High-Performance Training by Exploiting Hot-Embeddings in Recommendation Systems

By DeepAI - 2021-03-01

Description

03/01/21 - Recommendation models are commonly used learning models that suggest relevant items to a user for e-commerce and online advertisem...

Summary

Recommendation models are commonly used learning models that suggest relevant items to a user for e-commerce and online advertisement-based applications.
Current recommendation models include deep-learning-based (DLRM) and time-based sequence (TBSM) models.
This paper tries to leverage skewed embedded table accesses to efficiently use the GPU resources during training.
This framework efficiently estimates and varies the size of the hot portions of the embedding tables within GPUs and reallocates the rest of the embeddings on the CPU.

Topics

Machine_Learning (0.15)
NLP (0.13)
UX (0.07)

Similar Articles

Introducing ABENA: BERT Natural Language Processing for Twi

By Medium - 2020-10-23

Transformer Language Modeling for Akuapem and Asante Twi

ZeRO-Offload: Training Multi-Billion Parameter Models on a Single GPU

By Medium - 2021-01-27

Deep learning (DL) model size is growing exponentially — from the fewer than 100 million parameters in 2017’s largest language model to the whopping 175 billion parameters in 2020’s GPT-3. Training…

Reducing the High Cost of Training NLP Models With SRU

By ASAPP - 2021-02-24

Highly expressive and efficient neural models can be designed using SRU++ with little attention computation needed.

Key Concepts for Deploying Machine Learning Models to Mobile

By Spell - 2020-12-02

In this guest article from Fritz AI, we'll be outlining the key considerations in building a mobile-focused ML pipeline from end-to-end.

Fit More and Train Faster With ZeRO via DeepSpeed and FairScale

By huggingface - 2021-01-20

We’re on a journey to solve and democratize artificial intelligence through natural language.

FastFormers: 233x Faster Transformers inference on CPU

By Medium - 2020-11-04

Since the birth of BERT followed by that of Transformers have dominated NLP in nearly every language-related tasks whether it is Question-Answering, Sentiment Analysis, Text classification or Text…