spark-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "xiaobo " <guxiaobo1...@qq.com>
Subject What's relationship between the TensorflowOnSpark core modules?
Date Fri, 23 Feb 2018 08:43:19 GMT
Hi,
After reading the mnist example and the API of TensorflowOnSpark, I somehow got confused,
here are some questions:
1、 What's the relationship between TFCluster/TFManager/TFNode and TFSparkNode modules.
2、The conversion guide says we should replace the main function with a main_fun, but the
examples actually defines a map function, are they the same thing?
3、map_function parameter question:
      when the map_function is called, argv is the sys.args, what about the ctx parameter,
what's the type of ctx, and how its value got assigned,I can see the ctx object has various
properties and even functions .
4、The conversion guide says there should be step 3 and step 4, but the mnist example does
not do these things
5、By the way, where can I post questions about TFoS, I can  not join the official google
group.


I am sorry if I post this in a wrong place.
Mime
View raw message