Ovis-1.6: An Open-Source Multimodal Large Language Model (MLLM) Architecture Designed to Structurally Align Visual and Textual Embeddings
Marktechpost
SEPTEMBER 29, 2024
Artificial intelligence (AI) is transforming rapidly, particularly in multimodal learning. Similarly, in the RealWorldQA benchmark, Ovis outperformed leading proprietary models such as GPT4V and Qwen-VL-Plus, scoring 2230, compared to GPT4V’s 2038. to 14.1%, depending on the specific benchmark.
111 
Let's personalize your content