Image2Code on Real-world images corpus Flowchart D3
0.86CEr ScoreLlama3.2-11B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Llama3.2-11Bstatus=Vanilla2026.01 | 0.86 | 0.25 | 0.21 | 0.18 | |
| Qwen2.5-VL-7B2026.01 | 0.33 | 0.74 | 0.7 | 0.46 | |
| MiniCPM-V-2-62026.01 | 0.31 | 0.69 | 0.44 | 0.11 | |
| GPT-4o-mini2026.01 | 0.04 | 0.95 | 0.88 | 0.85 | |
| Llama-VL-TUGstatus=Fine-tuned, backbone=Llama3.2-11B-Vision-Instruct2026.01 | 0 | 0.86 | 0.63 | 0.29 | |
| Gemma3-12B2026.01 | 0 | 0.97 | 0.88 | 0.81 |