metadata

language:
  - fra
  - eng
multilinguality:
  - multilingual
task_categories:
  - translation
configs:
  - config_name: defaut
    data_files:
      - split: train
        path: train.csv
    sep: ;

Dataset origin: https://www.kaggle.com/datasets/adewoleakorede/english-french-translation

Dataset

I used this dataset for my project on translating from English to French using the transformer architecture.

To load the CSV file with pandas, use the parameter encoding_errors='ignore' - I couldn't fix the issues with the encoding.