singletongue
commited on
Commit
•
9c6cfa2
1
Parent(s):
2ef7386
Change to use ArrowBasedBuilder
Browse files- README.md +1 -1
- wikipedia-utils.py +33 -14
README.md
CHANGED
@@ -127,4 +127,4 @@ size_categories:
|
|
127 |
|
128 |
Preprocessed Wikipedia texts generated with the scripts in [singletongue/wikipedia-utils](https://github.com/singletongue/wikipedia-utils) repo.
|
129 |
|
130 |
-
For detailed information on how the texts are processed, please refer to the repo.
|
|
|
127 |
|
128 |
Preprocessed Wikipedia texts generated with the scripts in [singletongue/wikipedia-utils](https://github.com/singletongue/wikipedia-utils) repo.
|
129 |
|
130 |
+
For detailed information on how the texts are processed, please refer to the repo.
|
wikipedia-utils.py
CHANGED
@@ -1,4 +1,5 @@
|
|
1 |
# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
|
|
|
2 |
#
|
3 |
# Licensed under the Apache License, Version 2.0 (the "License");
|
4 |
# you may not use this file except in compliance with the License.
|
@@ -13,11 +14,11 @@
|
|
13 |
# limitations under the License.
|
14 |
"""Wikipedia-Utils: Preprocessed Wikipedia Texts for NLP"""
|
15 |
|
16 |
-
|
17 |
-
import
|
18 |
-
from typing import Dict, Iterator, List, Tuple, Union
|
19 |
|
20 |
import datasets
|
|
|
21 |
|
22 |
|
23 |
_DESCRIPTION = "Preprocessed Wikipedia texts generated with scripts in singletongue/wikipedia-utils repo."
|
@@ -40,7 +41,7 @@ _URLS = {
|
|
40 |
_VERSION = datasets.Version("1.0.0")
|
41 |
|
42 |
|
43 |
-
class WikipediaUtils(datasets.
|
44 |
"""Wikipedia-Utils dataset."""
|
45 |
|
46 |
BUILDER_CONFIGS = [datasets.BuilderConfig(name=name, version=_VERSION) for name in _URLS.keys()]
|
@@ -87,18 +88,36 @@ class WikipediaUtils(datasets.GeneratorBasedBuilder):
|
|
87 |
filepath = dl_manager.download_and_extract(url)
|
88 |
return [datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": filepath})]
|
89 |
|
90 |
-
def
|
91 |
if self.config.name.startswith("corpus"):
|
92 |
with open(filepath) as f:
|
93 |
-
|
94 |
-
|
95 |
-
|
96 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
97 |
elif self.config.name.startswith(("paragraphs", "passages")):
|
98 |
-
with open(filepath) as f:
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
103 |
else:
|
104 |
raise ValueError("Invalid dataset config name is specified.")
|
|
|
1 |
# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
|
2 |
+
# Copyright 2023 Masatoshi Suzuki (@singletongue)
|
3 |
#
|
4 |
# Licensed under the Apache License, Version 2.0 (the "License");
|
5 |
# you may not use this file except in compliance with the License.
|
|
|
14 |
# limitations under the License.
|
15 |
"""Wikipedia-Utils: Preprocessed Wikipedia Texts for NLP"""
|
16 |
|
17 |
+
import io
|
18 |
+
from typing import Iterator, List, Tuple
|
|
|
19 |
|
20 |
import datasets
|
21 |
+
import pyarrow as pa
|
22 |
|
23 |
|
24 |
_DESCRIPTION = "Preprocessed Wikipedia texts generated with scripts in singletongue/wikipedia-utils repo."
|
|
|
41 |
_VERSION = datasets.Version("1.0.0")
|
42 |
|
43 |
|
44 |
+
class WikipediaUtils(datasets.ArrowBasedBuilder):
|
45 |
"""Wikipedia-Utils dataset."""
|
46 |
|
47 |
BUILDER_CONFIGS = [datasets.BuilderConfig(name=name, version=_VERSION) for name in _URLS.keys()]
|
|
|
88 |
filepath = dl_manager.download_and_extract(url)
|
89 |
return [datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": filepath})]
|
90 |
|
91 |
+
def _generate_tables(self, filepath: str, chunksize: int = 10 << 20) -> Iterator[Tuple[int, pa.Table]]:
|
92 |
if self.config.name.startswith("corpus"):
|
93 |
with open(filepath) as f:
|
94 |
+
batch_idx = 0
|
95 |
+
while True:
|
96 |
+
batch = f.read(chunksize)
|
97 |
+
if not batch:
|
98 |
+
break
|
99 |
+
|
100 |
+
batch += f.readline()
|
101 |
+
batch = [line.rstrip("\n") for line in io.StringIO(batch).readlines()]
|
102 |
+
pa_table = pa.Table.from_arrays([pa.array(batch)], names=["text"])
|
103 |
+
|
104 |
+
yield batch_idx, pa_table
|
105 |
+
batch_idx += 1
|
106 |
elif self.config.name.startswith(("paragraphs", "passages")):
|
107 |
+
with open(filepath, "rb") as f:
|
108 |
+
batch_idx = 0
|
109 |
+
block_size = max(chunksize // 32, 16 << 10)
|
110 |
+
while True:
|
111 |
+
batch = f.read(chunksize)
|
112 |
+
if not batch:
|
113 |
+
break
|
114 |
+
|
115 |
+
batch += f.readline()
|
116 |
+
pa_table = pa.json.read_json(
|
117 |
+
io.BytesIO(batch), read_options=pa.json.ReadOptions(block_size=block_size)
|
118 |
+
)
|
119 |
+
|
120 |
+
yield batch_idx, pa_table
|
121 |
+
batch_idx += 1
|
122 |
else:
|
123 |
raise ValueError("Invalid dataset config name is specified.")
|