Vibeleaderboard
← All Intel
Intel / article

Introducing ERNIE-Image

Source
unknown
Date

Context

ERNIE-Image adds a prompt-expansion stage to text-to-image generation. Baidu's model card describes a Prompt Enhancer that turns brief requests into richer, structured descriptions before image generation. Its example pipeline lets users enable that component explicitly.

The design targets work where composition and written content matter, including posters and multi-panel images. The advertised eight-billion-parameter figure refers to the diffusion that generates the image, rather than an accounting of every component in the pipeline. The model weights are available under Apache 2.0, allowing developers to work with the released system directly.

Terms in this piece · Glossary
  • transformer — The neural network architecture behind modern AI models, built on attention — letting every word directly consider every other word in parallel.
  • open weights — A model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.
Source link unavailable.unknown
Recommended reads
Comments

Checking sign-in…

Loading comments…