Looking for a practical example of making KFP app using torchx

Hi.

Where can I find a good practical example of building KubeFlow Pipeline app using torchx?

I’ve been through the documents for torchx, and now I get how torchx work.
However, I have no idea how to design my app with torchx.


I’ve already developed my pytorch app to support functions below:

  • Fetch data from Feature store
  • EDA for the fetched data
  • Preprocessing the fetched data
  • Training model
  • Validate model

And, the each function has its own CLI, such as

my_app load_data --feature_store_root [URI]  --local_storage_root [URI]...
my_app run_eda --local_storage_root [URI] ...
my_app train ...
...

Now, I want to write torchx components of the each function above.

Is it enough to put pipeline.py under the project root directory and write all the torchx things in it?