kiddothe2b commited on
Commit
baa1d7c
1 Parent(s): d3b2cd9

Upload us-contracts xz file

Browse files
the_legal_pile_preprocessed.py ADDED
@@ -0,0 +1,158 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+ # Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
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.
6
+ # You may obtain a copy of the License at
7
+ #
8
+ # http://www.apache.org/licenses/LICENSE-2.0
9
+ #
10
+ # Unless required by applicable law or agreed to in writing, software
11
+ # distributed under the License is distributed on an "AS IS" BASIS,
12
+ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13
+ # See the License for the specific language governing permissions and
14
+ # limitations under the License.
15
+
16
+ """The Legal Pile (Preprocessed)"""
17
+
18
+
19
+ import datasets
20
+ import json
21
+
22
+ try:
23
+ import lzma as xz
24
+ except ImportError:
25
+ import pylzma as xz
26
+
27
+ logger = datasets.logging.get_logger(__name__)
28
+
29
+ _DESCRIPTION = """The Legal Pile (Preprocessed)"""
30
+
31
+ _HOMEPAGE = "https://iliaschalkidis.github.io/"
32
+
33
+ _VERSION = "1.0.0"
34
+
35
+
36
+ class LegalPileConfig(datasets.BuilderConfig):
37
+ """BuilderConfig for Legal Pile Corpora"""
38
+ def __init__(self, name, n_files, **kwargs):
39
+ super(LegalPileConfig, self).__init__(**kwargs)
40
+ self.name = name
41
+ self.filepath = f'https://huggingface.co/datasets/lexlms/the_legal_pile_preprocessed/' \
42
+ f'resolve/main/{self.name}.jsonl.xz'
43
+ self.n_files = n_files
44
+
45
+
46
+ class LegalPile(datasets.GeneratorBasedBuilder):
47
+ """
48
+ The Legal Pile
49
+ """
50
+
51
+ VERSION = datasets.Version(_VERSION)
52
+ BUILDER_CONFIG_CLASS = LegalPileConfig
53
+ BUILDER_CONFIGS = [
54
+ LegalPileConfig(
55
+ name=f"eu-legislation",
56
+ n_files=1,
57
+ description="EU Legislation",
58
+ ),
59
+ LegalPileConfig(
60
+ name=f"eu-court-cases",
61
+ n_files=1,
62
+ description="EU Court cases",
63
+ ),
64
+ LegalPileConfig(
65
+ name=f"ecthr-cases",
66
+ n_files=1,
67
+ description="ECtHR cases",
68
+ ),
69
+ LegalPileConfig(
70
+ name=f"uk-legislation",
71
+ n_files=1,
72
+ description="UK Legislation",
73
+ ),
74
+ LegalPileConfig(
75
+ name=f"uk-court-cases",
76
+ n_files=1,
77
+ description="UK Court cases",
78
+ ),
79
+ LegalPileConfig(
80
+ name=f"indian-court-cases",
81
+ n_files=1,
82
+ description="Indian Court cases",
83
+ ),
84
+ LegalPileConfig(
85
+ name=f"canadian-legislation",
86
+ n_files=1,
87
+ description="Canadian Legislation",
88
+ ),
89
+ LegalPileConfig(
90
+ name=f"canadian-court-cases",
91
+ n_files=1,
92
+ description="Canadian Legislation",
93
+ ),
94
+ LegalPileConfig(
95
+ name=f"us-legislation",
96
+ n_files=1,
97
+ description="US Legislation",
98
+ ),
99
+ LegalPileConfig(
100
+ name=f"us-contracts",
101
+ n_files=1,
102
+ description="US Contracts",
103
+ ),
104
+ LegalPileConfig(
105
+ name=f"us-court-cases",
106
+ n_files=1,
107
+ description="US Court cases",
108
+ )
109
+ ]
110
+
111
+ def _info(self):
112
+ features = datasets.Features(
113
+ {
114
+ "text": datasets.Value("string")
115
+ }
116
+ )
117
+ return datasets.DatasetInfo(
118
+ description=_DESCRIPTION,
119
+ features=features,
120
+ supervised_keys=None,
121
+ homepage=_HOMEPAGE,
122
+ )
123
+
124
+ def _split_generators(self, dl_manager):
125
+ train_data_dir = dl_manager.download([self.config.filepath.replace('.jsonl.xz', f'_train_{idx+1}.jsonl.xz')
126
+ for idx in range(self.config.n_files)])
127
+ # test_data_dir = dl_manager.download([self.config.filepath.replace('.jsonl.xz', 'test.jsonl.xz')])
128
+
129
+ return [
130
+ datasets.SplitGenerator(
131
+ name=datasets.Split.TRAIN,
132
+ gen_kwargs={"filepaths": train_data_dir, "split": 'train'},
133
+ ),
134
+ # datasets.SplitGenerator(
135
+ # name=datasets.Split.TEST,
136
+ # gen_kwargs={"filepaths": test_data_dir, "split": 'test'},
137
+ # )
138
+ ]
139
+
140
+ def _generate_examples(self, filepaths, split):
141
+ """
142
+ Reads line by line samples and generates examples.
143
+ :param filepath: Path to jsonl files with line by line examples.
144
+ """
145
+ id_ = 0
146
+ for filepath in filepaths:
147
+ logger.info("⏳ Generating examples from = %s", filepath)
148
+ with xz.open(open(filepath, "rb"), "rt", encoding='utf-8') as f:
149
+ for row in f:
150
+ try:
151
+ data = json.loads(row)
152
+ yield id_, {
153
+ "text": data["text"]
154
+ }
155
+ id_ += 1
156
+ except:
157
+ continue
158
+
us-contracts_train_1.jsonl.xz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6ece6ca34ea8f3bad3382ba646b7df27e5f3bc0f8793c25fdf0bdeeac7f7ce3
3
+ size 762802272