README.md

metadata

annotations_creators:
  - found
language_creators:
  - machine-generated
language:
  - pt
license:
  - lgpl-3.0
multilinguality:
  - monolingual
  - translation
size_categories:
  - 10K<n<100K
source_datasets:
  - extended|glue
task_categories:
  - text-classification
task_ids:
  - acceptability-classification
  - natural-language-inference
  - semantic-similarity-scoring
  - sentiment-classification
  - text-scoring
pretty_name: PLUE (Portuguese Language Understanding Evaluation benchmark)
tags:
  - paraphrase-identification
  - qa-nli
  - coreference-nli
configs:
  - config_name: CoLA
    data_files:
      - split: train
        path: datasets/CoLA/train.tsv
      - split: dev
        path: datasets/CoLA/dev.tsv
      - split: test
        path: datasets/CoLA/test.tsv
  - config_name: MRPC
    data_files:
      - split: train
        path: datasets/MRPC/train.tsv
      - split: dev
        path: datasets/MRPC/dev.tsv
      - split: test
        path: datasets/MRPC/test.tsv
  - config_name: QNLI
    data_files:
      - split: train
        path: datasets/QNLI/train.tsv
      - split: dev
        path: datasets/QNLI/dev.tsv
      - split: test
        path: datasets/QNLI/test.tsv
  - config_name: QNLI_v2
    data_files:
      - split: train
        path: datasets/QNLI_v2/train.tsv
      - split: dev
        path: datasets/QNLI_v2/dev.tsv
      - split: test
        path: datasets/QNLI_v2/test.tsv
  - config_name: QQP_v2
    data_files:
      - split: train
        path: datasets/QQP_v2/train.tsv
      - split: dev
        path: datasets/QQP_v2/dev.tsv
      - split: test
        path: datasets/QQP_v2/test.tsv
  - config_name: RTE
    data_files:
      - split: train
        path: datasets/RTE/train.tsv
      - split: dev
        path: datasets/RTE/dev.tsv
      - split: test
        path: datasets/RTE/test.tsv
  - config_name: SciTail
    data_files:
      - split: train
        path: datasets/SciTail/train.tsv
      - split: dev
        path: datasets/SciTail/dev.tsv
      - split: test
        path: datasets/SciTail/test.tsv
  - config_name: SST-2
    data_files:
      - split: train
        path: datasets/SST-2/train.tsv
      - split: dev
        path: datasets/SST-2/dev.tsv
      - split: test
        path: datasets/SST-2/test.tsv
  - config_name: STS-B
    data_files:
      - split: train
        path: datasets/STS-B/train.tsv
      - split: dev
        path: datasets/STS-B/dev.tsv
      - split: test
        path: datasets/STS-B/test.tsv
  - config_name: WNLI
    data_files:
      - split: train
        path: datasets/WNLI/train.tsv
      - split: dev
        path: datasets/WNLI/dev.tsv
      - split: test
        path: datasets/WNLI/test.tsv
  - config_name: SNLI
    data_files:
      - split: train_raw
        path: datasets/SNLI/train_raw.tsv
      - split: train
        path: datasets/SNLI/train.tsv
      - split: dev
        path: datasets/SNLI/dev.tsv
      - split: test
        path: datasets/SNLI/test.tsv
  - config_name: MNLI
    data_files:
      - split: test_matched
        path: datasets/MNLI/test_matched.tsv
      - split: test_mismatched
        path: datasets/MNLI/test_mismatched.tsv
      - split: dev_matched
        path: datasets/MNLI/dev_matched.tsv
      - split: dev_mismatched
        path: datasets/MNLI/dev_mismatched.tsv
      - split: train
        path: datasets/MNLI/train.tsv

PLUE

• Repository: https://github.com/ju-resplande/PLUE
• Paper:
• Leaderboard:
• Point of Contact:

Portuguese translation of the GLUE benchmark, SNLI, and Scitail using OPUS-MT model and Google Cloud Translation.

The language data in PLUE is Brazilian Portuguese (BCP-47 pt-BR)

Citation Information

@misc{Gomes2020,
  author = {GOMES, J. R. S.},
  title = {PLUE: Portuguese Language Understanding Evaluation},
  year = {2020},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/jubs12/PLUE}},
  commit = {CURRENT_COMMIT}
}

Contributions

Thanks to @ju-resplande for adding this dataset.