Machine Translation (ca-eo) on FLORES+
52.7ChrF++NLLB-200-3.3B
Evaluation Results
| Method | Links | |
|---|---|---|
| NLLB-200-3.3BModel Category=Neural MT, Parameters=3.3B2026.03 | 52.7 | |
| NLLB-200-distilled-600MModel Category=Neural MT, Parameters=600M, Strategy=distilled2026.03 | 52.26 | |
| NLLB-200-1.3BModel Category=Neural MT, Parameters=1.3B2026.03 | 51.97 | |
| NLLB-200-distilled-1.3BModel Category=Neural MT, Parameters=1.3B, Strategy=distilled2026.03 | 51.52 | |
| Llama-3.1-8B-Instruct-FTModel Category=Fine-tuned General-purpose LLMs, Parameters=8B2026.03 | 50.35 | |
| Llama-3.1-8B-InstructModel Category=General-purpose LLMs, Parameters=8B2026.03 | 49.59 | |
| Transformer-tinyModel Category=Neural MT from Scratch, Parameters=17.4M2026.03 | 49.57 | |
| Transformer-baseModel Category=Neural MT from Scratch, Parameters=60.6M2026.03 | 48.73 | |
| ApertiumModel Category=Rule-based MT2026.03 | 45.44 | |
| Tower-Plus-9BModel Category=MT-tuned LLMs, Parameters=9B2026.03 | 42.99 | |
| TowerInstruct-7B-v0.2Model Category=MT-tuned LLMs, Parameters=7B2026.03 | 23.97 |