---
dataset_info:
  features:
  - name: text1
    dtype: string
  - name: text2
    dtype: string
  - name: label
    dtype: float64
  - name: source
    dtype: string
  splits:
  - name: train
    num_bytes: 293947097
    num_examples: 241957
  - name: test
    num_bytes: 50716064
    num_examples: 39359
  download_size: 58828058
  dataset_size: 344663161
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
task_categories:
- text-classification
language:
- fr
size_categories:
- 1M<n<10M
---

This is a dataset for training a mixed cross-encoder. The purpose of the cross-encoder is to calculate not only a relevance score between a question and a context (whether the answer to the question can be found in the document or not) but also to calculate a similarity score between two sentences. This dataset is a combination of the PIAF, FQuAD, SQuAD-French, pandora-s-fr, and stsd-fr datasets.