Janus-Pro-7B
Janus-Pro-7B is a flexible multimodal model excelling in text and image tasks, built on DeepSeek-LLM with SigLIP-L encoder.
What is Janus-Pro-7B?
Janus-Pro-7B is a powerful multimodal model that handles both text and image data efficiently. Built on the DeepSeek-LLM architecture and featuring SigLIP-L as its visual encoder, it supports 384x384 image inputs. Ideal for developers and researchers, it excels in tasks like image generation and text understanding, offering flexible multimodal interaction capabilities.