The Living Thing / Notebooks :

Serious number crunching on Google Cloud

I want to do cloud machine learning using google’s CloudML offering.

Google cloud might interoperate well with a bunch of google products, such as tensorflow. I use that. Let’s see how it works.

Notes

Obviously this will work better if you commit to googly storage APIs etc.

Note (2017-02-14): Google supports python 2.7 only - so be prepared to party like it’s 2010. (Is this still so?)

Seems to also support VMs?

There is the usual thicket of weird service names.

Basic workflow bits are

Also good to know:

Painstaking run through

See How is Google number crunching awful?.

Google now has their own instruction manual for one common use case, tensorflow. Tensorflow without a phd.

Refs

Just saw a gentleman called @ankutkotwal claiming that Google cloudml does some neat AutoML. MMMV. Recommend hyptertuner for parameter selection?