Achieve up to 4.3x the BERT AI Deep Learning Performance with Google Cloud C3 High-CPU Virtual Machines with 4th Gen Intel® Xeon® Scalable Processors

Description

Organizations use natural language processing (NLP) frameworks for tasks such as analyzing textual data, making predictions, answering questions, and responding to conversations. Bidirectional Encoder Representations from Transformers (BERT) workloads are a popular NLP framework for analyzing and getting AI insights from text data. When running BERT workloads on Google Cloud VMs, it is important to choose carefully to get the best performance.