Machine Translation on FLORES EN-ZH
90.08COMETkiwiSSR-X-Zero-7B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SSR-X-Zero-7BReasoning/Thinking Mode=true2025.05 | 90.08 | — | 93.98 | — | |
| Qwen3-32BReasoning/Thinking Mode=true2025.05 | 89.69 | — | 94.07 | — | |
| Claude-3.5-Sonnet2025.05 | 89.47 | — | 94.32 | — | |
| Gemini-1.5-Pro2025.05 | 89.3 | — | 94.54 | — | |
| SSR-Zero-7B2025.05 | 89.25 | — | 93.52 | — | |
| DeepTrans-7B2025.05 | 89.23 | — | 92.85 | — | |
| Qwen3-8BReasoning/Thinking Mode=true2025.05 | 88.9 | — | 93.3 | — | |
| GPT-4o2025.05 | 88.3 | — | 93.3 | — | |
| Qwen2.5-72B-Instruct2025.05 | 88.04 | — | 93.2 | — | |
| Gemma2-27B-it2025.05 | 87.79 | — | 92.51 | — | |
| GemmaX2-28-9B-v0.12025.05 | 87.58 | — | 92.83 | — | |
| Qwen3-32B2025.05 | 87.45 | — | 92.18 | — | |
| Qwen2.5-32B-Instruct2025.05 | 87.13 | — | 91.99 | — | |
| QwQ-32BReasoning/Thinking Mode=true2025.05 | 86.51 | — | 90.93 | — | |
| TowerInstruct-13B-v0.12025.05 | 86.22 | — | 90.69 | — | |
| Qwen3-8B2025.05 | 86.08 | — | 91.02 | — | |
| Gemma2-9B-it2025.05 | 85.95 | — | 90.9 | — | |
| SSR-Zero-3B2025.05 | 85.43 | — | 88.66 | — | |
| Qwen2.5-7B-Instruct2025.05 | 85.18 | — | 89.9 | — | |
| TowerInstruct-7B-v0.22025.05 | 84.39 | — | 88.57 | — | |
| Qwen2.5-7B2025.05 | 82 | — | 86.07 | — | |
| Qwen2.5-3B-Instruct2025.05 | 78.79 | — | 85.19 | — | |
| PEGRLBackbone=Qwen3-0.6B, RL training=PEGRL, Rollouts per example=722026.02 | 74.39 | 34.03 | 80.89 | — | |
| MT-R1-ZeroBackbone=Qwen3-0.6B, RL training=MT-R1-Zero, Rollouts per example=722026.02 | 73.78 | 33.24 | 79.83 | — | |
| Qwen3-0.6BTraining=Pre-trained only2026.02 | 70.54 | 30.25 | 77.18 | — | |
| Qwen2.5-3B2025.05 | 23.31 | — | 69.16 | — | |
| Llama_Bi_NTPmode=instruction fine-tuned2026.04 | — | 33.6 | — | 82.8 | |
| Llama_CLmode=instruction fine-tuned2026.04 | — | 36.2 | — | 83.2 | |
| Llama_CLImode=instruction fine-tuned2026.04 | — | 37.6 | — | 83.9 | |
| Llama_NTPmode=instruction fine-tuned2026.04 | — | 29.9 | — | 82.9 | |
| Llama-3-8Bmode=instruction fine-tuned2026.04 | — | 28.6 | — | 75.74 |