FlashSloth: Lightning Multimodal Large Language Models via Embedded Visual Compression

Benchmark Model Rank Results
visual-question-answering-on-mm-vetFlashSloth-HD#49GPT-4 score: 49.0Params: 3.2B
visual-question-answering-on-mm-vetFlashSloth#68GPT-4 score: 41.9Params: 3.2B