model compression 썸네일형 리스트형 [ICML 2025] Q-VDiT: Towards Accurate Quantization and Distillation of Video-Generation Diffusion Transformers https://arxiv.org/abs/2505.22167 ICML 2025 Accepted Paper Q-VDiT: Towards Accurate Quantization and Distillation of Video-Generation Diffusion TransformersDiffusion transformers (DiT) have demonstrated exceptional performance in video generation. However, their large number of parameters and high computational complexity limit their deployment on edge devices. Quantization can reduce storage r.. 더보기 [ACM SAC 2025] Advanced Knowledge Transfer: Refined Feature Distillation for Zero-Shot Quantization in Edge Computing https://dl.acm.org/doi/abs/10.1145/3672608.3707816Abstract기존 Zero-Shot Quantization(ZSQ, Data-Free Quantizaton) 분야에서는 full-precision(FP) Model로부터 높은 quality의 데이터를 생성하는 데 초점을 두는 연구가 진행되고 있음. 하지만, low-bit(높은 압축률) 환경에서 Quantized Model을 학습할 때는 Quantized Model이 정보 수용량 관련 한계를 갖기 때문에 데이터를 생성하는 기법만으로는 적절한 학습이 이루어지지 않음. 이러한 한계를 개선하기 위해 본 논문에서는 Quantized Model을 효과적으로 학습하기 위한 AKT(Advanced Knowledge Transfe.. 더보기 [Neurocomputing 2025] A lightweight video anomaly detection model with weak supervision and adaptive instance selection https://www.sciencedirect.com/science/article/pii/S0925231224014693 Neurocomputing (IF: 5.5, Q1) A lightweight video anomaly detection model with weak supervision and adaptive instance selectionVideo anomaly detection is to determine whether there are any abnormal events, behaviors or objects in a given video, which enables effective and inte…www.sciencedirect.com AbstractWeakly supervised vid.. 더보기 [ICML2024] KIVI: A Tuning-Free Asymmetric 2bit Quantization for KV Cache Liu, Zirui, et al. "Kivi: A tuning-free asymmetric 2bit quantization for kv cache." ICML 2024 (Poster) https://icml.cc/virtual/2024/poster/34318 ICML Poster KIVI: A Tuning-Free Asymmetric 2bit Quantization for KV CacheAbstract: Efficiently serving large language models (LLMs) requires batching many requests together to reduce the cost per request. Yet, the key-value (KV) cache, which stores atte.. 더보기 [AAAI2021] Cross-Layer Distillation with Semantic Calibration Chen, Defang, et al. "Cross-layer distillation with semantic calibration." Proceedings of the AAAI conference on artificial intelligence. Vol. 35. No. 8. 2021. (AAAI 21) https://ojs.aaai.org/index.php/AAAI/article/view/16865 AbstractFeature map을 기반으로 지식을 전이하는 기존 feature distillation은 student model을 효과적으로 training시키는 방식임. 하지만, 의미론적 정보(semantic information)는 다양한 layer에 분포하며 이는 부정적인 regualrization.. 더보기 이전 1 다음
