Neural Machine Translation
NLLB-200 distilled-600M · 200+ languages · PyTorch 2.x
Source language
Target language
Examples
| Input text | Source language | Target language |
|---|
Record or upload audio. It will be transcribed, then translated into the target language.
Speech language
Target language
Enter one sentence per line — up to 50 lines per request.
Source language
Target language
Example
| Input sentences | Source language | Target language |
|---|
Compare a reference translation against model output using the BLEU metric.
60–100 Excellent
40–60 Good
20–40 Fair
0–20 Poor
Examples
| Reference translation | Model translation |
|---|
MODEL — FACEBOOK/NLLB-200-DISTILLED-600M · FRAMEWORK — PYTORCH 2.X + TRANSFORMERS · UI — GRADIO