Skip to main content

Open model registry

Download MobTranslate language models

Every public model gets a version, model card, dataset record, rights note, and evaluation report. Training-ready entries appear here before artifacts are published so the release path stays visible.

Rights visible

Release rows carry corpus rights status and keep training consent separate from display rights.

Versioned artifacts

Model, adapter, manifest, and eval files are tied to one immutable version.

Dataset trace

Each model points back to the exact corpus export used for training and evaluation.

Kuku Yalanji Translation

gvn

A versioned English -> Kuku Yalanji translation model line trained from approved MobTranslate parallel corpora.

NLLB LoRAFar North Queensland, Australia

Version 0.1.0-mini-pilot

internal proof
Base model
facebook/nllb-200-distilled-600M
Dataset
kuku_yalanji_ebible_parallel_v0.1.0
Directions
eng-gvn
Release date
2026-06-30
Test in v2

Kuku Yalanji eBible snapshot rights granted for MobTranslate model training by project owner attestation on 2026-06-30. Proof artifacts are local/internal and are not a public release.

train loss
6.295
validation loss
5.773
validation bleu
0.0313
validation chrf
4.744
test loss
5.817
test bleu
0.0275
test chrf
4.272
standalone bleu
0.0269
ArtifactDownload
Merged model

Local proof artifact. Publish only after community/project release decision.

Local
LoRA adapter

Includes resized language-token embeddings, so the proof adapter is large.

Local
Training manifest
Local
Evaluation report
Local
  • Budget-safe proof run on the high-confidence corpus, not a production translator.
  • A40 utilization was healthy: 100% max GPU, 26.5 GiB max VRAM, and 295 W max power draw.
  • Use this release to test the translate/v2 model-lab path and inference server wiring.

Version 0.1.0-smoke

internal proof
Base model
facebook/nllb-200-distilled-600M
Dataset
kuku_yalanji_ebible_parallel_smoke_v0.1.0
Directions
eng-gvn
Release date
2026-06-30
Test in v2

Kuku Yalanji eBible snapshot rights granted for MobTranslate model training by project owner attestation on 2026-06-30. Smoke artifacts are only a pipeline proof.

train loss
6.537
validation loss
6.013
validation bleu
0.1085
validation chrf
4.913
test loss
6.416
test bleu
0.0540
test chrf
5.485
standalone bleu
0.0552
ArtifactDownload
Merged model

Local smoke artifact.

Local
LoRA adapter
Local
Training manifest
Local
Evaluation report
Local
  • First end-to-end RunPod proof: dataset upload, CUDA validation, LoRA training, merge, eval, artifact pullback.
  • Not useful as a translator; kept for regression testing the model project.

Version 0.1.0-baseline

training ready
Base model
facebook/nllb-200-distilled-600M
Dataset
kuku_yalanji_ebible_parallel_v0.1.0
Directions
eng-gvn
Release date
2026-06-30
Awaiting training run

Kuku Yalanji eBible snapshot rights granted for MobTranslate model training by project owner attestation on 2026-06-30. Base model license and final release terms still apply to trained artifacts.

ArtifactDownload
Model card
Get
Merged model

Published after the full baseline passes evaluation and review.

Pending
LoRA adapter

Published alongside the merged model.

Pending
Training manifest
Pending
Evaluation report
Pending
  • Full baseline is intentionally blocked until the model project, download registry, and translate/v2 test bench are working.
  • Planned first publishable candidate: high-confidence rows only, English to Kuku Yalanji, 8 epochs, A40/RTX A6000 first.

Registry updated 30 June 2026.