A suit of multilingual MoE models with highly-sparse architectures
-
AIDC-AI/Marco-Nano-Base
Text Generation β’ 8B β’ Updated β’ 581 β’ 13 -
AIDC-AI/Marco-Mini-Base
Text Generation β’ 17B β’ Updated β’ 654 β’ 6 -
AIDC-AI/Marco-Mini-Global-Base
Text Generation β’ 17B β’ Updated β’ 525 β’ 6 -
AIDC-AI/Marco-Nano-Instruct
Text Generation β’ 8B β’ Updated β’ 2.33k β’ 33