1

I', looking to write a custom distribution strategy for tensorflow.

Currently there are several types strategies available (MirroredStrategy, TPUStrategy, ...), but i would like to implement a new way of distributing training across several machines.

I looked through the documentation and didn't find any tutorial or guide on how to implement a custom strategy

Is there already a predefined way to do this?

Joao P
  • 38
  • 1
  • 6
  • have you solved this problem and by how because I'm also preparing to implement the model parallelism strategy on Tensorflow. – Qiqin Zhan Apr 01 '20 at 03:59
  • I didn't manage to implement a new strategy. I implemented my distribution strategy outside the tensorflow enviroment. It is not ideal but it workerd. – Joao P Apr 01 '20 at 14:10

0 Answers0