giga_fren

ds = tfds.load('huggingface:giga_fren/en-fr')

Giga-word corpus for French-English from WMT2010 collected by Chris Callison-Burch
2 languages, total number of files: 452
total number of tokens: 1.43G
total number of sentence fragments: 47.55M

{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "translation": {
        "languages": [
            "en",
            "fr"
        ],
        "id": null,
        "_type": "Translation"
    }
}