Search
Step AI Unveils Open-Source Base Model: Step 3.5 Flash
2 day ago /
Read about 0 minute
Author：小编
On February 2nd, Step AI introduced its latest open-source Agent base model, the Step3.5Flash. This model is specifically crafted for real-time Agent workflow scenarios and boasts an impressive maximum inference speed of up to 350 tokens per second. Leveraging a sparse Mixture-of-Experts (MoE) architecture, the model activates around 11 billion parameters for each token, with an overall parameter count of 96 billion. Several chip manufacturers, such as Huawei Ascend, MetaX Technologies, Biren Technology, Enflame Technology, Iluvatar CoreX, and Alibaba T-Head, have already finished adapting this model for their respective platforms.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
