Datasets:

Modalities:
Text
Formats:
parquet
Languages:
French
Size:
< 1K
ArXiv:
Libraries:
Datasets
pandas
License:
bourdoiscatie commited on
Commit
09ae591
1 Parent(s): 8996905

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +87 -0
README.md ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - fr
4
+ license:
5
+ - cc-by-4.0
6
+ size_categories:
7
+ - 10K<n<100K
8
+ tags:
9
+ - coreference
10
+ ---
11
+
12
+ # xwinograd_fr_prompt_coreference
13
+ ## Summary
14
+
15
+ **xwinograd_fr_prompt_coreference** is a subset of the [**Dataset of French Prompts (DFP)**]().
16
+ It contains **X** rows that can be used for a coreference task.
17
+ The original data (without prompts) comes from the dataset [xwinograd](https://huggingface.co/datasets/Muennighoff/xwinograd) by Muennighoff. where only the French part has been kept.
18
+ 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.
19
+
20
+
21
+ ## Prompts used
22
+ ### List
23
+ 10 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.
24
+
25
+ ```
26
+ '"'+sentence+'"\nRemplacer le "_" dans la phrase ci-dessus par la bonne option :\n- "'+option1+'"\n- "'+option2+'"',
27
+ '"'+sentence+'"\nRemplace le "_" dans la phrase ci-dessus par la bonne option :\n- "'+option1+'"\n- "'+option2+'"',
28
+ '"'+sentence+'"\nRemplacez le "_" dans la phrase ci-dessus par la bonne option :\n- "'+option1+'"\n- "'+option2+'"',
29
+ '"'+sentence+'" Dans la phrase précédente, "_" fait-il référence à "'+option1+'" ou "'+option2+'" ?',
30
+ '"'+sentence+'" À quoi le "_" dans la phrase ci-dessus fait-il référence ? "'+option1+'" ou "'+option2+'" ?',
31
+ '"'+sentence+'" Le "_" dans la phrase ci-dessous fait référence à "'+option1+'"\n- "'+option2+'" ?',
32
+ 'Remplisser le "_" de la phrase suivante : "'+sentence+ '"\nChoix :\n- "'+option1+'"\n- "'+option2+'"\nRéponse :',
33
+ 'Remplis le "_" de la phrase suivante : "'+sentence+ '"\nChoix :\n- "'+option1+'"\n- "'+option2+'"\nRéponse :',
34
+ 'Remplissez le "_" de la phrase suivante : "'+sentence+ '"\nChoix :\n- "'+option1+'"\n- "'+option2+'"\nRéponse :',
35
+ 'Dans la phrase ci-dessous, le "_" renvoie-t-il à "'+option1+'" ou "'+option2+'" ? : '+sentence,
36
+ ```
37
+
38
+ ### Features used in the prompts
39
+ In the prompt list above, `option1`, `option2`, `sentence` and `targets` have been constructed from:
40
+ ```
41
+ xwinograd = load_dataset('Muennighoff/xwinograd','fr')
42
+ sentence = xwinograd['test'][i]['sentence']
43
+ option1 = xwinograd['test'][i]['option1']
44
+ option2 = xwinograd['test'][i]['option2']
45
+ targets = str(xwinograd['test'][i]['answer']).replace("1",xwinograd['test'][i]['option1']).replace("2",xwinograd['test'][i]['option2'])
46
+ ```
47
+
48
+
49
+ # Splits
50
+ - train with X samples
51
+ - dev with Y samples
52
+ - test with Z samples
53
+
54
+
55
+ # How to use?
56
+ ```
57
+ from datasets import load_dataset
58
+ dataset = load_dataset("CATIE-AQ/xwinograd_fr_prompt_coreference")
59
+ ```
60
+
61
+ # Citation
62
+ ## Original data
63
+ > @misc{muennighoff2022crosslingual,
64
+ title={Crosslingual Generalization through Multitask Finetuning},
65
+ author={Niklas Muennighoff and Thomas Wang and Lintang Sutawika and Adam Roberts and Stella Biderman and Teven Le Scao and M Saiful Bari and Sheng Shen and Zheng-Xin Yong and Hailey Schoelkopf and Xiangru Tang and Dragomir Radev and Alham Fikri Aji and Khalid Almubarak and Samuel Albanie and Zaid Alyafeai and Albert Webson and Edward Raff and Colin Raffel},
66
+ year={2022},
67
+ eprint={2211.01786},
68
+ archivePrefix={arXiv},
69
+ primaryClass={cs.CL}
70
+ }
71
+ > @misc{tikhonov2021heads,
72
+ title={It's All in the Heads: Using Attention Heads as a Baseline for Cross-Lingual Transfer in Commonsense Reasoning},
73
+ author={Alexey Tikhonov and Max Ryabinin},
74
+ year={2021},
75
+ eprint={2106.12066},
76
+ archivePrefix={arXiv},
77
+ primaryClass={cs.CL}
78
+ }
79
+
80
+
81
+
82
+ ## This Dataset
83
+
84
+
85
+
86
+ ## License
87
+ [cc-by-4.0](https://creativecommons.org/licenses/by/4.0/deed.en)