# Device-side optimization

MobileLLM is Meta's series of sub-billion parameter language models specifically optimized for resource-constrained devices, significantly improving on-device inference efficiency through deep-narrow architecture design.

Zamba2-2.7B is a hybrid model composed of state space and Transformer modules, using the Mamba2 module and shared attention module, featuring high performance and low latency.

MobiLlama-1B-Chat is an instruction-following model fine-tuned from MobiLlama-1B, specifically designed for resource-constrained devices, emphasizing efficiency, low memory footprint, and fast response.

This is a small 7B-parameter LLM optimized for device-side use, pruned and trained with DPO

A speech classification model fine-tuned on the SUPERB keyword spotting task, based on the Wav2Vec2-Large-LV60 pre-trained model