sielo

ds = tfds.load('huggingface:scielo/en-es')

A parallel corpus of full-text scientific articles collected from Scielo database in the following languages: English, Portuguese and Spanish. The corpus is sentence aligned for all language pairs, as well as trilingual aligned for a small subset of sentences. Alignment was carried out using the Hunalign algorithm.

{
    "translation": {
        "languages": [
            "en",
            "es"
        ],
        "id": null,
        "_type": "Translation"
    }
}

ds = tfds.load('huggingface:scielo/en-pt')

A parallel corpus of full-text scientific articles collected from Scielo database in the following languages: English, Portuguese and Spanish. The corpus is sentence aligned for all language pairs, as well as trilingual aligned for a small subset of sentences. Alignment was carried out using the Hunalign algorithm.

{
    "translation": {
        "languages": [
            "en",
            "pt"
        ],
        "id": null,
        "_type": "Translation"
    }
}

ds = tfds.load('huggingface:scielo/en-pt-es')

A parallel corpus of full-text scientific articles collected from Scielo database in the following languages: English, Portuguese and Spanish. The corpus is sentence aligned for all language pairs, as well as trilingual aligned for a small subset of sentences. Alignment was carried out using the Hunalign algorithm.

{
    "translation": {
        "languages": [
            "en",
            "pt",
            "es"
        ],
        "id": null,
        "_type": "Translation"
    }
}