Projects
全部
LLM
Audio
Image
Video
Multimodal
Benchmark
开源
Image Generation: Janus-4o
Janus-4o is a multimodal model that can generate high-quality images from text or from text and images, trained on just 91K samples made by GPT-4o.
开源
Medical LLM: HuatuoGPT-II
HuatuoGPT-II uses a unified input-output format for domain adaptation and achieves top performance in Medical, even surpassing GPT-4 in some tasks.