metadata
license: mit
dataset_info:
features:
- name: sub-domain
dtype: string
- name: target_text
dtype: string
- name: date
dtype: string
- name: language
dtype: string
- name: source_text
dtype: string
- name: unique id
dtype: int64
- name: hereafter_tagged
dtype: string
- name: reordering_tagged
list: string
- name: ctxpro_tagged
list: string
- name: muda_tagged
dtype: string
splits:
- name: train
num_bytes: 1342965218
num_examples: 112861
- name: test
num_bytes: 168675404
num_examples: 10000
download_size: 613384554
dataset_size: 1511640622
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
Dataset Card
Flipper was created as part of an academic research project led by Dragon LLM and the Laboratoire d'Informatique de Grenoble, group GETALP.
It is built from publicly available documents that are specialised in finance, containing mainly financial regulatory texts, that underwent a long processing pipeline. This dataset allows the training and evaluation of document-level machine translation in the domain of finance.
For more information about this dataset, please refer to the accompanying paper: [paper published soon].