Update README.md

README.md

@@ -132,66 +132,76 @@ The COLLE dataset suite hosts multiple French NLP benchmark tasks for evaluating
 
 ## Task Descriptions
 
-## Allo-ciné:
+## Allocine.fr :
 Allo-ciné tests language understanding in sentiment classification by feeding movie reviews which can be either positive and negative, the task consists in giving the correct sentiment for each review.
-## paws_x:
+## PAWS-X :
 This task aims to test paraphrase identification by giving two sentences and a label defining if these sentences are equivalent in meaning or not.
 
-## fquad:
+## FQuAD:
 Fquad is question/answer pair built on high-quality wikipedia articles. The goal of the model in this task is to accurately predict if the answer to the question really can be found in the provided answer.
 
-## gqnli: 
+## GQNLI-fr: 
 The dataset consists of carefully constructed premise-hypothesis pairs that involve quantifier logic (e.g., most, at least, more than half). The goal is to evaluate the model's ability to reason about these expressions and determine whether the hypothesis logically follows from the premise, contradicts it, or is neutral.
 
-## piaf: 
+## PIAF: 
 This task consists of pairs of questions and text answers with information of where in the answer is the truly relevant information.
 
-## sickfr:
+## SICK-fr :
 This task also has pairs of sentences and notes them on 2 dimensions, relatedness and entailment. While relatedness scales from 1 to 5, entailement is a choice between entails, contradicts or neutral.
 
-## xnli:
+## XNLI-fr:
 This task consists of pairs of sentences where the goal is to determine the relation between the two sentences, this relation can be either entailement, neutral or contradiction.
 
-## qfrcola: 
+## QFrCoLA: 
 QFrCoLA is a french dataset made from multiple french language sites such as académie-française.fr and vitrinelinguistique.com. It aims to tests models ability to determine a sentence's acceptability in french on subjects such as grammar and syntax. The answer is a binary label indicating if the sentence is correct or not.
 
-## qfrblimp:
+## QFrBLiMP:
 This task gives the model sentences pairs, the goal is to determine if the sentences are semantically equivalent, or, put more simply, if they mean the same thing, even with slightly different syntax and words.
 
-## sts22: 
+## STS22: 
 This task evaluates whether pairs of news articles, written in different languages, cover the same story. It focuses on document-level similarity, where systems rate article pairs on a 4-point scale from most to least similar
 
-## wino_x_lm
+## Wino-X-LM
 Pronoun resolution task: choose between two referents in a sentence with an ambiguous pronoun.
 
-## wino_x_mt
+## Wino-X-MT
 Translation-based pronoun resolution: choose which of two French translations uses the correct gendered pronoun.
 
-## expressions_quebecoises
-Given a Quebec proverb, predict its corresponding French equivalent.
 
-## termes_quebecoises
-Translate a Quebec French term to standard French or English based on a phrase.
+## QFrCoRE
+QFrCoRE is a definition matching task where the model selects the correct standard French  definition for a Quebec French expression from a list of candidates.
+## QFrCoRT
+QFrCoRE is a definition matching task where the model selects the correct standard French  definition for a Quebec French term from a list of candidates.
+
+## FraCaS 
+
+Fracas is a natural language inference (NLI) taskthe where the model must classify the relationship between a premise and a hypothesis—entailment, contradiction, or neutral—based on complex linguistic phenomena such as quantifiers, plurality, anaphora, and ellipsis.
 
-## daccord
+## DACCORD
 Determine if a French sentence makes sense semantically (binary label).
 
-## french_boolq
+## Fr-BoolQ
 Boolean question answering in French: answer true/false based on context.
 
-## mnli-nineeleven-fr-mt
+## MNLI-nineeleven-Fr-MT 
 French machine-translated version of MNLI using 9/11 context, for entailment classification.
 
-## rte3-french
+## RTE3-Fr
 French version of RTE3 for textual entailment recognition.
 
+## MultiBLiMP-Fr 
+MultiBLiMP-Fr is a grammatical judgment task where the model must identify the grammatically correct sentence from a minimal pair differing by a single targeted feature, thereby assessing its knowledge of French syntax, morphology, and agreement.
 
+## MMS-fr 
+MMS-fr is a sentiment analysis task where the model classifies a French text as positive (2), neutral (1), or negative (0), assessing its ability to detect sentiment across diverse domains and sources.
+## WSD-Fr
+WSD-Fr is a word sense disambiguation task where the model must identify the correct meaning of an ambiguous verb in context, as part of the FLUE benchmark.
 
 ## Language
 The language data in COLLE is in French 
 
 ### Dataset structure
-## Allo-ciné:
+## Allocine.fr:
 ```json
 
 {
@@ -199,7 +209,7 @@ The language data in COLLE is in French
   "label": 1
 }
 ```
-## paws_x:
+## PAWS-X :
  ```json
  {
   "id": 12,
@@ -208,7 +218,7 @@ The language data in COLLE is in French
   "label": 0
 }
  ```
-## fquad:
+## FQuAD:
  ```json
 {
   "title": "pégase_23_3",
@@ -228,7 +238,7 @@ The language data in COLLE is in French
   "sent2": "Le gâteau à la glace."
 }
 ```
-## gqnli: 
+## GQNLI-fr: 
 ```json
 {
   "uid": 214,
@@ -240,7 +250,7 @@ The language data in COLLE is in French
   "hypothesis_original": "One beige bear runs."
 }
 ```
-## piaf: 
+## PIAF: 
 ```json
 {
 
@@ -251,7 +261,7 @@ The language data in COLLE is in French
   "answer_start":[222]}
 }
 ```
-## sickfr:
+## SICK-fr :
 ```json
     {
       "Unnamed: 0": 5,
@@ -261,7 +271,7 @@ The language data in COLLE is in French
       "sentence_B": "Il n'y a pas de lutte et d'étreinte de chiens."
     }
  ```
-## xnli:
+## XNLI-fr:
 ```json
 
     {
@@ -270,8 +280,7 @@ The language data in COLLE is in French
       "label": 1
     }
  ```
-
-## qfrcola:
+## QFrCoLA:
  ```json
     {
   "label": 1,
@@ -280,7 +289,7 @@ The language data in COLLE is in French
   "category": "anglicism"
 }
 ```
-## wino_x_lm:
+## Wino-X-LM:
  ```json
 {
   "qID": "3UDTAB6HH8D37OQL3O6F3GXEEOF09Z-1",
@@ -296,7 +305,7 @@ The language data in COLLE is in French
   "context_referent_of_option2_fr": "vase"
 }
 ```
-## wino_x_mt
+## Wino-X-MT
  ```json
 {
   "qID": "3FULMHZ7OUVKJ7S9R0LMS753751M44-1",
@@ -314,7 +323,7 @@ The language data in COLLE is in French
   "false_translation_referent_of_pronoun2_fr": "arme"
 }
 ```
-## rte3-french
+## RTE3-Fr
  ```json
 {
   "id": "1",
@@ -329,7 +338,7 @@ The language data in COLLE is in French
 
 ```
 
-## daccord
+## DACCORD
  ```json
 {
   "id": "a001",
@@ -341,7 +350,7 @@ The language data in COLLE is in French
   "genre": "conflit ukrainien-russe"
 }
 ```
-## qfrblimp:
+## QFrBLiMP:
    ```json
     {
   "id": 250,
@@ -359,7 +368,7 @@ The language data in COLLE is in French
   "answer": "accept"
 }
    ```
-## sts22:
+## STS22:
  ```json
 {
   "id": "1559147599_1558534688",
@@ -368,7 +377,7 @@ The language data in COLLE is in French
   "sentence2": "Le décret n° 2020-293 du 23 mars 2020..."
 }
 ```
-## french_boolq
+## Fr-BoolQ
  ```json
 
 {
@@ -377,7 +386,7 @@ The language data in COLLE is in French
   "label": 1
 }
 ```
-## expressions_quebecoises
+## QFrCoRE
 
  ```json
 {
@@ -398,7 +407,55 @@ The language data in COLLE is in French
   "reference": "https://canada-media.ca/expressions-quebecoises/"
 }
 ```
-## mnli-nineeleven-fr-mt
+## QFrCoRT
+
+ ```json
+{
+  "terme": "Avoir la chienne",
+  "choices": [
+    "Prendre une chaise et s'asseoir.",
+    "Avoir du plaisir, parfois avec une connotation sexuelle.",
+    "Prépare-toi, ça va brasser.",
+    "Tomber amoureux.",
+    "Être en pleine forme.",
+    "Critiquer sévèrement.",
+    "Personne inefficace, qui ne travaille pas bien.",
+    "Il se comporte mal en public.",
+    "Se détendre, arrêter de s'énerver.",
+    "Avoir peur."
+  ],
+  "correct_index": 9,
+  "reference": "https://canada-media.ca/expressions-quebecoises/"
+}
+```
+## FraCaS
+```json
+{
+  "id": "1",
+  "premise": "Un Italien est devenu le plus grand ténor du monde.",
+  "hypothesis": "Il y a eu un Italien qui est devenu le plus grand ténor du monde.",
+  "label": "0",
+  "question": "Y a-t-il eu un Italien qui soit devenu le plus grand ténor du monde ?",
+  "answer": "yes",
+  "premises_original": "An Italian became the world's greatest tenor.",
+  "premise1": "Un Italien est devenu le plus grand ténor du monde.",
+  "premise1_original": "An Italian became the world's greatest tenor.",
+  "premise2": "",
+  "premise2_original": "",
+  "premise3": "",
+  "premise3_original": "",
+  "premise4": "",
+  "premise4_original": "",
+  "premise5": "",
+  "premise5_original": "",
+  "hypothesis_original": "There was an Italian who became the world's greatest tenor.",
+  "question_original": "Was there an Italian who became the world's greatest tenor?",
+  "note": "",
+  "topic": "GENERALIZED QUANTIFIERS"
+}
+```
+
+## MNLI-nineeleven-Fr-MT 
  ```json
 {
   "premise": "La faillite du nationalisme laïque et autocratique était évidente dans le monde musulman à la fin des années 1970.",
@@ -412,31 +469,58 @@ The language data in COLLE is in French
 }
 
 ```
-## Allo-ciné:
+## MultiBLiMP-Fr
+
+ ```json
+{
+  "sentence_a": "C'est le genre à lequel appartiennent les espèces de kiwi.",
+  "sentence_b": "C'est le genre à lequel appartenez les espèces de kiwi.",
+  "label": 0
+}
+```
+## WSD-Fr 
+
+ ```json
+{
+  "sentence": "Il rend hommage au roi de France et des négociations aboutissent au traité du Goulet , formalisant la paix entre les deux pays .",
+  "labels_idx": [10],
+  "label": "négociations"
+}
+
+```
+## MMS
+
+ ```json
+{
+  "text": "Cadeaux pour ma fille.",
+  "label": 2
+}
+```
+## Allocine.fr:
 | split      | # examples |
 |------------|-----------:|
 | train      |             |
 | validation |       20000 |
 | test       |       20000 |
-## paws_x:
+## PAWS-X :
 | split      | # examples |
 |------------|-----------:|
 | train      |     49401  |
 | validation |       2000 |
 | test       |       2000 |
-## fquad:
+## FQuAD:
 | split      | # examples |
 |------------|-----------:|
 | validation |       100 |
 | test       |       400 |
 
-## gqnli: 
+## GQNLI-fr: 
 | split      | # examples |
 |------------|-----------:|
 | train      |       243 |
 | validation |       27 |
 | test       |       30 |
-## piaf: 
+## PIAF: 
 | split      | # examples |
 |------------|-----------:|
 | train      |       3105 |
@@ -448,49 +532,51 @@ The language data in COLLE is in French
 | train      |       4439 |
 | validation |       495 |
 | test       |       4906 |
-## xnli:
+## XNLI-fr:
 | split      | # examples |
 |------------|-----------:|
-| train      |       NA |
+| train      |       393,000 |
 | validation |       2490 |
 | test       |       5010 |
-## qfrcola: 
+## QFrCoLA: 
 | split      | # examples |
 |------------|-----------:|
 | train      |       15846 |
 | validation |       1761 |
 | test       |       7546 |
-## qfrblimp:
+## QFrBLiMP:
 | split      | # examples |
 |------------|-----------:|
 | train      |       NA |
 | validation |       2061 |
 | test       |       2290 |
-## sts22: 
+## STS22: 
 
 | split      | # examples |
 |------------|-----------:|
 | train      |       101 |
 | test       |       72 |
 
-## french_boolq
+## Fr-BoolQ
 
 | split      | # examples |
 |------------|-----------:|
 | test      |       178 |
-## sts22: 
+## SICK-fr : 
 
 | split      | # examples |
 |------------|-----------:|
-| train      |       101 |
-| test       |       72 |
-## mnli-nineeleven-fr-mt
+| train      |       4,439  |
+| test       |       2,000 |
+| validation |       2,000 |
+
+## MNLI-nineeleven-Fr-MT 
 
 | split      | # examples |
 |------------|-----------:|
 | test       |       2000 |
 
-## daccord
+## DACCORD
 | split      | # examples |
 |------------|-----------:|
 | test       |       1034 |
@@ -498,15 +584,46 @@ The language data in COLLE is in French
 | split      | # examples |
 |------------|-----------:|
 | test       |       800 |
-| dev       |       800 |
-## wino_x_lm
+| validation       |       800 |
+## Wino-X-LM
 | split      | # examples |
 |------------|-----------:|
 | test       |       2793 |
-## wino_x_mt
+## Wino-X-MT
 | split      | # examples |
 |------------|-----------:|
 | test       |       2988 |
+## QFrCoRE
+| split      | # examples |
+|------------|-----------:|
+| test       |       4,633 |
+## QFrCoRT
+| split      | # examples |
+|------------|-----------:|
+| test       |       201 |
+## MultiBLiMP-Fr : 
+
+| split      | # examples |
+|------------|-----------:|
+| train      |       160  |
+| test       |       77 |
+| validation |       18 |
+## MMS: 
+
+| split      | # examples |
+|------------|-----------:|
+| train      |      132,696  |
+| test       |      63,190 |
+| validation |      14,745 |
+## FraCaS
+| split      | # examples |
+|------------|-----------:|
+| test       |       346 |
+## rte3-french
+| split      | # examples |
+|------------|-----------:|
+| test       |       3,121 |
+| train       |       269,821  |