List your company

Visual Question Answering (Perception + Reasoning) on MathVista, MMStar, and MMVet

46.9MathVista ScoreLATTE
30.46834.7343943.266Dec 7, 2024

Evaluation results

MethodLinks
LATTELanguage / Vision=Qwen2-7B/SigLIP, Starting checkpoint=LLaVA-OV Stage 1.52024.1246.950.850.951.2
Vanilla ITLanguage / Vision=Qwen2-7B/SigLIP, Starting checkpoint=LLaVA-OV Stage 1.52024.1242.450.139.343.9
LATTELanguage / Vision=LLaMA3-8B/SigLIP, Starting checkpoint=Mantis Instruct-tuned2024.1238.945.15044.7
LATTELanguage / Vision=LLaMA3-8B/CLIP, Starting checkpoint=Mantis Pretrained2024.1236.640.845.240.8
Vanilla ITLanguage / Vision=LLaMA3-8B/SigLIP, Starting checkpoint=Mantis Instruct-tuned2024.1236.240.729.735.5
LATTELanguage / Vision=LLaMA3-8B/CLIP, Starting checkpoint=Mantis Pretrained (Configuration 2)2024.1234.944.645.241.6
Vanilla ITLanguage / Vision=LLaMA3-8B/CLIP, Starting checkpoint=Mantis Pretrained2024.1233.136.728.932.9
Vanilla ITLanguage / Vision=LLaMA3-8B/CLIP, Starting checkpoint=Mantis Pretrained (Configuration 2)2024.1231.140.53334.9