Qwen2.5-VL released in 3 sizes (3B, 7B, 72B) for Hugging Face and ModelScope, with base and instruct models; includes visual understanding capabilities
Read the original at qwenlm.github.io→QWEN CHAT GITHUB HUGGING FACE MODELSCOPE DISCORD We release Qwen2.5-VL, the new flagship vision-language model of Qwen and also a significant leap from the previous Qwen2-VL. To try the latest model, feel free to...
Original headline: "Qwen2.5 VL! Qwen2.5 VL! Qwen2.5 VL!"