---
pretty_name: Translations
license: other
license_name: per-upstream-corpus
task_categories:
  - translation
language:
  - en
  - ca
  - be
tags:
  - parallel-corpus
  - bitext
  - mtdata
  - machine-translation
size_categories:
  - 10M<n<100M
configs:
  - config_name: bel-eng
    data_files:
      - split: train
        path: data/bel-eng/*.parquet
  - config_name: cat-eng
    data_files:
      - split: train
        path: data/cat-eng/*.parquet
---

# natgillin/translations

Parquet bitext corpus assembled from upstream mtdata / OPUS releases.

## Schema

| column        | type   | description |
| ------------- | ------ | ----------- |
| `source`      | string | source-language sentence |
| `target`      | string | target-language sentence |
| `source_lang` | string | source language, ISO 639-3 (3-char), e.g. `bel`, `cat`, `eng` |
| `target_lang` | string | target language, ISO 639-3 (3-char) |
| `origin`      | string | upstream sub-corpus, e.g. `opus-ccmatrix`, `opus-nllb`, `elr`, `mtdata-merged` |

Source / target ordering is fixed by **alphabetical ISO 639-3 code**
(`bel` before `eng`, `cat` before `eng`). This keeps the pair name
deterministic and bidirectional joins unambiguous.

## Configs / layout

One config per language pair. The OPUS sub-corpus is encoded as a
row-level column (`origin`), not as a separate config.

```
data/
  bel-eng/
    opus-ccmatrix-00000-of-00200.parquet
    opus-nllb-00000-of-00200.parquet
    mtdata-merged-00000-of-00200.parquet
    ...
  cat-eng/
    ...
```

Each language pair is one config (`bel-eng`, `cat-eng`) with everything
under that pair pinned to `split: train`. The configs are listed
explicitly in the YAML frontmatter so that HF's auto-split detection
does not mistake filenames containing substrings like `-test-` for an
actual test split.

## `origin` values

| value             | meaning |
| ----------------- | ------- |
| `mtdata-merged`   | mtdata's final deduplicated merged bitext for the pair |
| `opus-<source>`   | one upstream OPUS sub-corpus, e.g. `opus-ccmatrix`, `opus-nllb`, `opus-ccaligned`, `opus-multiccaligned`, `opus-wikimatrix`, `opus-eubookshop`, `opus-kde4`, `opus-gnome`, `opus-neulab-tedtalks`, ... |
| `elr`             | ELRA / ELR slice of the pair when distinct from the merged top-level |

## License

Inherits per-corpus terms from the upstream OPUS sources. Cite OPUS:
<https://opus.nlpl.eu>.