Train TensorFlow models at cloud scale with TensorFlow Cloud
summary
Demo shows moving a NasNet training from Colab or Kaggle to Google Cloud by packaging the notebook and supplying three project parameters, authenticating, then calling tfc.run to launch remote jobs and inspect results in TensorBoard. It also runs hyperparameter search with Keras Tuner and tfc.run_cloudtuner, returns a link to live jobs, reports explored parameter trials and their scores, and confirms remote training and tuning complete successfully at cloud scale.
AI-written summary · about
related talks
37mEnd-2-End Machine Learning on GCP
22mThe Containerization of Machine Learning: TensorFlow: Kubernetes, and Kubeflow
42mDeploying ML at Scale with Kubernetes and TFX
13mTaking Machine Learning to the Next Level
04mDistributed Cloud for retail and manufacturing
37mWhich Serverless Prediction Platforms on Google Cloud Platform?
23mPyTorch on Google Cloud TPUs
32mGPU Machine Learning From Laptop to Cloud