Machine Translation on WMT23 En-Ru (test)
31.5BLEU ScoreX-ALMA-13B (Only SFT, MoE)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| X-ALMA-13B (Only SFT, MoE)Parameters=13B, Training Strategy=Only SFT, Architecture=MoE2025.02 | 31.5 | 85.9 | |
| NLLB-3.3BParameters=3.3B2025.02 | 29.1 | 83.8 | |
| X-ALMA-13B-DAT (MoE)Parameters=13B, Training Strategy=DAT, Architecture=MoE2025.02 | 27.6 | 85 | |
| X-ALMA-13B-DATM (MoE)Parameters=13B, Training Strategy=DATM, Architecture=MoE2025.02 | 27.6 | 85 | |
| Aya-23-35BParameters=35B2025.02 | 27.5 | 84.7 | |
| Aya-23-8BParameters=8B2025.02 | 24.3 | 84.3 | |
| LLaMAX3-Alpaca-8BParameters=8B2025.02 | 23.5 | 81.6 | |
| Aya-1012025.02 | 22.1 | 83.1 |