Ordicio on-device translator
Small translation models for the 16 languages of the Ordicio apps, which translate attribute values and data sheets of collectors' items on the device. There is one model per direction between English and each language; other pairs go through English. A language detector of 2.2 MB (0.75 MB to download) tells which model to use.
| Languages | en, de, fr, es, it, nl, pt (Brazil), pl, sv, da, nb, fi, ja, ko, zh-Hans, zh-Hant |
| Models | 30 directions (en↔x), transformer encoder + SSRU decoder, int8 (bergamot-translator intgemm8 with alphas) |
| Download | about 23 MB per direction, 13 MB for sv, da, nb and 34 MB for ja, ko, zh (gzip compressed); the detector 0.75 MB |
| Speed | about 18 ms per value in the WebAssembly engine on one core |
| Base | the Firefox Translations students of each direction |
| Training | fine-tuned on collecting texts translated by Gemma 4 31B, plus a quantization-aware epoch |
Files
manifest.json lists every file with its sha256 and size (of the stored, gzip-compressed file) and the
unpacked size:
{
"id": "ordicio-translator-v1",
"languages": ["en", "de", "..."],
"pivot": "en",
"engine": {"beam-size": "1", "gemm-precision": "int8shiftAlphaAll", "...": "..."},
"pairs": {
"en-de": {"from": "en", "to": "de",
"model": {"path": "v1/en-de/model.bin.gz", "sha256": "…", "size": 0, "gzip": true, "unpackedSize": 0},
"srcVocab": {"path": "v1/en-de/vocab.spm.gz", "...": "..."}, "trgVocab": {"...": "..."}}
},
"lid": {"model": {"path": "v1/lid.bin.gz", "...": "..."}, "labels": ["en", "...", "zxx"], "buckets": 131072,
"ngrams": [1, 4], "scale": 0.0, "bias": [], "maxChars": 1000, "hash": "fnv1a32-utf8"}
}
- Paths are relative to the manifest.
srcVocabandtrgVocabname the same file when the direction has one vocabulary. engineholds the bergamot-translator options the models were evaluated with. They run without a lexical shortlist (it cost quality on short values).source/v1/<src>-<tgt>.npzare the float Marian weights of each model (the form to change them in); the apps do not download them.
Language detector
A linear model over hashed character n-grams with 17 labels: the 16 languages and zxx (names, codes, model
numbers: nothing to translate). lid.bin is int8, buckets × labels (the weights of one bucket together).
- Text → lower case → every digit
0-9to0→ runs of whitespace (space, tab, line feed, carriage return, form feed, vertical tab, U+00A0, U+1680, U+2000–U+200A, U+2028, U+2029, U+202F, U+205F, U+3000) to one space → spaces trimmed → the firstmaxCharscode points. No Unicode normalization. - Each word
w(split on spaces) is padded to" " + w + " "; every n-gram of 1 to 4 code points of the padded word except a single space counts once. - The bucket of an n-gram is the 32-bit FNV-1a hash of its UTF-8 bytes modulo
buckets. The feature vector holds the count per bucket, L2 normalized. - Score of label
i=bias[i] + scale × Σ feature × weight[bucket][i]; softmax gives probabilities.
| Accuracy (16 languages) | this detector | Lingua | fastText lid.176 |
|---|---|---|---|
| short values (held out, 6,877) | 0.89 | 0.83 | 0.74 |
| sentences (WMT / NTREX held out, 2,649) | 1.00 | 0.99 | 0.98 |
Short values are often ambiguous (Danish and Norwegian, Spanish and Portuguese, Chinese characters in Japanese). On them, 78 % of the predictions reach a probability of 0.9, and 98.7 % of those are right.
Quality
chrF++ in the bergamot WebAssembly engine without shortlist, against the shipped Firefox model of the same direction. Test: 150 English values of a collecting catalog (examples, list values, member notes) with reference translations.
| en → | de | fr | es | pl | fi | ja | ko | zh-Hans |
|---|---|---|---|---|---|---|---|---|
| Firefox | 69.1 | 67.5 | 67.8 | 60.0 | 60.7 | 32.5 | 42.5 | 25.1 |
| this | 70.7 | 70.9 | 69.2 | 64.4 | 65.1 | 40.6 | 51.0 | 36.3 |
Values that every reference keeps unchanged (names, titles, codes) stay unchanged twice as often (9 to 16 of 19 instead of 1 to 11). Korean writes most of them in Hangul, as Korean catalogs do (4 of 19 unchanged).
| → en (the references as source) | de | fr | es | pl | fi | ja | ko | zh-Hans |
|---|---|---|---|---|---|---|---|---|
| Firefox | 75.0 | 74.4 | 73.6 | 69.3 | 66.0 | 58.2 | 57.9 | 55.0 |
| this | 76.8 | 74.7 | 75.4 | 69.4 | 67.4 | 61.1 | 61.9 | 59.0 |
it, nl, pt, sv, da, nb and zh-Hant have no reference set. On 300 held-out teacher translations they gain 4 to 32 chrF++ over the Firefox models from English and 1 to 9 into English.
Training
- 10,900 English collecting texts (attribute values of a classification catalog, member notes) translated by Gemma 4 31B (4-bit) into the 15 languages.
- Each Firefox student (float weights from Mozilla's published training runs) fine-tuned on these pairs (both directions; for x → en the teacher's translation is the source) mixed with WMT and NTREX-128 sentences: 4 epochs, learning rate 2e-5, then one quantization-aware epoch (8 bit), so the int8 model scores like the float one.
- For Korean, the teacher prompt asks for brands, people, places and characters in their usual Hangul form (as Korean catalogs write them), keeps inscriptions, codes and scientific species names unchanged, and leaves noun phrases without a sentence ending. The Japanese → English model learns from Japanese translations made with a Japanese prompt in the same spirit (names in katakana, inscriptions and codes unchanged, notes in the plain form).
License
The models are derived from the Firefox Translations models and are released under the Mozilla Public License 2.0 like them. The teacher, Gemma 4, is licensed under Apache 2.0. Fine-tuning data also includes NTREX-128 (CC BY-SA 4.0) and WMT test sets.