Repository navigation
Remote worker configuration #12
Description
Activity
This release does not support distributed computation.
Hi kmatzen: As mentioned in our FAQ, this release does not have an RPC and distributed implementation available yet: http://tensorflow.org/resources/faq.md#running_a_tensorflow_computation, but we are working on it and hope to make it available when it's ready.
As a short-term proxy, you can see the Cifar tutorial for an example of how training and distributing among multiple GPUs works -- http://tensorflow.org/tutorials/deep_cnn/index.md
- added a commit that references this issue
on Mar 30, 2017 - added a commit that references this issue
on Jun 20, 2017 - added a commit that references this issue
on Aug 2, 2018 - added a commit that references this issue
on Dec 8, 2018 - added a commit that references this issue
on Jan 23, 2020 - added a commit that references this issue
on Aug 28, 2020 - added a commit that references this issue
on Sep 9, 2020 10 remaining items
- added 7 commits that reference this issue
on Apr 9, 2025 - added 5 commits that reference this issue
on Jul 28, 2025 - added a commit that references this issue
on Nov 26, 2025
Does this release support remote workers out of the box? I couldn't find details in the tutorials, but the whitepaper and
tensorflow/core/common_runtime/device.htalked about adding remote devices through RPC. Is a parameter server or alternative already implemented in here somewhere? If remote workers are supported out of the box, what is the execution model? Is there a daemon or do you have to configure a session in a certain way to act as one?