Logo image
Access Frequency-Aware Storage Reduction for Deep Learning Recommendation Model
會議論文

Access Frequency-Aware Storage Reduction for Deep Learning Recommendation Model

Chia-Chun Wang, Chuan-Yao Lai 和 Ren-Shuo Liu
Proceedings - IEEE International Conference on Computer Design, 頁碼.710-717
IEEE
2025 IEEE 43rd International Conference on Computer Design (ICCD) (Richardson, TX, USA, 10/11/2025–12/11/2025)
10/11/2025
Web of Science ID: WOS:001684940200100

摘要

Accuracy Computational modeling Data models Deep learning Embedding Guidelines Memory management Pruning Quantization Quantization (signal) Recommendation System Recommender systems Training Stress
Personalized recommendation is one of the flourishing AI applications in recent years. However, its powerful functionality comes with the need of a significant amount of embeddings, which hinders the model performance by causing large memory footprints. In addition, each inference request requires multiple memory access to these embeddings, which puts even more stress on the already-suffering memory bandwidth. In this paper, we aim to reduce the storage size of the embeddings required by personalized recommendation while inducing minimal impact on the accuracy. To achieve such goals, we take advantage of the fact that not all data contribute equally towards the final result. Small values and less frequently accessed embeddings have little impact on the accuracy. In addition, we opt to avoid any extra steps of training to restore the accuracy since training recommendation models can take hours, consuming a huge amount of computing power, which is undesirable. Instead, to restore accuracy, we further identify the insight that some data, although less frequently accessed, do not offer good storage reduction-accuracy trade-offs. Following these key guidelines, we judiciously choose certain embeddings to apply element-wise pruning, leaving the rest untouched. We are able to prune the embedding data by more than 99% while inducing less than 1% accuracy drop.

相關連結

指標

1 檢視次數

詳細資料

Logo image