jrsilva commited on
Commit
7487d10
1 Parent(s): 94ddb82

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +26 -12
README.md CHANGED
@@ -11,15 +11,29 @@ license: cc-by-4.0
11
  ExtraGLUE
12
  ===
13
 
14
- ExtraGLUE comprises 4 tasks from GLUE (viz. MRPC, RTE, STS-B, and WNLI) and 4 tasks from SuperGLUE (viz. BoolQ, CB, COPA, and MultiRC), automatically translated from English into two variants of Portuguese, namely American Portuguese and European Portuguese.
15
-
16
- The tasks cover different aspects of understanding, such as semantic similarity and entailment.
17
-
18
- - MRPC (Microsoft Research Paraphrase Corpus): A data set of sentence pairs, annotated as to whether they are paraphrases of each other.
19
- - RTE (Recognizing Textual Entailment): A data set of sentence pairs, annotated as to whether one (the premise) entails the other (the hypothesis).
20
- - STS-B (Semantic Textual Similarity Benchmark): A data set of sentence pairs annotated with a 0-5 score indicating the semantic similarity between the two sentences.
21
- - WNLI (Winograd Natural Language Inference): A data set of sentence pairs where the first sentence contains a pronoun whose referent must be correctly resolved in order to determine whether the first sentence entails the second sentence.
22
- - BoolQ (Boolean Questions): A data set of text excerpts and questions with yes/no answers.
23
- - CB (CommitmentBank): A data set of excerpt-clause pairs, where the clause has been extracted from the excerpt. Each pair is classified as to whether the excerpt implies, contradicts, or is neutral in relation to the clause.
24
- - COPA (Choice of Plausible Alternatives): A data set containing a premise, two alternative sentences, and a cause/effect indication. The task consists of indicating which of the two alternative sentences is the cause/effect of the premise.
25
- - MultiRC (Multi-Sentence Reading Comprehension): A data set where each instance consists of a context paragraph, a question about that paragraph, and an answer, labeled as to whether the answer is true or false. For the a given context paragraph there may be multiple questions, and for each question there may be multiple answers, some true and some false.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
11
  ExtraGLUE
12
  ===
13
 
14
+ ExtraGLUE is a Portuguese dataset obtained by the automatic translation of some of the tasks in the GLUE and SuperGLUE benchmarks.
15
+ The 14 datasets in extraGLUE cover different aspects of language understanding:
16
+
17
+ *Semantic similarity*
18
+
19
+ - SST-2
20
+ - MRPC
21
+ - STS-B
22
+
23
+ *Inference*
24
+
25
+ - MNLI: matched and mismatched
26
+ - QNLI
27
+ - RTE
28
+ - WNLI
29
+ - CB
30
+ - AB: b and g
31
+
32
+ *Question answering*
33
+
34
+ - BoolQ
35
+ - MultiRC
36
+
37
+ *Reasoning*
38
+
39
+ - CoPA