Datasets:
Size:
10B<n<100B
License:
Finalized the script.
Browse files- bertic_data.py +28 -48
bertic_data.py
CHANGED
@@ -1,26 +1,26 @@
|
|
1 |
import datasets
|
2 |
import gzip
|
3 |
-
import
|
4 |
from typing import List
|
5 |
|
6 |
|
7 |
_URL = "http://nl.ijs.si/nikola/dedup_hbs/"
|
8 |
-
_URLS =
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
"
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
_HOMEPAGE = _URL
|
25 |
|
26 |
_DESCRIPTION = """\
|
@@ -54,18 +54,6 @@ _CITATION = r"""
|
|
54 |
}"""
|
55 |
|
56 |
|
57 |
-
class BerticDataConfig(datasets.BuilderConfig):
|
58 |
-
"""BuilderConfig for Bertic data sample."""
|
59 |
-
|
60 |
-
def __init__(self, *args, subsets, **kwargs):
|
61 |
-
"""BuilderConfig for BerticData.
|
62 |
-
Args:
|
63 |
-
**kwargs: keyword arguments forwarded to super.
|
64 |
-
"""
|
65 |
-
super(BerticDataConfig, self).__init__(**kwargs)
|
66 |
-
self.subsets = subsets
|
67 |
-
|
68 |
-
|
69 |
class BerticData(datasets.GeneratorBasedBuilder):
|
70 |
"""Bertic dataset, used for training Bertic model."""
|
71 |
|
@@ -82,14 +70,6 @@ class BerticData(datasets.GeneratorBasedBuilder):
|
|
82 |
# You will be able to load one or the other configurations in the following list with
|
83 |
# data = datasets.load_dataset('my_dataset', 'first_domain')
|
84 |
# data = datasets.load_dataset('my_dataset', 'second_domain')
|
85 |
-
BUILDER_CONFIGS = [
|
86 |
-
BerticDataConfig(
|
87 |
-
name="bertic",
|
88 |
-
subsets=["bertic"],
|
89 |
-
version=VERSION,
|
90 |
-
description="All subsets",
|
91 |
-
)
|
92 |
-
]
|
93 |
|
94 |
def _info(self):
|
95 |
features = datasets.Features(
|
@@ -104,23 +84,23 @@ class BerticData(datasets.GeneratorBasedBuilder):
|
|
104 |
citation=_CITATION,
|
105 |
)
|
106 |
|
107 |
-
def _split_generators(
|
108 |
-
|
109 |
-
) -> List[datasets.SplitGenerator]:
|
110 |
-
urls_to_download = {i: url for i, url in enumerate(_URLS)}
|
111 |
-
downloaded_files = dl_manager.download_and_extract(urls_to_download)
|
112 |
|
113 |
return [
|
114 |
datasets.SplitGenerator(
|
115 |
-
name=
|
|
|
116 |
)
|
117 |
-
for
|
118 |
]
|
119 |
|
120 |
-
|
|
|
121 |
key = 0
|
122 |
-
for name in
|
123 |
-
with gzip.open(name, "rb") as f:
|
|
|
124 |
for line in f.readlines():
|
125 |
-
yield key, {"text": line.
|
126 |
key += 1
|
|
|
1 |
import datasets
|
2 |
import gzip
|
3 |
+
import os
|
4 |
from typing import List
|
5 |
|
6 |
|
7 |
_URL = "http://nl.ijs.si/nikola/dedup_hbs/"
|
8 |
+
_URLS = {
|
9 |
+
"macocu_hbs": _URL + "macocu.hbs.translit.dedup.lines.gz",
|
10 |
+
"hr_news": _URL + "hr_news.translit.dedup.lines.gz",
|
11 |
+
"bswac": _URL + "bswac.translit.dedup.lines.gz",
|
12 |
+
"cc100-hr": _URL + "cc100-hr.translit.dedup.lines.gz",
|
13 |
+
"cc100-sr": _URL + "cc100-sr.translit.dedup.lines.gz",
|
14 |
+
"classla-sr": _URL + "classla-sr.translit.dedup.lines.gz",
|
15 |
+
"classla-hr": _URL + "classla-hr.translit.dedup.lines.gz",
|
16 |
+
"classla-bs": _URL + "classla-bs.translit.dedup.lines.gz",
|
17 |
+
"cnrwac": _URL + "cnrwac.translit.dedup.lines.gz",
|
18 |
+
"hrwac": _URL + "hrwac.translit.dedup.lines.gz",
|
19 |
+
"mC4": _URL + "mC4.sr.translit.dedup.lines.gz",
|
20 |
+
"riznica": _URL + "riznica.translit.dedup.lines.gz",
|
21 |
+
"srwac": _URL + "srwac.translit.dedup.lines.gz",
|
22 |
+
}
|
23 |
+
|
24 |
_HOMEPAGE = _URL
|
25 |
|
26 |
_DESCRIPTION = """\
|
|
|
54 |
}"""
|
55 |
|
56 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
57 |
class BerticData(datasets.GeneratorBasedBuilder):
|
58 |
"""Bertic dataset, used for training Bertic model."""
|
59 |
|
|
|
70 |
# You will be able to load one or the other configurations in the following list with
|
71 |
# data = datasets.load_dataset('my_dataset', 'first_domain')
|
72 |
# data = datasets.load_dataset('my_dataset', 'second_domain')
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
73 |
|
74 |
def _info(self):
|
75 |
features = datasets.Features(
|
|
|
84 |
citation=_CITATION,
|
85 |
)
|
86 |
|
87 |
+
def _split_generators(self, dl_manager):
|
88 |
+
downloaded_files = dl_manager.download_and_extract(_URLS)
|
|
|
|
|
|
|
89 |
|
90 |
return [
|
91 |
datasets.SplitGenerator(
|
92 |
+
name=url,
|
93 |
+
gen_kwargs={"filepath": downloaded_files[url]},
|
94 |
)
|
95 |
+
for url in _URLS
|
96 |
]
|
97 |
|
98 |
+
# method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
|
99 |
+
def _generate_examples(self, filepath):
|
100 |
key = 0
|
101 |
+
for name in [filepath]:
|
102 |
+
# with gzip.open(name, "rb") as f:
|
103 |
+
with open(name, "r") as f:
|
104 |
for line in f.readlines():
|
105 |
+
yield key, {"text": line.rstrip()}
|
106 |
key += 1
|