Skip to content
This repository has been archived by the owner on Nov 8, 2018. It is now read-only.

How dist-keras achieve parallelism ? #61

Open
anishsharma opened this issue Apr 11, 2018 · 0 comments
Open

How dist-keras achieve parallelism ? #61

anishsharma opened this issue Apr 11, 2018 · 0 comments

Comments

@anishsharma
Copy link

Hi,

I know their are spark functions such as map,mapPartitions which executes tasks in different nodes but I am interested to know about the basic fundamental concepts.how dist-keras go about training and predicting in achieving parallelism ? In other words, does training and predictions of batches happens in different nodes simultaneously ? If so then how do they combine the final results ?

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant