Description2Code Generation on Synthetic Eval Dataset Packet 1.0
1BLEUGPT-4o-mini
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GPT-4o-mini2026.01 | 1 | 100 | 1 | 100 | 0.8295 | 1 | |
| Llama-VL-TUG2026.01 | 0.958 | 95.8048 | 0.9902 | 97.0555 | 0.8043 | 0.9803 | |
| Gemma3-12B-Instruction-TunedModel Size=12B, Variant=Instruction-Tuned2026.01 | 0.1243 | 12.7378 | 0.3533 | 45.9207 | -0.3478 | 0.4548 | |
| MiniCPM-V-2-62026.01 | 0.0556 | 5.8855 | 0.2378 | 34.0349 | -0.5503 | 0.3326 | |
| Qwen2.5-VL-7B-InstructModel Size=7B, Variant=Instruct2026.01 | 0.0544 | 5.7997 | 0.2202 | 27.7765 | -0.8272 | 0.2582 | |
| Llama3.2-11B-Vision-InstructModel Size=11B, Variant=Vision-Instruct2026.01 | 0.0235 | 3.6337 | 0.1563 | 27.1367 | -0.8113 | 0.2517 |