AI2D厂商自报 · 越高越好
77.5%
HF model card table (AI2D_TEST via VLMEvalKit)
成绩出处BLINK厂商自报 · 越高越好
52.7%
HF model card table (BLINK via VLMEvalKit)
成绩出处ChartQA厂商自报 · 越高越好
80.8%
HF model card table (ChartQA Test via VLMEvalKit)
成绩出处CharXiv-D厂商自报 · 越高越好
60%
HF model card table (CharXiv DQ via VLMEvalKit)
成绩出处CountBench厂商自报 · 越高越好
0.73%
HF model card table (CountBench via VLMEvalKit)
成绩出处DocVQA厂商自报 · 越高越好
92.1%
HF model card table (DocVQA VAL via VLMEvalKit)
成绩出处Hallusion Bench厂商自报 · 越高越好
61.5%
HF model card table (HallusionBench via VLMEvalKit)
成绩出处IFEval厂商自报 · 越高越好
74.9%
HF model card table (IFEval via VLMEvalKit)
成绩出处InfoVQA厂商自报 · 越高越好
65.2%
HF model card table (InfoVQA VAL via VLMEvalKit)
成绩出处MMBench-V1.1厂商自报 · 越高越好
68.7%
HF model card table (MMBenchDEV_EN_V11 via VLMEvalKit)
成绩出处MMLU厂商自报 · 越高越好
50.4%
HF model card table (MMLU test via VLMEvalKit)
成绩出处MMLU-Pro厂商自报 · 越高越好
30.7%
HF model card table (MMLU-Pro test via VLMEvalKit)
成绩出处MMMU (val)厂商自报 · 越高越好
32.9%
HF model card table (MMMU DEV_VAL via VLMEvalKit)
成绩出处MMStar厂商自报 · 越高越好
51.8%
HF model card table (MMStar via VLMEvalKit)
成绩出处Multi-IF厂商自报 · 越高越好
37.3%
HF model card table (Multi-If via VLMEvalKit)
成绩出处OCRBench厂商自报 · 越高越好
79.2%
HF model card table (OCRBench via VLMEvalKit)
成绩出处OCRBench-V2 (en)厂商自报 · 越高越好
36.7%
HF model card table (OCRBench v2_en via VLMEvalKit)
成绩出处RealWorldQA厂商自报 · 越高越好
62.2%
HF model card table (RealWorldQA via VLMEvalKit)
成绩出处RefCOCO-avg厂商自报 · 越高越好
0.73%
HF model card table (RefCOCO avg P@1 over 8 splits via VLMEvalKit)
成绩出处
