Skip to main content

Quick Start

By default, the PE model is loaded by the diffusion model server, which may not provide optimal performance. For higher performance, the PE model can be deployed as a separate SGLang server. This document uses baidu/ERNIE-Image as an example. Run the model with the built-in Transformers PE implementation (default):
Run the model with an SGLang-served PE model (high performance):

Support matrix

Ascend NPU Environment

See Diffusion models with AR stage like GLM-Image.