Translated data: Adept Corporation has unveiled Fuyu-8B, a multi-modal large model with 8 billion parameters, capable of understanding various types of images. It employs a pure decoder Transformer architecture, eliminating the need for an image encoder and supporting any image resolution, excelling in multiple tasks. The company aims to create an AI Copilot and has secured $415 million in funding, demonstrating significant potential. Founded by industry professionals including Transformer creator Ashish Vaswani, Adept emphasizes open-source projects with the goal of enhancing human-computer collaboration efficiency.