ReachQA on ReachQA (test)
98.2Execution RateMSRL
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| MSRLParams=7B, Setting=Zero-shot2025.08 | 98.2 | 86.1 | 89.9 | |
| GPT-4oParams=-, Setting=Zero-shot2025.08 | 92.8 | 81.8 | 84 | |
| MSRL-SFTParams=7B, Setting=Zero-shot2025.08 | 92.2 | 78.6 | 80 | |
| Claude-3-opusParams=-, Setting=Zero-shot2025.08 | 89 | 51.7 | 61.1 | |
| GPT-4VParams=-, Setting=Zero-shot2025.08 | 88 | 69.5 | 78.6 | |
| ChartCoderParams=7B, Setting=Zero-shot2025.08 | 83.8 | 67.9 | 69.4 | |
| Qwen2-VL-72BParams=72B, Setting=Zero-shot2025.08 | 77.2 | 50 | 48.1 | |
| GeminiProVisionParams=-, Setting=Zero-shot2025.08 | 74 | 67 | 67.8 | |
| Chart2CodeParams=7B, Setting=Zero-shot2025.08 | 63.6 | 52.3 | 49.7 | |
| Qwen2.5-VL-7BParams=7B, Setting=Zero-shot2025.08 | 62.2 | 36.9 | 37.6 | |
| Qwen2-VL-7BParams=7B, Setting=Zero-shot2025.08 | 55.4 | 22.6 | 29.3 | |
| InternVL2-26BParams=26B, Setting=Zero-shot2025.08 | 55.4 | 29 | 28.8 | |
| ChartLlamaParams=13B, Setting=Zero-shot2025.08 | 54.8 | 11.1 | 8.1 | |
| InternVL2-8BParams=8B, Setting=Zero-shot2025.08 | 50.8 | 24.1 | 24.2 | |
| TinyChartParams=3B, Setting=Zero-shot2025.08 | 34.4 | 11.6 | 11.2 | |
| ChartVLM-LParams=14B, Setting=Zero-shot2025.08 | 8.2 | 2.1 | 3.9 |