DeepRec | BestHub

Ximalaya Technology Team

Oct 9, 2023 · Artificial Intelligence

DeepRec-Based High-Dimensional Sparse Feature Support and Real-Time Model Training in Ximalaya AI Cloud

Ximalaya AI Cloud leverages DeepRec’s Embedding Variable to elastically manage high‑dimensional sparse features with low collision, supporting admission/eviction, multi‑level storage and minute‑level incremental model updates, which together boost GPU utilization, halve training time and improve recommendation CTR by 2‑3 % while maintaining latency.

AI CloudDeepRecRecommendation systems

0 likes · 13 min read

DeepRec-Based High-Dimensional Sparse Feature Support and Real-Time Model Training in Ximalaya AI Cloud

DataFunTalk

Apr 17, 2022 · Artificial Intelligence

DeepRec: Alibaba’s Sparse Model Training Engine – Architecture, Features, and Open‑Source Status

DeepRec, developed since 2016 by Alibaba, is a specialized sparse‑model training engine that addresses feature elasticity, training performance, and deployment challenges through dynamic elastic features, optimized runtimes, distributed training frameworks, incremental model export, and multi‑level storage, and is now being open‑sourced for broader industry collaboration.

AI InfrastructureDeepRecdistributed training

0 likes · 15 min read

DeepRec: Alibaba’s Sparse Model Training Engine – Architecture, Features, and Open‑Source Status