Image model

CogView

CogView is THUDM’s text to image system that uses a transformer over discrete image tokens. It is strong with Chinese and English prompts, can render readable Chinese text, and supports image generation, captioning, and simple edits in later versions.

Type
Image
Weights
Closed