Learn how Tencent Deployed an Advertising System on the Merlin GPU Recommender Framework
, Tencent
, NVIDIA
The performance of recommendation models influences the income of internet companies. With massive data, the speed of the training system determines how much data you can train and the performance of the recommendation model. In the first half of this talk, we'll introduce Merlin, an end-to-end recommender-on-GPU framework that aims to provide fast feature engineering and high throughput of DL recommender models' training/inference. In the second half, as a case study we will introduce how Tencent deployed their real advertising recommendation training with Merlin and achieved more than 7x speedup over the original Tensorflow solution on the same GPU platform ... and income has increased in certain scenarios.