Update files from the datasets library (from 1.6.0)
Browse filesRelease notes: https://github.com/huggingface/datasets/releases/tag/1.6.0
README.md
CHANGED
@@ -1,4 +1,27 @@
|
|
1 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
---
|
3 |
|
4 |
# Dataset Card for "fquad"
|
@@ -30,9 +53,8 @@
|
|
30 |
## Dataset Description
|
31 |
|
32 |
- **Homepage:** [https://fquad.illuin.tech/](https://fquad.illuin.tech/)
|
33 |
-
- **
|
34 |
-
- **
|
35 |
-
- **Point of Contact:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
36 |
- **Size of downloaded dataset files:** 3.14 MB
|
37 |
- **Size of the generated dataset:** 6.62 MB
|
38 |
- **Total amount of disk used:** 9.76 MB
|
@@ -40,16 +62,19 @@
|
|
40 |
### Dataset Summary
|
41 |
|
42 |
FQuAD: French Question Answering Dataset
|
43 |
-
We introduce FQuAD, a native French Question Answering Dataset.
|
44 |
-
|
|
|
|
|
|
|
45 |
|
46 |
### Supported Tasks
|
47 |
|
48 |
-
|
49 |
|
50 |
### Languages
|
51 |
|
52 |
-
|
53 |
|
54 |
## Dataset Structure
|
55 |
|
@@ -90,51 +115,57 @@ The data fields are the same among all splits.
|
|
90 |
|
91 |
### Data Splits Sample Size
|
92 |
|
93 |
-
|
94 |
-
|
95 |
-
|
|
|
|
|
|
|
|
|
|
96 |
|
97 |
## Dataset Creation
|
98 |
|
99 |
### Curation Rationale
|
100 |
-
|
101 |
-
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
102 |
|
103 |
### Source Data
|
104 |
|
105 |
-
|
106 |
-
|
107 |
### Annotations
|
108 |
|
109 |
-
|
|
|
|
|
110 |
|
111 |
### Personal and Sensitive Information
|
112 |
|
113 |
-
|
114 |
|
115 |
## Considerations for Using the Data
|
116 |
|
|
|
|
|
117 |
### Social Impact of Dataset
|
118 |
|
119 |
-
|
120 |
|
121 |
### Discussion of Biases
|
122 |
|
123 |
-
|
124 |
|
125 |
### Other Known Limitations
|
126 |
|
127 |
-
|
128 |
|
129 |
## Additional Information
|
130 |
|
131 |
### Dataset Curators
|
132 |
|
133 |
-
[
|
134 |
|
135 |
### Licensing Information
|
136 |
|
137 |
-
[
|
138 |
|
139 |
### Citation Information
|
140 |
|
@@ -159,4 +190,5 @@ archivePrefix = {arXiv},
|
|
159 |
|
160 |
### Contributions
|
161 |
|
162 |
-
Thanks to [@thomwolf](https://github.com/thomwolf), [@mariamabarham](https://github.com/mariamabarham), [@patrickvonplaten](https://github.com/patrickvonplaten), [@lewtun](https://github.com/lewtun), [@albertvillanova](https://github.com/albertvillanova) for adding this dataset.
|
|
|
|
1 |
---
|
2 |
+
annotations_creators:
|
3 |
+
- crowdsourced
|
4 |
+
extended:
|
5 |
+
- original
|
6 |
+
language_creators:
|
7 |
+
- crowdsourced
|
8 |
+
- found
|
9 |
+
languages:
|
10 |
+
- fr
|
11 |
+
licenses:
|
12 |
+
- cc-by-nc-sa-3-0
|
13 |
+
multilinguality:
|
14 |
+
- monolingual
|
15 |
+
size_categories:
|
16 |
+
- 1K<n<10K
|
17 |
+
source_datasets:
|
18 |
+
- original
|
19 |
+
task_categories:
|
20 |
+
- question-answering
|
21 |
+
- text-retrieval
|
22 |
+
task_ids:
|
23 |
+
- extractive-qa
|
24 |
+
- closed-domain-qa
|
25 |
---
|
26 |
|
27 |
# Dataset Card for "fquad"
|
|
|
53 |
## Dataset Description
|
54 |
|
55 |
- **Homepage:** [https://fquad.illuin.tech/](https://fquad.illuin.tech/)
|
56 |
+
- **Paper:** [FQuAD: French Question Answering Dataset](https://arxiv.org/abs/2002.06071)
|
57 |
+
- **Point of Contact:** [https://www.illuin.tech/contact/](https://www.illuin.tech/contact/)
|
|
|
58 |
- **Size of downloaded dataset files:** 3.14 MB
|
59 |
- **Size of the generated dataset:** 6.62 MB
|
60 |
- **Total amount of disk used:** 9.76 MB
|
|
|
62 |
### Dataset Summary
|
63 |
|
64 |
FQuAD: French Question Answering Dataset
|
65 |
+
We introduce FQuAD, a native French Question Answering Dataset.
|
66 |
+
|
67 |
+
FQuAD contains 25,000+ question and answer pairs.
|
68 |
+
Finetuning CamemBERT on FQuAD yields a F1 score of 88% and an exact match of 77.9%.
|
69 |
+
Developped to provide a SQuAD equivalent in the French language. Questions are original and based on high quality Wikipedia articles.
|
70 |
|
71 |
### Supported Tasks
|
72 |
|
73 |
+
- `closed-domain-qa`, `text-retrieval`: This dataset is intended to be used for `closed-domain-qa`, but can also be used for information retrieval tasks.
|
74 |
|
75 |
### Languages
|
76 |
|
77 |
+
This dataset is exclusively in French, with context data from Wikipedia and questions from French university students (`fr`).
|
78 |
|
79 |
## Dataset Structure
|
80 |
|
|
|
115 |
|
116 |
### Data Splits Sample Size
|
117 |
|
118 |
+
The FQuAD dataset has 3 splits: _train_, _validation_, and _test_. The _test_ split is however not released publicly at the moment. The splits contain disjoint sets of articles. The following table contains stats about each split.
|
119 |
+
|
120 |
+
Dataset Split | Number of Articles in Split | Number of paragraphs in split | Number of questions in split
|
121 |
+
--------------|------------------------------|--------------------------|-------------------------
|
122 |
+
Train | 117 | 4921 | 20731
|
123 |
+
Validation | 768 | 51.0% | 3188
|
124 |
+
Test | 10 | 532 | 2189
|
125 |
|
126 |
## Dataset Creation
|
127 |
|
128 |
### Curation Rationale
|
129 |
+
The FQuAD dataset was created by Illuin technology. It was developped to provide a SQuAD equivalent in the French language. Questions are original and based on high quality Wikipedia articles.
|
|
|
130 |
|
131 |
### Source Data
|
132 |
|
133 |
+
The text used for the contexts are from the curated list of French High-Quality Wikipedia [articles](https://fr.wikipedia.org/wiki/Cat%C3%A9gorie:Article_de_qualit%C3%A9).
|
|
|
134 |
### Annotations
|
135 |
|
136 |
+
Annotations (spans and questions) are written by students of the CentraleSupélec school of engineering.
|
137 |
+
Wikipedia articles were scraped and Illuin used an internally-developped tool to help annotators ask questions and indicate the answer spans.
|
138 |
+
Annotators were given paragraph sized contexts and asked to generate 4/5 non-trivial questions about information in the context.
|
139 |
|
140 |
### Personal and Sensitive Information
|
141 |
|
142 |
+
No personal or sensitive information is included in this dataset. This has been manually verified by the dataset curators.
|
143 |
|
144 |
## Considerations for Using the Data
|
145 |
|
146 |
+
Users should consider this dataset is sampled from Wikipedia data which might not be representative of all QA use cases.
|
147 |
+
|
148 |
### Social Impact of Dataset
|
149 |
|
150 |
+
The social biases of this dataset have not yet been investigated.
|
151 |
|
152 |
### Discussion of Biases
|
153 |
|
154 |
+
The social biases of this dataset have not yet been investigated, though articles have been selected by their quality and objectivity.
|
155 |
|
156 |
### Other Known Limitations
|
157 |
|
158 |
+
The limitations of the FQuAD dataset have not yet been investigated.
|
159 |
|
160 |
## Additional Information
|
161 |
|
162 |
### Dataset Curators
|
163 |
|
164 |
+
Illuin Technology: [https://fquad.illuin.tech/](https://fquad.illuin.tech/)
|
165 |
|
166 |
### Licensing Information
|
167 |
|
168 |
+
The FQuAD dataset is licensed under the [CC BY-NC-SA 3.0](https://creativecommons.org/licenses/by-nc-sa/3.0/fr/) license.
|
169 |
|
170 |
### Citation Information
|
171 |
|
|
|
190 |
|
191 |
### Contributions
|
192 |
|
193 |
+
Thanks to [@thomwolf](https://github.com/thomwolf), [@mariamabarham](https://github.com/mariamabarham), [@patrickvonplaten](https://github.com/patrickvonplaten), [@lewtun](https://github.com/lewtun), [@albertvillanova](https://github.com/albertvillanova) for adding this dataset.
|
194 |
+
Thanks to [@ManuelFay](https://github.com/manuelfay) for providing information on the dataset creation process.
|
fquad.py
CHANGED
@@ -1,6 +1,5 @@
|
|
1 |
"""TODO(fquad): Add a description here."""
|
2 |
|
3 |
-
from __future__ import absolute_import, division, print_function
|
4 |
|
5 |
import json
|
6 |
import os
|
|
|
1 |
"""TODO(fquad): Add a description here."""
|
2 |
|
|
|
3 |
|
4 |
import json
|
5 |
import os
|