Thursday, December 30, 2021

MirroredStrategy demo for distributed training


Google Cloud Developer Advocate Nikita Namjoshi demonstrates how to get started with distributed training on Google Cloud. Learn how to distribute training across multiple GPUs within a managed Jupyter Lab environment. Intro to distributed training → https://goo.gle/3FolcIz Creating and managing GCP projects → https://goo.gle/3mDZ4m7 Vertex AI → https://goo.gle/3FzfmUU Cloud Storage Quickstart → https://goo.gle/3qsNWcN Notebook Executor Codelab → https://goo.gle/3etrvPh Chapters: 00:00 - Introduction 00:40 - Get started with GCP 01:00 - Walking through the notebook’s code 02:46 - Creating a notebook on GCP 05:35 - Modifying the code for distributed training 06:59 - Running the distributed training job 08:41 - Another way to run distributed training 11:18 - The results 11:54 - Thanks for watching! Watch more ML Tech Talks → https://goo.gle/ml-tech-talks Subscribe to TensorFlow → https://goo.gle/TensorFlow #TensorFlow #MachineLearning #ML product: TensorFlow - General;

No comments:

Post a Comment