I have read the datalake 3.0 blog on tensorflow. Please provide suggestions for customers currently on HDP 2.5.3 cluster.
HDP cluster is installed on RHEL 7.3 , how about running tensorflow in docker on HDP edge node ?
Does anyone have experience running tensorflow on spark ?
If you perform a search on tensorflow in the HCC Search, there are a number of excellent articles, such as:
plus articles on DZone:
@Graham Martin Does the same steps work for CPU servers ?
Tensorflow on Spark from Yahoo is very solid.
I have run it as a regular install on Hadoop nodes and NiFi nodes and it works fine.
I also like the paradigm of running separate GPU heavy nodes with TensorFlow Serving and communicating via gRPC. Also TensorFlow can save data to HDFS