MM-Vet v2

ImagesTextsCC BY-NC 4.0Introduced 2024-08-01

MM-Vet v2: A Challenging Benchmark to Evaluate Large Multimodal Models for Integrated Capabilities

Benchmarks

Visual Question Answering/GPT-4 score Visual Question Answering/Params Visual Question Answering (VQA)/GPT-4 score Visual Question Answering (VQA)/Params