bourdoiscatie commited on
Commit
9bd972a
1 Parent(s): 2e01145

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +19 -7
README.md CHANGED
@@ -3,16 +3,27 @@ language:
3
  - fr
4
  license: apache-2.0
5
  size_categories:
6
- - 10K<n<100K
7
  task_categories:
8
  - question-answering
 
 
 
 
 
 
 
 
 
 
 
9
  ---
10
 
11
  # squad_v2_french_translated_fr_prompt_qa
12
  ## Summary
13
 
14
- **squad_v2_french_translated_fr_prompt_qa** is a subset of the [**Dataset of French Prompts (DFP)**]().
15
- It contains **48,069** rows that can be used for a question-answering task.
16
  The original data (without prompts) comes from the dataset [pragnakalp/squad_v2_french_translated](https://huggingface.co/datasets/pragnakalp/squad_v2_french_translated) and was augmented by questions in SQUAD 2.0 format in the [FrenchQA]( https://huggingface.co/datasets/CATIE-AQ/frenchQA) dataset.
17
  A list of prompts (see below) was then applied in order to build the input and target columns and thus obtain the same format as the [xP3](https://huggingface.co/datasets/bigscience/xP3) dataset by Muennighoff et al.
18
 
@@ -22,7 +33,7 @@ A list of prompts (see below) was then applied in order to build the input and t
22
  42 prompts were created for this dataset. The logic applied consists in proposing prompts in the indicative tense, in the form of tutoiement and in the form of vouvoiement.
23
 
24
  ```
25
- # SQUAD V1 format
26
  'Question : "'+question+'"\nContexte : "'+context+'" Réponse :',
27
  'La réponse à la question "'+question+'" se trouve dans "'+context+'" Pouvez-vous me la dire ?',
28
  'La réponse à la question "'+question+'" se trouve dans "'+context+'" Peux-tu me la dire ?',
@@ -35,7 +46,7 @@ A list of prompts (see below) was then applied in order to build the input and t
35
  """La réponse à la question : " """+question+""" " se trouve dans le texte : " """+context+""" "\n Peux-tu l'indiquer ?""",
36
  """La réponse à la question : " """+question+""" " se trouve dans le texte : " """+context+""" "\n Pouvez-vous l'indiquer ?""",
37
  """La réponse à la question : " """+question+""" " se trouve dans le texte : " """+context+""" "\n Qu'elle est-elle ?""",
38
- # SQUAD V2 format
39
  '"'+question+'"\n Répondre à la question ci-dessus en se basant sur le contexte suivant : "'+context+'"\n Si vous ne trouvez pas la réponse, répondre "sans réponse".',
40
  '"'+question+'"\n Réponds à la question ci-dessus en te basant sur le contexte suivant : "'+context+'"\n Si tu ne trouves pas la réponse, répondre "sans réponse".',
41
  '"'+question+'"\n Répondez à la question ci-dessus en vous basant sur le contexte suivant : "'+context+'"\n Si vous ne trouvez pas la réponse, répondre "sans réponse".',
@@ -70,8 +81,9 @@ A list of prompts (see below) was then applied in order to build the input and t
70
 
71
 
72
  # Splits
73
- - train with 48,069 samples
74
- - no valid and test splits in the original data
 
75
 
76
 
77
  # How to use?
 
3
  - fr
4
  license: apache-2.0
5
  size_categories:
6
+ - 1M<n<10M
7
  task_categories:
8
  - question-answering
9
+ tags:
10
+ - DFP
11
+ - french prompts
12
+ annotations_creators:
13
+ - found
14
+ language_creators:
15
+ - found
16
+ multilinguality:
17
+ - monolingual
18
+ source_datasets:
19
+ - amazon_reviews_multi
20
  ---
21
 
22
  # squad_v2_french_translated_fr_prompt_qa
23
  ## Summary
24
 
25
+ **squad_v2_french_translated_fr_prompt_qa** is a subset of the [**Dataset of French Prompts (DFP)**](https://huggingface.co/datasets/CATIE-AQ/DFP).
26
+ It contains **3,320,898** rows that can be used for a question-answering task.
27
  The original data (without prompts) comes from the dataset [pragnakalp/squad_v2_french_translated](https://huggingface.co/datasets/pragnakalp/squad_v2_french_translated) and was augmented by questions in SQUAD 2.0 format in the [FrenchQA]( https://huggingface.co/datasets/CATIE-AQ/frenchQA) dataset.
28
  A list of prompts (see below) was then applied in order to build the input and target columns and thus obtain the same format as the [xP3](https://huggingface.co/datasets/bigscience/xP3) dataset by Muennighoff et al.
29
 
 
33
  42 prompts were created for this dataset. The logic applied consists in proposing prompts in the indicative tense, in the form of tutoiement and in the form of vouvoiement.
34
 
35
  ```
36
+ # SQUAD 1.0 format
37
  'Question : "'+question+'"\nContexte : "'+context+'" Réponse :',
38
  'La réponse à la question "'+question+'" se trouve dans "'+context+'" Pouvez-vous me la dire ?',
39
  'La réponse à la question "'+question+'" se trouve dans "'+context+'" Peux-tu me la dire ?',
 
46
  """La réponse à la question : " """+question+""" " se trouve dans le texte : " """+context+""" "\n Peux-tu l'indiquer ?""",
47
  """La réponse à la question : " """+question+""" " se trouve dans le texte : " """+context+""" "\n Pouvez-vous l'indiquer ?""",
48
  """La réponse à la question : " """+question+""" " se trouve dans le texte : " """+context+""" "\n Qu'elle est-elle ?""",
49
+ # SQUAD 2.0 format
50
  '"'+question+'"\n Répondre à la question ci-dessus en se basant sur le contexte suivant : "'+context+'"\n Si vous ne trouvez pas la réponse, répondre "sans réponse".',
51
  '"'+question+'"\n Réponds à la question ci-dessus en te basant sur le contexte suivant : "'+context+'"\n Si tu ne trouves pas la réponse, répondre "sans réponse".',
52
  '"'+question+'"\n Répondez à la question ci-dessus en vous basant sur le contexte suivant : "'+context+'"\n Si vous ne trouvez pas la réponse, répondre "sans réponse".',
 
81
 
82
 
83
  # Splits
84
+ - `train` with 3,320,898 samples
85
+ - no `valid` split
86
+ - no `test` split
87
 
88
 
89
  # How to use?