Microsoft reveals cutting-edge MAI-Base-2 model with 2T parameters

SiliconANGLE
Read the full article
Microsoft has developed MAI-Base-2, a 2 trillion parameter base language model using mixture of experts architecture, which is up to 12 times more hardware-efficient than leading open-source models. It was trained on 50 trillion tokens including vision data, enhancing its scientific chart understanding.

More in LLM & Text Generation

Meta releases Muse Spark Open, a new ‘open-weight’ AI model

Covered by 14 sources

Scaling Indic Pretraining: 2.8 Trillion Tokens and an Open Data Generation Pipeline

Covered by 2 sources

Welcome Beam: a 501B open model you can own, customize, and scale

Covered by 2 sources