Model Family: Mistral and Mixtral

6.29.1Mistral 7B and Sliding-Window Attention#

6.29.2Mixtral 8×7B and Sparse MoE#

6.29.3Mixtral 8×22B and Model Scaling#

6.29.4Mistral's Instruction, Code, and Multimodal Branches#

6.29.5Routing, Caching, and Deployment Characteristics of Mistral/Mixtral#