Hi Griselda, Sorry for the late response. I wanted to use TF transform ( https://github.com/tensorflow/transform) for Preprocessing as part of Deep Learning training/prediction on a Spark cluster. Hence the interest in Spark runner support with Python.
Thanks, Manoj On Wed, Sep 20, 2017 at 12:34 AM, Griselda Cuevas <[email protected]> wrote: > Hi Manoj, > > As far as I know, today Python only works with DataFlow & direct runners > and there isn't a concrete timeframe for a Spark Runner at the moment. > > Could you share a bit more of your use case or why you're interested? > > G > > On 18 September 2017 at 06:36, Manoj Mohan <[email protected]> > wrote: > >> Hello, >> >> As per the response posted for this SO question in May, Spark runner >> support would be available in Python relatively soon. >> >> https://stackoverflow.com/questions/43923776/python-support- >> for-sparkrunner-in-apache-beam >> >> Still don't see it here: >> https://beam.apache.org/get-started/wordcount-example/ >> >> Can you please let me know if there is a timeframe for adding Spark >> Runner support for the Python SDK? >> >> Thanks, >> Manoj >> > >
