Cloud ML interface for R
Switch branches/tags
bufix/keras-cache-creation bugfig/gs-sync-local-path bugfix/allow-account-change bugfix/appveyor-enable bugfix/appveyor bugfix/cloud-binary-windows bugfix/cloudml-terminal-path bugfix/config-files bugfix/config-imports-appveyor bugfix/cran-warnings bugfix/default-bucket bugfix/default-region-uscentral bugfix/default-region bugfix/deploy-using-cran-packages bugfix/fix-hypertuning bugfix/fix-keras-cache-warning bugfix/gcloud-binary-path bugfix/gcloud-exec-echo bugfix/gcloud-init-windows bugfix/gcloud-install-osx bugfix/gcloud-install-path bugfix/gsutil-windows bugfix/hyperparam-examples-maxtrials bugfix/hyperparam-usability bugfix/import-tfruns bugfix/improve-job-trials-collect bugfix/job-trials-categorical bugfix/job-trials-numeric bugfix/keras-cache-creation bugfix/learning-phase-remove bugfix/max-params bugfix/predict-print bugfix/predict-savedmodel-multiple bugfix/prediction-response-flat bugfix/print-python3 bugfix/read-runtimeversion bugfix/remove-gcloud-add-region bugfix/remove-remote-tfruns bugfix/revert-devel-deps bugfix/summary-writer-example bugfix/tensorflow-1.9 bugfix/terminal-windows-type bugfix/testthat-fix-cran bugfix/todos bugfix/train-alternate-cloudml bugfix/training-time bugfix/travis-fix-debian-setup bugfix/travis-not-cran bugfix/travis-python-envs bugfix/travis-python-versions bugfix/trials-no-folders bugfix/tuning-accuracy-metric bugfix/tuning-flags-job-dir bugfix/validate-entrypoint bugfix/verbose-doc bugfix/view-run-regression bugfix/windows-exec-warnings bugfix/windows-git-bash bugfix/windows-processx-crash bugfix/windows-system-path bugfix/windows-terminals-check feature/cloudml-command-line feature/cloudml-deploy feature/cloudml-predict-verbose feature/config-driven-training feature/configuration-filter feature/copy-directory-recursive feature/cran-devtools-only feature/deploy-mnist-vignette feature/deploy-models-dir feature/deployment-vignette feature/docs-training feature/docs feature/gcloud-install feature/getting-started-vignette feature/gloud-all-jobs feature/hyperparameters feature/install-github-shim feature/job-collect-multiple feature/keras feature/latest-r feature/machine-types-config feature/machine-types feature/master-type-docs feature/overlay-refactor feature/pip-cache feature/publish-docs feature/python-cache feature/runtime-version-in-cloudml feature/sdk feature/simplify-shell-arguments feature/tests feature/train-improvements feature/training-flags feature/travis-pythin-versions feature/tuning feature/zero-config master
Nothing to show
Clone or download

README.md

R interface to Google CloudML

Build Status AppVeyor Build Status CRAN_Status_Badge

The cloudml package provides an R interface to Google Cloud Machine Learning Engine, a managed service that enables:

  • Scalable training of models built with the keras, tfestimators, and tensorflow R packages.

  • On-demand access to training on GPUs, including the new Tesla P100 GPUs from NVIDIA®.

  • Hyperparameter tuning to optimize key attributes of model architectures in order to maximize predictive accuracy.

  • Deployment of trained models to the Google global prediction platform that can support thousands of users and TBs of data.

CloudML is a managed service where you pay only for the hardware resources that you use. Prices vary depending on configuration (e.g. CPU vs. GPU vs. multiple GPUs). See https://cloud.google.com/ml-engine/pricing for additional details.

For documentation on using the R interface to CloudML see the package website at https://tensorflow.rstudio.com/tools/cloudml/