alibaba/AliceMind
Collection of pre-trained encoder-decoder models and multimodal large language models developed by Alibaba's MinD Lab.

AliceMind is Alibaba’s comprehensive library of pre-trained language and multimodal models. It includes large language models (mPLUG-Owl, ChatPLUG), multimodal models supporting text, image, and video (mPLUG-2, mPLUG-DocOwl), vision-language understanding systems, and Chinese NLP models. The repository provides model weights, training techniques, and benchmarks for research on encoder-decoder architectures, multimodal learning, and dialogue systems.