谷歌云平台和谷歌机器学习

Question

I have to use several services of the google cloud platform but I'm pretty confused between the several services (Google Machine learning engine, google Data prep, Data lab).

How do they interact together ? And I have a more specific question : I ran a python script (to use SVM classifier) in the cloud shell ? So am I using the the google machine learning engine by doing so ?

If I ran another script python using the tensorflow library am I using google machine learning engine ?

And the only advantage of using google machine learning engine is the google machine learning library ? Because tensorflow, scikitlearn , etc can be used with other python interpreters ... Thank you very much in advance for your answers.

Answer 1

When you create a project on the Google Cloud Platform (GCP), you can configure the project to access different services. Many of these services, such as Cloud Storage, Datastore, BigTable, and Dataprep, involve storing and transforming data at high speed.

Another service, the Google Compute Engine (GCE), makes it possible to execute code on Google's powerful CPUs and GPUs. When you run an application using Datalab or the cloud shell, the GCP configures the required processor(s) and deploys your code.

The Machine Learning engine runs training and prediction jobs on the GCE's CPUs and GPUs. One advantage of using the engine is that you can configure a job to execute on a cluster of processors. From what Google says, you can also access custom processors called Tensor Processing Units (TPUs). You have to specifically request permission to use TPUs, and last I checked, Google won't give that permission to everybody.