Text-to-Image Generation on GenEval (GenEval Score and R Metric)
87GenEval ScoreBase model
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Base modelModel Context=Qwen-Image, P.(B)↓=20, M.(%)↓=100, L.(ms)↓=26252025.11 | 87 | 0 | |
| PPCL(14B)Model Context=Qwen-Image, P.(B)↓=14, M.(%)↓=79.4, L.(ms)↓=17922025.11 | 84.7 | 0.42 | |
| PPCL(Fine-tuning)Model Context=Qwen-Image, P.(B)↓=10, M.(%)↓=66.9, L.(ms)↓=14622025.11 | 82.8 | 3.29 | |
| PPCL(12B)Model Context=Qwen-Image, P.(B)↓=12, M.(%)↓=71.4, L.(ms)↓=15142025.11 | 80.1 | 3.03 | |
| PPCL(10B)Model Context=Qwen-Image, P.(B)↓=10, M.(%)↓=66.9, L.(ms)↓=14622025.11 | 78.4 | 6.88 | |
| HierarchicalPruneModel Context=Qwen-Image, P.(B)↓=14, M.(%)↓=79.4, L.(ms)↓=17862025.11 | 76.6 | 6.49 | |
| TinyFusionModel Context=Qwen-Image, P.(B)↓=14, M.(%)↓=79.4, L.(ms)↓=17892025.11 | 73.9 | 8.75 | |
| Base modelModel Context=FLUX.1-dev, P.(B)↓=12, M.(%)↓=100, L.(ms)↓=7152025.11 | 66.5 | 0 | |
| FLUX.1 LiteModel Context=FLUX.1-dev, P.(B)↓=8, M.(%)↓=78.8, L.(ms)↓=5722025.11 | 62.3 | 6.09 | |
| Base modelModel Context=FLUX.1 Lite, P.(B)↓=8, M.(%)↓=78.8, L.(ms)↓=5722025.11 | 62.3 | 0 | |
| PPCL(8B)Model Context=FLUX.1-dev, P.(B)↓=8, M.(%)↓=74.4, L.(ms)↓=5352025.11 | 60.5 | 4.03 | |
| Chroma1-HDModel Context=FLUX.1-dev, P.(B)↓=8.9, M.(%)↓=82.5, L.(ms)↓=17142025.11 | 59.3 | 1.02 | |
| PPCL(6.5B)Model Context=FLUX.1 Lite, P.(B)↓=6.5, M.(%)↓=69.2, L.(ms)↓=4282025.11 | 59.3 | 0.07 | |
| TinyFusionModel Context=FLUX.1-dev, P.(B)↓=8, M.(%)↓=74.4, L.(ms)↓=5342025.11 | 51.1 | 13.8 | |
| HierarchicalPruneModel Context=FLUX.1-dev, P.(B)↓=8, M.(%)↓=74.4, L.(ms)↓=5432025.11 | 50.3 | 13.38 | |
| Dense2MoEModel Context=FLUX.1-dev, P.(B)↓=12, M.(%)↓=100, L.(ms)↓=3122025.11 | 40.3 | 21.52 |