Add Operation Partitioning

[sangkeun00]

System information

• AutoDist version: 0.7.0
• Are you willing to contribute it (Yes/No): Yes

Describe the new feature and the current behavior/state
Enable operation partitioning as in Mesh Tensorflow to improve parallelization of very large neural networks.

Will this change the current API? How?
May add an additional (operation) partitioner to the kernel.

Describe alternatives you've considered

Additional context