Image2Code Generation on Synthetic Eval Dataset Packet
0.519BLEULlama-VL-TUG
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Llama-VL-TUG2026.01 | 0.519 | 51.8954 | 0.6775 | 68.7837 | 0.0244 | 0.5898 | |
| GPT-4o-mini2026.01 | 0.0525 | 5.7991 | 0.2362 | 38.8445 | -0.3666 | 0.3684 | |
| Qwen2.5-VL-7B-Instruct2026.01 | 0.045 | 4.8357 | 0.2034 | 29.4239 | -0.7568 | 0.2976 | |
| Llama3.2-11B-Vision-Instruct2026.01 | 0.0298 | 3.2621 | 0.1502 | 24.7333 | -0.8368 | 0.248 | |
| Gemma3-12B-Instruction-Tuned2026.01 | 0.0278 | 2.8232 | 0.0941 | 14.5225 | -1.283 | 0.1566 | |
| MiniCPM-V-2-62026.01 | 0 | 0.1241 | 0.0212 | 5.5846 | -1.2374 | 0.023 |