Machine Translation on WMT23 En-XX Average (test)
33.9BLEUX-ALMA-13B (Only SFT, MoE)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| X-ALMA-13B (Only SFT, MoE)Parameters=13B, Training Strategy=Only SFT, Architecture=MoE2025.02 | 33.9 | 85.6 | |
| Aya-23-35BParameters=35B2025.02 | 29.3 | 84.1 | |
| X-ALMA-13B-DAT (MoE)Parameters=13B, Training Strategy=DAT, Architecture=MoE2025.02 | 29.1 | 84.8 | |
| X-ALMA-13B-DATM (MoE)Parameters=13B, Training Strategy=DATM, Architecture=MoE2025.02 | 29.1 | 84.8 | |
| Aya-23-8BParameters=8B2025.02 | 28.3 | 84.2 | |
| NLLB-3.3BParameters=3.3B2025.02 | 27.3 | 81.5 | |
| LLaMAX3-Alpaca-8BParameters=8B2025.02 | 22 | 79.8 | |
| Aya-1012025.02 | 21.3 | 80.8 |