GCP supports training scikit / xgboost/ Tensorflow models and exporting them to Cloud Storage for use elsewhere. More here: https://cloud.google.com/ml-engine

GCP supports training scikit / xgboost/ Tensorflow models and exporting them to Cloud Storage for use elsewhere. More here: https://cloud.google.com/ml-engine/docs/scikit/custom-pipeline https://cloud.google.com/ml-engine/docs/scikit/custom-pipeli... https://cloud.google.com/ml-engine/docs/algorithms/xgboost-start https://cloud.google.com/ml-engine/docs/algorithms/xgboost-s... https://cloud.google.com/ml-engine/docs/tensorflow/getting-started-keras https://cloud.google.com/ml-engine/docs/tensorflow/getting-s... Creating custom notebook containers is now also supported with AI Platform: https://cloud.google.com/ai-platform/notebooks/docs/custom-container https://cloud.google.com/ai-platform/notebooks/docs/custom-c... Disclaimer: I work for Google.