Fetching the paper…

Training Recommender Systems at Scale: Communication-Efficient Model and Data Parallelism · Around