Machine Translation on FLORES-22 1.0 (Evaluation)
0.1946BLEUFinetuned (alternate with and w/o punct) (x)
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Finetuned (alternate with and w/o punct) (x)Training Mode=alternate punctuation, Data Scale=1x2025.12 | 0.1946 | 0.7825 | 0.7533 | 55.11 | 50.9 | 0.9218 | 0.7421 | |
| Finetuned (with and w/o punct) (2x)Training Data=with and without punctuation, Data Scale=2x2025.12 | 0.1937 | 0.781 | 0.7521 | 55.11 | 50.9 | 0.9204 | 0.7418 | |
| Fine-tuned t5-base + Original ModelRestoration Backbone=t5-base, Translation Backbone=IndicTrans22025.12 | 0.1934 | 0.7818 | 0.7531 | 55.02 | 50.82 | 0.9213 | 0.7413 | |
| Finetuned (w/o punct) (x)Training Data=without punctuation, Data Scale=1x2025.12 | 0.1927 | 0.7795 | 0.7528 | 54.73 | 50.57 | 0.9201 | 0.7413 | |
| Finetuned (w/ punct) (x)Training Data=with punctuation, Data Scale=1x2025.12 | 0.192 | 0.7786 | 0.7513 | 54.78 | 50.57 | 0.9198 | 0.74 | |
| IndicTrans2 en indic 200 M (Original)Model Parameters=200 M2025.12 | 0.1869 | 0.7894 | 0.7616 | 54.48 | 50.19 | 0.9226 | 0.7425 |