AIConfPaper
← Search

Pegah Kharazmi

2 accepted papers

2023

Distill-Quantize-Tune - Leveraging Large Teachers for Low-Footprint Efficient Multilingual NLU on Edge

Pegah Kharazmi, Zhewei Zhao, Clement Chung, Samridhi Choudhary
ICASSP 2023accepted

This paper describes Distill-Quantize-Tune (DQT), a pipeline to create viable small-footprint multilingual models that can perform NLU on extremely resource-constrained Edge devices. We distill semantic knowledge from a large-sized teacher (transformer-based), that has been trained on huge amount of…

Cited by 0Source ↗Scholar ↗
2023

Pyramid Dynamic Inference: Encouraging Faster Inference Via Early Exit Boosting

Ershad Banijamali, Pegah Kharazmi, Sepehr Eghbali, Jixuan Wang, Clement Chung, Samridhi Choudhary
ICASSP 2023accepted

Transformer-based models demonstrate state of the art results on several natural language understanding tasks. However, their deployment comes at the cost of increased footprint and inference latency, limiting their adoption to real-time applications. Early exit strategies are designed to speed-up t…

Cited by 0Source ↗Scholar ↗