Text-to-Image Generation on FaithT2I (test)
74.8VQA Score (Gemini)FaithRewriter
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| FaithRewriterBackbone=Qwen3-8B-Base2026.06 | 74.8 | 73.71 | 78.91 | 4.075 | 4.098 | 4.392 | 1.52 | 1.59 | 1.55 | 84.7 | 22.19 | |
| FaithRewriterBackbone=Hunyuan-7B-Instruct2026.06 | 70.83 | 73.41 | 77.94 | 3.949 | 4.001 | 4.164 | 1.84 | 1.95 | 1.94 | 80.8 | 22.87 | |
| RePromptBackbone=Qwen3-8B-Base2026.06 | 69.15 | 68.91 | 76.12 | 3.194 | 3.857 | 3.891 | 2.23 | 1.97 | 1.89 | 82.1 | 22.58 | |
| Text modalBackbone=Qwen3-8B-Base2026.06 | 68.78 | 69.4 | 75.65 | 3.263 | 3.814 | 3.963 | 2.19 | 2.03 | 1.82 | 82.8 | 22.63 | |
| PromptEnhancerBackbone=Hunyuan-7B-Instruct2026.06 | 68.59 | 72.69 | 74.43 | 3.472 | 3.491 | 3.483 | 2.24 | 1.89 | 2.28 | 81.3 | 22.84 | |
| Text modalBackbone=Hunyuan-7B-Instruct2026.06 | 66.37 | 68.15 | 67.06 | 3.176 | 3.863 | 3.925 | 2.81 | 2.77 | 2.71 | 79.6 | 23.25 | |
| ZeroshotBackbone=Hunyuan-7B-Instruct2026.06 | 64.18 | 65.82 | 62.94 | 3.327 | 3.364 | 3.381 | 3.11 | 3.39 | 3.07 | 76.8 | 23.67 | |
| FaithRewriterBackbone=LLaMa3.1-8B2026.06 | 63.27 | 65.61 | 70.65 | 4.037 | 4.249 | 4.403 | 1.71 | 1.83 | 1.68 | 78.4 | 23.41 | |
| T0 (raw)Backbone=None2026.06 | 62.12 | 60.28 | 64.39 | 3.303 | 3.939 | 3.994 | — | — | — | 75.2 | 23.94 | |
| ZeroshotBackbone=Qwen3-8B-Base2026.06 | 59.73 | 61.34 | 70.83 | 4.122 | 3.745 | 3.773 | 2.29 | 2.38 | 2.63 | 76.2 | 23.76 | |
| Text modalBackbone=LLaMa3.1-8B2026.06 | 59.12 | 62.35 | 64.52 | 3.229 | 3.689 | 3.977 | 2.09 | 1.95 | 2.07 | 77.4 | 23.58 | |
| ZeroshotBackbone=LLaMa3.1-8B2026.06 | 55 | 58.23 | 60.32 | 3.314 | 3.421 | 3.456 | 2.2 | 2.22 | 2.25 | 76.2 | 23.76 |