Baidu Releases ERNIE 5.5 Flash Base: A 280B MoE Base Model With 14B Active Parameters
MarkTechPost
Baidu has released ERNIE 5.5 Flash Base, a 280B sparse MoE base model.
Baidu has introduced the ERNIE 5.5 Flash Base model. The new offering is a 280 billion parameter sparse Mixture of Experts base model featuring 14 billion active parameters. It was trained on 33.5 trillion tokens and published before undergoing mid- or post-training.
Why it matters: The release provides the AI community with a new large-scale sparse Mixture of Experts base model trained on a vast corpus of tokens before post-training phases.