Datasets:

Modalities:
Text
Formats:
parquet
Languages:
French
ArXiv:
DOI:
Libraries:
Datasets
Dask
bourdoiscatie commited on
Commit
e02f4d6
1 Parent(s): 4951beb

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +18 -17
README.md CHANGED
@@ -1466,13 +1466,13 @@ In our case, for a given text, we have hidden one word per sentence of the text.
1466
  ## 24 prompts
1467
 
1468
 
1469
- 'Remplacer le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text,<br>
1470
- 'Remplace le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text, <br>
1471
  'Remplacez le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text, <br>
1472
  'Remplacer le \<mask\> dans le texte suivant par le mot le plus probable : '+text, <br>
1473
  'Remplace le \<mask\> dans le texte suivant par le mot le plus probable : '+text, <br>
1474
  'Remplacez le \<mask\> dans le texte suivant par le mot le plus probable : '+text, <br>
1475
- 'Substituer le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text,<br>
1476
  'Substitue le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text, <br>
1477
  'Substituez le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text, <br>
1478
  'Substituer le \<mask\> dans le texte suivant par le mot le plus probable : '+text, <br>
@@ -1499,7 +1499,7 @@ An example:
1499
 
1500
  ## Datasets
1501
  ### orange_sum
1502
- Note: we use the split `abstract`
1503
  **Original**: https://huggingface.co/datasets/orange_sum
1504
  #### Citation
1505
  Identical to the first citation of this dataset earlier in the card.
@@ -1521,7 +1521,7 @@ The aim is to generate a text positive ou negative depending on the prompt selec
1521
  ## 36 prompts
1522
 
1523
  <code>
1524
- # negative
1525
  'Rédiger un commentaire négatif dont le titre est : "'+title+'"".', <br>
1526
  'Rédige un commentaire négatif dont le titre est : "'+title+'"".', <br>
1527
  'Rédigez un commentaire négatif dont le titre est : "'+title+'"".', <br>
@@ -1558,7 +1558,7 @@ The aim is to generate a text positive ou negative depending on the prompt selec
1558
  'Titre : "'+title +'"\n Ecrire une évaluation négative de 1 à 5 phrases sur le titre précédent : ',<br>
1559
  'Titre : "'+title +'"\n Ecris une évaluation négative de 1 à 5 phrases sur le titre précédent : ',<br>
1560
  'Titre : "'+title +'"\n Ecrivez une évaluation négative de 1 à 5 phrases sur le titre précédent : ',<br>
1561
- # positive
1562
  'Rédiger un commentaire positif dont le titre est : '+title+'.',<br>
1563
  'Rédige un commentaire positif dont le titre est : '+title+'.',<br>
1564
  'Rédigez un commentaire positif dont le titre est : '+title+'.',<br>
@@ -1618,7 +1618,7 @@ The `https://huggingface.co/datasets/CATIE-AQ/french_book_reviews_fr_prompt_text
1618
 
1619
 
1620
  ### amazon_reviews_multi
1621
- **Original**: https://huggingface.co/datasets/amazon_reviews_multi
1622
  Note: only the French portion of this multilingual dataset is kept for our use.
1623
  #### Citation
1624
  Identical to the first citation of this dataset earlier in the card.
@@ -1745,7 +1745,7 @@ An example:
1745
  ## Datasets
1746
  ## Datasets
1747
  ### orange_sum
1748
- Note: we use the split `abstract`
1749
  **Original**: https://huggingface.co/datasets/orange_sum
1750
  #### Citation
1751
  Identical to the first citation of this dataset earlier in the card.
@@ -1794,7 +1794,7 @@ An example:
1794
  | Qualité très mauvaise. Après quelques semaines d'utilisation il était déjà cassé (sans l'avoir fait tomber) et il ne protège absolument pas le téléphone. Générez un titre pour cet avis : |Cassé après quelques semaines|
1795
 
1796
  ### amazon_reviews_multi
1797
- **Original**: https://huggingface.co/datasets/amazon_reviews_multi
1798
  Note: only the French portion of this multilingual dataset is kept for our use.
1799
  #### Citation
1800
  Identical to the first citation of this dataset earlier in the card.
@@ -1846,7 +1846,7 @@ An example:
1846
 
1847
  ## Datasets
1848
  ### amazon_reviews_multi
1849
- **Original**: https://huggingface.co/datasets/amazon_reviews_multi
1850
  Note: only the French portion of this multilingual dataset is kept for our use.
1851
  #### Citation
1852
  Identical to the first citation of this dataset earlier in the card.
@@ -1913,7 +1913,7 @@ The `https://huggingface.co/datasets/CATIE-AQ/french_book_reviews_fr_prompt_star
1913
 
1914
 
1915
  ### amazon_reviews_multi
1916
- **Original**: https://huggingface.co/datasets/amazon_reviews_multi
1917
  Note: only the French portion of this multilingual dataset is kept for our use.
1918
  #### Citation
1919
  Identical to the first citation of this dataset earlier in the card.
@@ -1994,7 +1994,7 @@ The `https://huggingface.co/datasets/CATIE-AQ/amazon_massive_intent_fr_prompt_in
1994
 
1995
 
1996
  ### mteb/mtop_domain
1997
- **Original**: https://huggingface.co/datasets/mteb/mtop_domain
1998
  Note: only the French portion of this multilingual dataset is kept for our use.
1999
  #### Citation
2000
  > @misc{li2021mtop,
@@ -2055,7 +2055,7 @@ An example:
2055
 
2056
  ## Datasets
2057
  ### tner/wikiann
2058
- **Original**: https://huggingface.co/datasets/tner/wikiann
2059
  Note: only the French portion of this multilingual dataset is kept for our use.
2060
  #### Citation
2061
  > @inproceedings{pan-etal-2017-cross,
@@ -2080,7 +2080,7 @@ The `https://huggingface.co/datasets/CATIE-AQ/wikiann_fr_prompt_ner` dataset has
2080
 
2081
 
2082
  ### tner/wikiann
2083
- **Original**: https://huggingface.co/datasets/tner/wikineural
2084
  Note: only the French portion of this multilingual dataset is kept for our use.
2085
  #### Citation
2086
  > @inproceedings{tedeschi-etal-2021-wikineural-combined,
@@ -2105,7 +2105,7 @@ The `https://huggingface.co/datasets/CATIE-AQ/wikineural_fr_prompt_ner` dataset
2105
 
2106
 
2107
  ### tner/multinerd
2108
- **Original**: https://huggingface.co/datasets/tner/multinerd
2109
  Note: only the French portion of this multilingual dataset is kept for our use.
2110
  #### Citation
2111
  > @inproceedings{tedeschi-navigli-2022-multinerd,
@@ -2193,7 +2193,7 @@ An example:
2193
 
2194
  #### Citation
2195
  ### universal_dependencies
2196
- **Original**: https://huggingface.co/datasets/universal_dependencies
2197
  Note: only the French portion of this multilingual dataset is kept for our use. These are the `fr_fqb`, `fr_gsd`, `fr_partut`, `fr_pud`, `fr_sequoia` and `fr_spoken` splits.
2198
  #### Citation
2199
  > @inproceedings{nivre-etal-2020-universal,
@@ -2227,4 +2227,5 @@ See the DOI at the end of this dataset card.
2227
  The `universal_dependencies_fr_fqb_fr_prompt_pos`, `universal_dependencies_fr_gsd_fr_prompt_pos`, `universal_dependencies_fr_partut_fr_prompt_pos`, `universal_dependencies_fr_pud_fr_prompt_pos`, `universal_dependencies_fr_sequoia_fr_prompt_pos`, `universal_dependencies_fr_spoken_fr_prompt_pos` datasets have the same license as the original dataset from which they are derived.</details>
2228
 
2229
 
2230
- # Citation
 
 
1466
  ## 24 prompts
1467
 
1468
 
1469
+ 'Remplacer le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text,<br>
1470
+ 'Remplace le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text, <br>
1471
  'Remplacez le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text, <br>
1472
  'Remplacer le \<mask\> dans le texte suivant par le mot le plus probable : '+text, <br>
1473
  'Remplace le \<mask\> dans le texte suivant par le mot le plus probable : '+text, <br>
1474
  'Remplacez le \<mask\> dans le texte suivant par le mot le plus probable : '+text, <br>
1475
+ 'Substituer le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text,<br>
1476
  'Substitue le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text, <br>
1477
  'Substituez le \<mask\> dans le texte suivant par le mot le plus vraisemblable : '+text, <br>
1478
  'Substituer le \<mask\> dans le texte suivant par le mot le plus probable : '+text, <br>
 
1499
 
1500
  ## Datasets
1501
  ### orange_sum
1502
+ Note: we use the split `abstract`.
1503
  **Original**: https://huggingface.co/datasets/orange_sum
1504
  #### Citation
1505
  Identical to the first citation of this dataset earlier in the card.
 
1521
  ## 36 prompts
1522
 
1523
  <code>
1524
+ # negative<br>
1525
  'Rédiger un commentaire négatif dont le titre est : "'+title+'"".', <br>
1526
  'Rédige un commentaire négatif dont le titre est : "'+title+'"".', <br>
1527
  'Rédigez un commentaire négatif dont le titre est : "'+title+'"".', <br>
 
1558
  'Titre : "'+title +'"\n Ecrire une évaluation négative de 1 à 5 phrases sur le titre précédent : ',<br>
1559
  'Titre : "'+title +'"\n Ecris une évaluation négative de 1 à 5 phrases sur le titre précédent : ',<br>
1560
  'Titre : "'+title +'"\n Ecrivez une évaluation négative de 1 à 5 phrases sur le titre précédent : ',<br>
1561
+ # positive<br>
1562
  'Rédiger un commentaire positif dont le titre est : '+title+'.',<br>
1563
  'Rédige un commentaire positif dont le titre est : '+title+'.',<br>
1564
  'Rédigez un commentaire positif dont le titre est : '+title+'.',<br>
 
1618
 
1619
 
1620
  ### amazon_reviews_multi
1621
+ **Original**: https://huggingface.co/datasets/amazon_reviews_multi
1622
  Note: only the French portion of this multilingual dataset is kept for our use.
1623
  #### Citation
1624
  Identical to the first citation of this dataset earlier in the card.
 
1745
  ## Datasets
1746
  ## Datasets
1747
  ### orange_sum
1748
+ Note: we use the split `abstract`.
1749
  **Original**: https://huggingface.co/datasets/orange_sum
1750
  #### Citation
1751
  Identical to the first citation of this dataset earlier in the card.
 
1794
  | Qualité très mauvaise. Après quelques semaines d'utilisation il était déjà cassé (sans l'avoir fait tomber) et il ne protège absolument pas le téléphone. Générez un titre pour cet avis : |Cassé après quelques semaines|
1795
 
1796
  ### amazon_reviews_multi
1797
+ **Original**: https://huggingface.co/datasets/amazon_reviews_multi
1798
  Note: only the French portion of this multilingual dataset is kept for our use.
1799
  #### Citation
1800
  Identical to the first citation of this dataset earlier in the card.
 
1846
 
1847
  ## Datasets
1848
  ### amazon_reviews_multi
1849
+ **Original**: https://huggingface.co/datasets/amazon_reviews_multi
1850
  Note: only the French portion of this multilingual dataset is kept for our use.
1851
  #### Citation
1852
  Identical to the first citation of this dataset earlier in the card.
 
1913
 
1914
 
1915
  ### amazon_reviews_multi
1916
+ **Original**: https://huggingface.co/datasets/amazon_reviews_multi
1917
  Note: only the French portion of this multilingual dataset is kept for our use.
1918
  #### Citation
1919
  Identical to the first citation of this dataset earlier in the card.
 
1994
 
1995
 
1996
  ### mteb/mtop_domain
1997
+ **Original**: https://huggingface.co/datasets/mteb/mtop_domain
1998
  Note: only the French portion of this multilingual dataset is kept for our use.
1999
  #### Citation
2000
  > @misc{li2021mtop,
 
2055
 
2056
  ## Datasets
2057
  ### tner/wikiann
2058
+ **Original**: https://huggingface.co/datasets/tner/wikiann
2059
  Note: only the French portion of this multilingual dataset is kept for our use.
2060
  #### Citation
2061
  > @inproceedings{pan-etal-2017-cross,
 
2080
 
2081
 
2082
  ### tner/wikiann
2083
+ **Original**: https://huggingface.co/datasets/tner/wikineural
2084
  Note: only the French portion of this multilingual dataset is kept for our use.
2085
  #### Citation
2086
  > @inproceedings{tedeschi-etal-2021-wikineural-combined,
 
2105
 
2106
 
2107
  ### tner/multinerd
2108
+ **Original**: https://huggingface.co/datasets/tner/multinerd
2109
  Note: only the French portion of this multilingual dataset is kept for our use.
2110
  #### Citation
2111
  > @inproceedings{tedeschi-navigli-2022-multinerd,
 
2193
 
2194
  #### Citation
2195
  ### universal_dependencies
2196
+ **Original**: https://huggingface.co/datasets/universal_dependencies
2197
  Note: only the French portion of this multilingual dataset is kept for our use. These are the `fr_fqb`, `fr_gsd`, `fr_partut`, `fr_pud`, `fr_sequoia` and `fr_spoken` splits.
2198
  #### Citation
2199
  > @inproceedings{nivre-etal-2020-universal,
 
2227
  The `universal_dependencies_fr_fqb_fr_prompt_pos`, `universal_dependencies_fr_gsd_fr_prompt_pos`, `universal_dependencies_fr_partut_fr_prompt_pos`, `universal_dependencies_fr_pud_fr_prompt_pos`, `universal_dependencies_fr_sequoia_fr_prompt_pos`, `universal_dependencies_fr_spoken_fr_prompt_pos` datasets have the same license as the original dataset from which they are derived.</details>
2228
 
2229
 
2230
+ # Citation
2231
+ Add DOI