|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
import csv |
|
import json |
|
import os |
|
|
|
import datasets |
|
|
|
|
|
_CITATION = """\ |
|
""" |
|
|
|
_DESCRIPTION = """\ |
|
Online dataset mockup. |
|
""" |
|
|
|
_HOMEPAGE = "" |
|
|
|
_LICENSE = "" |
|
|
|
_URLS = {} |
|
|
|
class MockupDataset(datasets.GeneratorBasedBuilder): |
|
"""TODO: Short description of my dataset.""" |
|
|
|
VERSION = datasets.Version("0.0.0") |
|
BUILDER_CONFIGS = [] |
|
|
|
def __init__(self, name=None, data_config={}, **kwargs): |
|
super().__init__(**kwargs) |
|
|
|
if 'length' not in data_config: |
|
data_config['length'] = 20 |
|
if 'size' not in data_config: |
|
data_config['size'] = 100 |
|
|
|
self.data_config = data_config |
|
|
|
def _info(self): |
|
features = datasets.Features( |
|
{ |
|
"x": datasets.Value("null"), |
|
"y": datasets.Value("null") |
|
} |
|
) |
|
|
|
return datasets.DatasetInfo( |
|
description=_DESCRIPTION, |
|
features=features, |
|
homepage=_HOMEPAGE, |
|
license=_LICENSE, |
|
citation=_CITATION, |
|
) |
|
|
|
def _split_generators(self, dl_manager): |
|
return [ |
|
datasets.SplitGenerator( |
|
name=datasets.Split.TRAIN, |
|
gen_kwargs={ |
|
"split": "train", |
|
}, |
|
) |
|
] |
|
|
|
|
|
def _generate_examples(self, split): |
|
for i in range(self.data_config['size']): |
|
print('generator iter', i) |
|
yield i, { |
|
"x": [0]*self.data_config['length'], |
|
"y": [1]*self.data_config['length'] |
|
} |