parallel_workers=num_parallel_workers, sampler=sampler, num_samples=num_samples,
shuffle=shuffle_to_bool(shuffle), num_shards=num_shards, shard_id=shard_id, cache=cache)
if num_samples and shuffle in (Shuffle.FILES, Shuffle.INFILE):
raise ValueError("'Shuffle.FILES' or 'Shuffle.INFILE' and 'num_samples' "
"cannot be specified at the same time.")
self.shuffle_option = shuffle
self.load_dataset = True
if isinstance(dataset_files, list):
self.load_dataset = False
self.dataset_files = dataset_files
if platform.system().lower() == "windows":
if isinstance(dataset_files, list):
file_tuple = []
for item in dataset_files:
item.replace("\\", "/")
file_tuple.append(item)
self.dataset_files = file_tuple
else:
self.dataset_files = dataset_files.replace("\\", "/")
# do decrypt & integrity check
if not isinstance(self.dataset_files, list):
if _get_enc_key() is not None or _get_hash_mode() is not None:
logger.warning("When a single mindrecord file which is generated by " +
"`mindspore.mindrecord.FileWriter` with `shard_num` > 1 is used as the input, " +
"enabling decryption/integrity check may fail. Please use file list as the input.")
# decrypt the data file and index file
index_file_name = self.dataset_files + ".db"
self.dataset_files = decrypt(self.dataset_files, _get_enc_key(), _get_dec_mode())
decrypt(index_file_name, _get_enc_key(), _get_dec_mode())
# verify integrity check
verify_file_hash(self.dataset_files)
verify_file_hash(self.dataset_files + ".db")
else:
file_tuple = []
for item in self.dataset_files:
# decrypt the data file and index file
index_file_name = item + ".db"
decrypt_filename = decrypt(item, _get_enc_key(), _get_dec_mode())
file_tuple.append(decrypt_filename)
decrypt(index_file_name, _get_enc_key(), _get_dec_mode())
# verify integrity check
verify_file_hash(decrypt_filename)
verify_file_hash(decrypt_filename + ".db")
self.dataset_files = file_tuple
self.columns_list = replace_none(columns_list, [])
if sampler is not None:
if isinstance(sampler, (
samplers.SubsetRandomSampler, samplers.SubsetSampler, samplers.PKSampler,
samplers.DistributedSampler,
samplers.RandomSampler, samplers.SequentialSampler)) is False:
raise ValueError("The sampler is not supported yet.")
self.padded_sample = padded_sample
self.num_padded = replace_none(num_padded, 0)
self.new_padded_sample = {}
if padded_sample:
for k, v in padded_sample.items():
if isinstance(v, np.ndarray):
self.new_padded_sample[k] = v.tobytes()
else:
self.new_padded_sample[k] = v
[docs]class TFRecordDataset(SourceDataset, UnionBaseDataset):
"""
A source dataset that reads and parses datasets stored on disk in TFData format.
The columns of generated dataset depend on the source TFRecord files.
Note:
'TFRecordDataset' is not support on Windows platform yet.
Args:
dataset_files (Union[str, list[str]]): String or list of files to be read or glob strings to search for a
pattern of files. The list will be sorted in lexicographical order.
schema (Union[str, Schema], optional): Data format policy, which specifies the data types and shapes of the data
column to be read. Both JSON file path and objects constructed by :class:`mindspore.dataset.Schema` are
acceptable. Default: ``None`` .
columns_list (list[str], optional): List of columns to be read. Default: ``None`` , read all columns.
num_samples (int, optional): The number of samples (rows) to be included in the dataset. Default: ``None`` .
When `num_shards` and `shard_id` are specified, it will be interpreted as number of rows per shard.
Processing priority for `num_samples` is as the following:
- If specify `num_samples` with value > 0, read `num_samples` samples.
- If no `num_samples` and specify numRows(parsed from `schema`) with value > 0, read numRows samples.
- If no `num_samples` and no `schema`, read the full dataset.
num_parallel_workers (int, optional): Number of worker threads to read the data.
Default: ``None`` , will use global default workers(8), it can be set
by :func:`mindspore.dataset.config.set_num_parallel_workers` .
shuffle (Union[bool, Shuffle], optional): Perform reshuffling of the data every epoch.
Default: ``Shuffle.GLOBAL`` . Bool type and Shuffle enum are both supported to pass in.
If `shuffle` is ``False``, no shuffling will be performed.
If `shuffle` is ``True``, perform global shuffle.
There are three levels of shuffling, desired shuffle enum defined by :class:`mindspore.dataset.Shuffle` .
- ``Shuffle.GLOBAL`` : Shuffle both the files and samples, same as setting `shuffle` to ``True``.
- ``Shuffle.FILES`` : Shuffle files only.
num_shards (int, optional): Number of shards that the dataset will be divided
into. Default: ``None`` . When this argument is specified, `num_samples` reflects
the maximum sample number per shard.
shard_id (int, optional): The shard ID within `num_shards` . Default: ``None`` . This
argument can only be specified when `num_shards` is also specified.
shard_equal_rows (bool, optional): Get equal rows for all shards. Default: ``False``. If `shard_equal_rows`
is False, the number of rows of each shard may not be equal, and may lead to a failure in distributed
training. When the number of samples per TFRecord file are not equal, it is suggested to set it to ``True``.
This argument should only be specified when `num_shards` is also specified.
When `compression_type` is not ``None``, and `num_samples` or numRows (parsed from `schema` ) is provided,
`shard_equal_rows` will be implied as ``True``.
cache (DatasetCache, optional): Use tensor caching service to speed up dataset processing. More details:
`Single-Node Data Cache <https://www.mindspore.cn/tutorials/experts/en/r2.3.0/dataset/cache.html>`_ .
Default: ``None`` , which means no cache is used.
compression_type (str, optional): The type of compression used for all files, must be either ``''``,
``'GZIP'``, or ``'ZLIB'``. Default: ``None`` , as in empty string. It is highly recommended to
provide `num_samples` or numRows (parsed from `schema`) when `compression_type` is ``"GZIP"`` or
``"ZLIB"`` to avoid performance degradation caused by multiple decompressions of the same file
to obtain the file size.
Raises:
ValueError: If dataset_files are not valid or do not exist.
ValueError: If `num_parallel_workers` exceeds the max thread numbers.
RuntimeError: If `num_shards` is specified but `shard_id` is None.
RuntimeError: If `shard_id` is specified but `num_shards` is None.
ValueError: If `shard_id` is not in range of [0, `num_shards` ).
ValueError: If `compression_type` is not ``''``, ``'GZIP'`` or ``'ZLIB'`` .
ValueError: If `compression_type` is provided, but the number of dataset files < `num_shards` .
ValueError: If `num_samples` < 0.
Examples:
>>> import mindspore.dataset as ds
>>> from mindspore import dtype as mstype
>>>
>>> tfrecord_dataset_dir = ["/path/to/tfrecord_dataset_file"] # contains 1 or multiple TFRecord files
>>> tfrecord_schema_file = "/path/to/tfrecord_schema_file"
>>>
>>> # 1) Get all rows from tfrecord_dataset_dir with no explicit schema.
>>> # The meta-data in the first row will be used as a schema.
>>> dataset = ds.TFRecordDataset(dataset_files=tfrecord_dataset_dir)
>>>
>>> # 2) Get all rows from tfrecord_dataset_dir with user-defined schema.
>>> schema = ds.Schema()
>>> schema.add_column(name='col_1d', de_type=mstype.int64, shape=[2])
>>> dataset = ds.TFRecordDataset(dataset_files=tfrecord_dataset_dir, schema=schema)
>>>
>>> # 3) Get all rows from tfrecord_dataset_dir with the schema file.
>>> dataset = ds.TFRecordDataset(dataset_files=tfrecord_dataset_dir, schema=tfrecord_schema_file)
"""
@check_tfrecorddataset
def __init__(self, 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, compression_type=None):
super().__init__(num_parallel_workers=num_parallel_workers, num_samples=num_samples, shuffle=shuffle,
num_shards=num_shards, shard_id=shard_id, cache=cache)
if platform.system().lower() == "windows":
raise NotImplementedError("TFRecordDataset is not supported for windows.")
self.dataset_files = self._find_files(dataset_files)
self.dataset_files.sort()
self.schema = schema
self.columns_list = replace_none(columns_list, [])
self.shard_equal_rows = replace_none(shard_equal_rows, False)
self.compression_type = replace_none(compression_type, "")
# Only take numRows from schema when num_samples is not provided
if self.schema is not None and (self.num_samples is None or self.num_samples == 0):
self.num_samples = Schema.get_num_rows(self.schema)
if self.compression_type in ['ZLIB', 'GZIP'] and (self.num_samples is None or self.num_samples == 0):
logger.warning("Since compression_type is set, but neither num_samples nor numRows (from schema file) " +
"is provided, performance might be degraded.")
def parse(self, children=None):
schema = self.schema.cpp_schema if isinstance(self.schema, Schema) else self.schema
return cde.TFRecordNode(self.dataset_files, schema, self.columns_list, self.num_samples, self.shuffle_flag,
self.num_shards, self.shard_id, self.shard_equal_rows, self.compression_type)
[docs]class OBSMindDataset(GeneratorDataset):
"""
A source dataset that reads and parses MindRecord dataset which stored in cloud storage
such as OBS, Minio or AWS S3.
The columns of generated dataset depend on the source MindRecord files.
Args:
dataset_files (list[str]): List of files in cloud storage to be read and file path is in
the format of s3://bucketName/objectKey.
server (str): Endpoint for accessing cloud storage.
If it's OBS Service of Huawei Cloud, the endpoint is
like ``<obs.cn-north-4.myhuaweicloud.com>`` (Region cn-north-4).
If it's Minio which starts locally, the endpoint is like ``<https://127.0.0.1:9000>``.
ak (str): The access key ID used to access the OBS data.
sk (str): The secret access key used to access the OBS data.
sync_obs_path (str): Remote dir path used for synchronization, users need to
create it on cloud storage in advance. Path is in the format of s3://bucketName/objectKey.
columns_list (list[str], optional): List of columns to be read. Default: ``None`` , read all columns.
shuffle (Union[bool, Shuffle], optional): Perform reshuffling of the data every epoch.
Default: ``Shuffle.GLOBAL``. Bool type and Shuffle enum are both supported to pass in.
If `shuffle` is ``False`` , no shuffling will be performed.
If `shuffle` is ``True`` , performs global shuffle.
There are three levels of shuffling, desired shuffle enum defined by :class:`mindspore.dataset.Shuffle` .
- ``Shuffle.GLOBAL`` : Global shuffle of all rows of data in dataset, same as setting shuffle to True.
- ``Shuffle.FILES`` : Shuffle the file sequence but keep the order of data within each file.
- ``Shuffle.INFILE`` : Keep the file sequence the same but shuffle the data within each file.
num_shards (int, optional): Number of shards that the dataset will be divided
into. Default: ``None`` .
shard_id (int, optional): The shard ID within num_shards. Default: ``None`` . This
argument can only be specified when `num_shards` is also specified.
shard_equal_rows (bool, optional): Get equal rows for all shards. Default: ``True``. If shard_equal_rows
is false, number of rows of each shard may be not equal, and may lead to a failure in distributed training.
When the number of samples of per MindRecord file are not equal, it is suggested to set to ``True``.
This argument should only be specified when `num_shards` is also specified.
Raises:
RuntimeError: If `sync_obs_path` do not exist.
ValueError: If `columns_list` is invalid.
RuntimeError: If `num_shards` is specified but `shard_id` is None.
RuntimeError: If `shard_id` is specified but `num_shards` is None.
ValueError: If `shard_id` is not in range of [0, `num_shards` ).
Note:
- It's necessary to create a synchronization directory on cloud storage in
advance which be defined by parameter: `sync_obs_path` .
- If training is offline(no cloud), it's recommended to set the
environment variable `BATCH_JOB_ID` .
- In distributed training, if there are multiple nodes(servers), all 8
devices must be used in each node(server). If there is only one
node(server), there is no such restriction.
Examples:
>>> import mindspore.dataset as ds
>>> # OBS
>>> bucket = "iris" # your obs bucket name
>>> # the bucket directory structure is similar to the following:
>>> # - imagenet21k
>>> # | - mr_imagenet21k_01
>>> # | - mr_imagenet21k_02
>>> # - sync_node
>>> dataset_obs_dir = ["s3://" + bucket + "/imagenet21k/mr_imagenet21k_01",
... "s3://" + bucket + "/imagenet21k/mr_imagenet21k_02"]
>>> sync_obs_dir = "s3://" + bucket + "/sync_node"
>>> num_shards = 8
>>> shard_id = 0
>>> dataset = ds.OBSMindDataset(dataset_obs_dir, "obs.cn-north-4.myhuaweicloud.com",
... "AK of OBS", "SK of OBS",
... sync_obs_dir, shuffle=True, num_shards=num_shards, shard_id=shard_id)
"""
@check_obsminddataset
def __init__(self, dataset_files, server, ak, sk, sync_obs_path,
columns_list=None,
shuffle=Shuffle.GLOBAL,
num_shards=None,
shard_id=None,
shard_equal_rows=True):
from .obs.config_loader import config
config.AK = ak
config.SK = sk
config.SERVER = server
config.SYNC_OBS_PATH = sync_obs_path
if shuffle is not None and not isinstance(shuffle, (bool, Shuffle)):
raise TypeError("shuffle must be of boolean or enum of 'Shuffle' values like 'Shuffle.GLOBAL' or "
"'Shuffle.FILES'.")
self.num_shards = replace_none(num_shards, 1)
self.shard_id = replace_none(shard_id, 0)
self.shuffle = replace_none(shuffle, True)
dataset = MindRecordFromOBS(dataset_files, columns_list, shuffle, self.num_shards, self.shard_id,
shard_equal_rows, config.DATASET_LOCAL_PATH)
if not columns_list:
columns_list = dataset.get_col_names()
else:
full_columns_list = dataset.get_col_names()
if not set(columns_list).issubset(full_columns_list):
raise ValueError("columns_list: {} can not found in MindRecord fields: {}".format(columns_list,
full_columns_list))
super().__init__(source=dataset, column_names=columns_list, num_shards=None, shard_id=None, shuffle=False)
def add_sampler(self, new_sampler):
raise NotImplementedError("add_sampler is not supported for OBSMindDataset.")
def use_sampler(self, new_sampler):
raise NotImplementedError("use_sampler is not supported for OBSMindDataset.")