Datasets:
Update files from the datasets library (from 1.18.0)
Browse filesRelease notes: https://github.com/huggingface/datasets/releases/tag/1.18.0
README.md
CHANGED
@@ -1,5 +1,27 @@
|
|
1 |
---
|
|
|
|
|
|
|
|
|
|
|
2 |
paperswithcode_id: null
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
4 |
|
5 |
# Dataset Card for "wiki_dpr"
|
@@ -44,6 +66,9 @@ This is the wikipedia split used to evaluate the Dense Passage Retrieval (DPR) m
|
|
44 |
It contains 21M passages from wikipedia along with their DPR embeddings.
|
45 |
The wikipedia articles were split into multiple, disjoint text blocks of 100 words as passages.
|
46 |
|
|
|
|
|
|
|
47 |
### Supported Tasks and Leaderboards
|
48 |
|
49 |
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
@@ -58,6 +83,8 @@ We show detailed information for up to 5 configurations of the dataset.
|
|
58 |
|
59 |
### Data Instances
|
60 |
|
|
|
|
|
61 |
#### psgs_w100.multiset.compressed
|
62 |
|
63 |
- **Size of downloaded dataset files:** 67678.16 MB
|
@@ -68,12 +95,20 @@ An example of 'train' looks as follows.
|
|
68 |
```
|
69 |
This example was too long and was cropped:
|
70 |
|
71 |
-
{
|
72 |
-
|
73 |
-
|
74 |
-
|
75 |
-
|
76 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
77 |
```
|
78 |
|
79 |
#### psgs_w100.multiset.exact
|
@@ -86,12 +121,20 @@ An example of 'train' looks as follows.
|
|
86 |
```
|
87 |
This example was too long and was cropped:
|
88 |
|
89 |
-
{
|
90 |
-
|
91 |
-
|
92 |
-
|
93 |
-
|
94 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
95 |
```
|
96 |
|
97 |
#### psgs_w100.multiset.no_index
|
@@ -104,12 +147,20 @@ An example of 'train' looks as follows.
|
|
104 |
```
|
105 |
This example was too long and was cropped:
|
106 |
|
107 |
-
{
|
108 |
-
|
109 |
-
|
110 |
-
|
111 |
-
|
112 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
113 |
```
|
114 |
|
115 |
#### psgs_w100.nq.compressed
|
@@ -122,12 +173,20 @@ An example of 'train' looks as follows.
|
|
122 |
```
|
123 |
This example was too long and was cropped:
|
124 |
|
125 |
-
{
|
126 |
-
|
127 |
-
|
128 |
-
|
129 |
-
|
130 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
131 |
```
|
132 |
|
133 |
#### psgs_w100.nq.exact
|
@@ -140,12 +199,20 @@ An example of 'train' looks as follows.
|
|
140 |
```
|
141 |
This example was too long and was cropped:
|
142 |
|
143 |
-
{
|
144 |
-
|
145 |
-
|
146 |
-
|
147 |
-
|
148 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
149 |
```
|
150 |
|
151 |
### Data Fields
|
@@ -264,4 +331,4 @@ The data fields are the same among all splits.
|
|
264 |
|
265 |
### Contributions
|
266 |
|
267 |
-
Thanks to [@thomwolf](https://github.com/thomwolf), [@lewtun](https://github.com/lewtun), [@lhoestq](https://github.com/lhoestq) for adding this dataset.
|
|
|
1 |
---
|
2 |
+
annotations_creators:
|
3 |
+
- no-annotation
|
4 |
+
language_creators:
|
5 |
+
- crowdsourced
|
6 |
+
pretty_name: Wiki-DPR
|
7 |
paperswithcode_id: null
|
8 |
+
licenses:
|
9 |
+
- cc-by-sa-3-0
|
10 |
+
- gfdl-1-3-or-later
|
11 |
+
task_categories:
|
12 |
+
- sequence-modeling
|
13 |
+
- other-text-search
|
14 |
+
task_ids:
|
15 |
+
- language-modeling
|
16 |
+
- other-neural-search
|
17 |
+
source_datasets:
|
18 |
+
- original
|
19 |
+
multilinguality:
|
20 |
+
- multilingual
|
21 |
+
size_categories:
|
22 |
+
- 10M<n<100M
|
23 |
+
languages:
|
24 |
+
- en
|
25 |
---
|
26 |
|
27 |
# Dataset Card for "wiki_dpr"
|
|
|
66 |
It contains 21M passages from wikipedia along with their DPR embeddings.
|
67 |
The wikipedia articles were split into multiple, disjoint text blocks of 100 words as passages.
|
68 |
|
69 |
+
The wikipedia dump is the one from Dec. 20, 2018.
|
70 |
+
|
71 |
+
|
72 |
### Supported Tasks and Leaderboards
|
73 |
|
74 |
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
|
|
83 |
|
84 |
### Data Instances
|
85 |
|
86 |
+
Each instance contains a paragraph of at most 100 words, as well as the title of the wikipedia page it comes from, and the DPR embedding (a 768-d vector).
|
87 |
+
|
88 |
#### psgs_w100.multiset.compressed
|
89 |
|
90 |
- **Size of downloaded dataset files:** 67678.16 MB
|
|
|
95 |
```
|
96 |
This example was too long and was cropped:
|
97 |
|
98 |
+
{'id': '1',
|
99 |
+
'text': 'Aaron Aaron ( or ; "Ah盲r么n") is a prophet, high priest, and the brother of Moses in the Abrahamic religions. Knowledge of Aaron, along with his brother Moses, comes exclusively from religious texts, such as the Bible and Quran. The Hebrew Bible relates that, unlike Moses, who grew up in the Egyptian royal court, Aaron and his elder sister Miriam remained with their kinsmen in the eastern border-land of Egypt (Goshen). When Moses first confronted the Egyptian king about the Israelites, Aaron served as his brother\'s spokesman ("prophet") to the Pharaoh. Part of the Law (Torah) that Moses received from'],
|
100 |
+
'title': 'Aaron',
|
101 |
+
'embeddings': [-0.07233893871307373,
|
102 |
+
0.48035329580307007,
|
103 |
+
0.18650995194911957,
|
104 |
+
-0.5287084579467773,
|
105 |
+
-0.37329429388046265,
|
106 |
+
0.37622880935668945,
|
107 |
+
0.25524479150772095,
|
108 |
+
...
|
109 |
+
-0.336689829826355,
|
110 |
+
0.6313082575798035,
|
111 |
+
-0.7025573253631592]}
|
112 |
```
|
113 |
|
114 |
#### psgs_w100.multiset.exact
|
|
|
121 |
```
|
122 |
This example was too long and was cropped:
|
123 |
|
124 |
+
{'id': '1',
|
125 |
+
'text': 'Aaron Aaron ( or ; "Ah盲r么n") is a prophet, high priest, and the brother of Moses in the Abrahamic religions. Knowledge of Aaron, along with his brother Moses, comes exclusively from religious texts, such as the Bible and Quran. The Hebrew Bible relates that, unlike Moses, who grew up in the Egyptian royal court, Aaron and his elder sister Miriam remained with their kinsmen in the eastern border-land of Egypt (Goshen). When Moses first confronted the Egyptian king about the Israelites, Aaron served as his brother\'s spokesman ("prophet") to the Pharaoh. Part of the Law (Torah) that Moses received from'],
|
126 |
+
'title': 'Aaron',
|
127 |
+
'embeddings': [-0.07233893871307373,
|
128 |
+
0.48035329580307007,
|
129 |
+
0.18650995194911957,
|
130 |
+
-0.5287084579467773,
|
131 |
+
-0.37329429388046265,
|
132 |
+
0.37622880935668945,
|
133 |
+
0.25524479150772095,
|
134 |
+
...
|
135 |
+
-0.336689829826355,
|
136 |
+
0.6313082575798035,
|
137 |
+
-0.7025573253631592]}
|
138 |
```
|
139 |
|
140 |
#### psgs_w100.multiset.no_index
|
|
|
147 |
```
|
148 |
This example was too long and was cropped:
|
149 |
|
150 |
+
{'id': '1',
|
151 |
+
'text': 'Aaron Aaron ( or ; "Ah盲r么n") is a prophet, high priest, and the brother of Moses in the Abrahamic religions. Knowledge of Aaron, along with his brother Moses, comes exclusively from religious texts, such as the Bible and Quran. The Hebrew Bible relates that, unlike Moses, who grew up in the Egyptian royal court, Aaron and his elder sister Miriam remained with their kinsmen in the eastern border-land of Egypt (Goshen). When Moses first confronted the Egyptian king about the Israelites, Aaron served as his brother\'s spokesman ("prophet") to the Pharaoh. Part of the Law (Torah) that Moses received from'],
|
152 |
+
'title': 'Aaron',
|
153 |
+
'embeddings': [-0.07233893871307373,
|
154 |
+
0.48035329580307007,
|
155 |
+
0.18650995194911957,
|
156 |
+
-0.5287084579467773,
|
157 |
+
-0.37329429388046265,
|
158 |
+
0.37622880935668945,
|
159 |
+
0.25524479150772095,
|
160 |
+
...
|
161 |
+
-0.336689829826355,
|
162 |
+
0.6313082575798035,
|
163 |
+
-0.7025573253631592]}
|
164 |
```
|
165 |
|
166 |
#### psgs_w100.nq.compressed
|
|
|
173 |
```
|
174 |
This example was too long and was cropped:
|
175 |
|
176 |
+
{'id': '1',
|
177 |
+
'text': 'Aaron Aaron ( or ; "Ah盲r么n") is a prophet, high priest, and the brother of Moses in the Abrahamic religions. Knowledge of Aaron, along with his brother Moses, comes exclusively from religious texts, such as the Bible and Quran. The Hebrew Bible relates that, unlike Moses, who grew up in the Egyptian royal court, Aaron and his elder sister Miriam remained with their kinsmen in the eastern border-land of Egypt (Goshen). When Moses first confronted the Egyptian king about the Israelites, Aaron served as his brother\'s spokesman ("prophet") to the Pharaoh. Part of the Law (Torah) that Moses received from'],
|
178 |
+
'title': 'Aaron',
|
179 |
+
'embeddings': [0.013342111371457577,
|
180 |
+
0.582173764705658,
|
181 |
+
-0.31309744715690613,
|
182 |
+
-0.6991612911224365,
|
183 |
+
-0.5583199858665466,
|
184 |
+
0.5187504887580872,
|
185 |
+
0.7152731418609619,
|
186 |
+
...
|
187 |
+
-0.5385938286781311,
|
188 |
+
0.8093984127044678,
|
189 |
+
-0.4741983711719513]}
|
190 |
```
|
191 |
|
192 |
#### psgs_w100.nq.exact
|
|
|
199 |
```
|
200 |
This example was too long and was cropped:
|
201 |
|
202 |
+
{'id': '1',
|
203 |
+
'text': 'Aaron Aaron ( or ; "Ah盲r么n") is a prophet, high priest, and the brother of Moses in the Abrahamic religions. Knowledge of Aaron, along with his brother Moses, comes exclusively from religious texts, such as the Bible and Quran. The Hebrew Bible relates that, unlike Moses, who grew up in the Egyptian royal court, Aaron and his elder sister Miriam remained with their kinsmen in the eastern border-land of Egypt (Goshen). When Moses first confronted the Egyptian king about the Israelites, Aaron served as his brother\'s spokesman ("prophet") to the Pharaoh. Part of the Law (Torah) that Moses received from'],
|
204 |
+
'title': 'Aaron',
|
205 |
+
'embeddings': [0.013342111371457577,
|
206 |
+
0.582173764705658,
|
207 |
+
-0.31309744715690613,
|
208 |
+
-0.6991612911224365,
|
209 |
+
-0.5583199858665466,
|
210 |
+
0.5187504887580872,
|
211 |
+
0.7152731418609619,
|
212 |
+
...
|
213 |
+
-0.5385938286781311,
|
214 |
+
0.8093984127044678,
|
215 |
+
-0.4741983711719513]}
|
216 |
```
|
217 |
|
218 |
### Data Fields
|
|
|
331 |
|
332 |
### Contributions
|
333 |
|
334 |
+
Thanks to [@thomwolf](https://github.com/thomwolf), [@lewtun](https://github.com/lewtun), [@lhoestq](https://github.com/lhoestq) for adding this dataset.
|