orionweller commited on
Commit
16a506c
1 Parent(s): 51beb38

fixed readme

Browse files
Files changed (1) hide show
  1. neuclir-2022.py +46 -28
neuclir-2022.py CHANGED
@@ -1,20 +1,3 @@
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
- # Lint as: python3
17
-
18
  import json
19
 
20
  import datasets
@@ -38,13 +21,19 @@ _DESCRIPTION = 'dataset load script for NeuCLIR 2022'
38
 
39
  _DATASET_URLS = {
40
  lang: {
41
- 'test': f'https://huggingface.co/datasets/MTEB/neuclir-2022/resolve/main/neuclir-{lang}/queries.jsonl.gz',
42
  } for lang in _LANGUAGES
43
  }
44
 
45
  _DATASET_CORPUS_URLS = {
46
  f'corpus-{lang}': {
47
- 'corpus': f'https://huggingface.co/datasets/MTEB/neuclir-2022/resolve/main/neuclir-{lang}/corpus.jsonl.gz'
 
 
 
 
 
 
48
  } for lang in _LANGUAGES
49
  }
50
 
@@ -52,12 +41,17 @@ _DATASET_CORPUS_URLS = {
52
  class MLDR(datasets.GeneratorBasedBuilder):
53
  BUILDER_CONFIGS = [datasets.BuilderConfig(
54
  version=datasets.Version('1.0.0'),
55
- name=lang, description=f'MLDR dataset in language {lang}.'
56
  ) for lang in _LANGUAGES
57
  ] + [
58
  datasets.BuilderConfig(
59
  version=datasets.Version('1.0.0'),
60
- name=f'corpus-{lang}', description=f'corpus of MLDR dataset in language {lang}.'
 
 
 
 
 
61
  ) for lang in _LANGUAGES
62
  ]
63
 
@@ -68,10 +62,16 @@ class MLDR(datasets.GeneratorBasedBuilder):
68
  '_id': datasets.Value('string'),
69
  'text': datasets.Value('string'),
70
  })
71
- else:
72
  features = datasets.Features({
73
  '_id': datasets.Value('string'),
74
- 'query': datasets.Value('string'),
 
 
 
 
 
 
75
  })
76
 
77
  return datasets.DatasetInfo(
@@ -81,9 +81,9 @@ class MLDR(datasets.GeneratorBasedBuilder):
81
  features=features, # Here we define them above because they are different between the two configurations
82
  supervised_keys=None,
83
  # Homepage of the dataset for documentation
84
- homepage='https://github.com/FlagOpen/FlagEmbedding',
85
  # License for the dataset if available
86
- license='mit',
87
  # Citation for the dataset
88
  citation=_CITATION,
89
  )
@@ -100,6 +100,16 @@ class MLDR(datasets.GeneratorBasedBuilder):
100
  },
101
  ),
102
  ]
 
 
 
 
 
 
 
 
 
 
103
  else:
104
  downloaded_files = dl_manager.download_and_extract(_DATASET_URLS[name])
105
  splits = [
@@ -118,10 +128,18 @@ class MLDR(datasets.GeneratorBasedBuilder):
118
  with open(filepath, encoding='utf-8') as f:
119
  for line in f:
120
  data = json.loads(line)
121
- yield data['docid'], data
 
 
 
 
 
 
122
  else:
123
  with open(filepath, encoding="utf-8") as f:
124
  for line in f:
125
  data = json.loads(line)
126
- qid = data['query_id']
127
- yield qid, data
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  import json
2
 
3
  import datasets
 
21
 
22
  _DATASET_URLS = {
23
  lang: {
24
+ 'test': f'https://huggingface.co/datasets/MTEB/neuclir-2022/resolve/main/neuclir-{lang}/test.jsonl',
25
  } for lang in _LANGUAGES
26
  }
27
 
28
  _DATASET_CORPUS_URLS = {
29
  f'corpus-{lang}': {
30
+ 'corpus': f'https://huggingface.co/datasets/MTEB/neuclir-2022/resolve/main/neuclir-{lang}/corpus.jsonl'
31
+ } for lang in _LANGUAGES
32
+ }
33
+
34
+ _DATASET_QUERIES_URLS = {
35
+ f'queries-{lang}': {
36
+ 'queries': f'https://huggingface.co/datasets/MTEB/neuclir-2022/resolve/main/neuclir-{lang}/queries.jsonl'
37
  } for lang in _LANGUAGES
38
  }
39
 
 
41
  class MLDR(datasets.GeneratorBasedBuilder):
42
  BUILDER_CONFIGS = [datasets.BuilderConfig(
43
  version=datasets.Version('1.0.0'),
44
+ name=lang, description=f'NeuCLIR dataset in language {lang}.'
45
  ) for lang in _LANGUAGES
46
  ] + [
47
  datasets.BuilderConfig(
48
  version=datasets.Version('1.0.0'),
49
+ name=f'corpus-{lang}', description=f'corpus of NeuCLIR dataset in language {lang}.'
50
+ ) for lang in _LANGUAGES
51
+ ] + [
52
+ datasets.BuilderConfig(
53
+ version=datasets.Version('1.0.0'),
54
+ name=f'queries-{lang}', description=f'queries of NeuCLIR dataset in language {lang}.'
55
  ) for lang in _LANGUAGES
56
  ]
57
 
 
62
  '_id': datasets.Value('string'),
63
  'text': datasets.Value('string'),
64
  })
65
+ elif name.startswith("queries-"):
66
  features = datasets.Features({
67
  '_id': datasets.Value('string'),
68
+ 'text': datasets.Value('string'),
69
+ })
70
+ else:
71
+ features = datasets.Features({
72
+ 'query-id': datasets.Value('string'),
73
+ 'corpus-id': datasets.Value('string'),
74
+ 'score': datasets.Value('int32'),
75
  })
76
 
77
  return datasets.DatasetInfo(
 
81
  features=features, # Here we define them above because they are different between the two configurations
82
  supervised_keys=None,
83
  # Homepage of the dataset for documentation
84
+ homepage='https://arxiv.org/abs/2304.12367',
85
  # License for the dataset if available
86
+ license=None,
87
  # Citation for the dataset
88
  citation=_CITATION,
89
  )
 
100
  },
101
  ),
102
  ]
103
+ elif name.startswith("queries-"):
104
+ downloaded_files = dl_manager.download_and_extract(_DATASET_QUERIES_URLS[name])
105
+ splits = [
106
+ datasets.SplitGenerator(
107
+ name='queries',
108
+ gen_kwargs={
109
+ 'filepath': downloaded_files['queries'],
110
+ },
111
+ ),
112
+ ]
113
  else:
114
  downloaded_files = dl_manager.download_and_extract(_DATASET_URLS[name])
115
  splits = [
 
128
  with open(filepath, encoding='utf-8') as f:
129
  for line in f:
130
  data = json.loads(line)
131
+ yield data['_id'], data
132
+ elif name.startswith("queries-"):
133
+ with open(filepath, encoding="utf-8") as f:
134
+ for line in f:
135
+ data = json.loads(line)
136
+ qid = data['_id']
137
+ yield qid, data
138
  else:
139
  with open(filepath, encoding="utf-8") as f:
140
  for line in f:
141
  data = json.loads(line)
142
+ qid = data['query-id']
143
+ pid = data['corpus-id']
144
+ yield qid + "-----" + pid, data
145
+