add setup.py, workflows, unit tests, and examples #1
Conversation
Signed-off-by: kta-intel <[email protected]>
… some basic tests Signed-off-by: kta-intel <[email protected]>
Signed-off-by: kta-intel <[email protected]>
Signed-off-by: kta-intel <[email protected]>
Signed-off-by: kta-intel <[email protected]>
Signed-off-by: kta-intel <[email protected]>
Signed-off-by: kta-intel <[email protected]>
Signed-off-by: kta-intel <[email protected]>
Signed-off-by: kta-intel <[email protected]>
…example Signed-off-by: kta-intel <[email protected]>
… and workflows Signed-off-by: kta-intel <[email protected]>
…e test Signed-off-by: kta-intel <[email protected]>
There was a problem hiding this comment.
Thanks for submitting PR no. 1, @kta-intel 😉 ! It's great that you have created some basic infrastructure and examples for contributors to start building and testing with.
I have a couple of suggestions for making the repo structure a bit more intuitive - but otherwise I believe it's in pretty good shape already!
|
|
||
| setup( | ||
| name='openfl-contrib', | ||
| version='1.6', |
There was a problem hiding this comment.
openfl-contrib being a newly minted repo, shouldn't we start from 0.1.0 as per semantic versioning?
| @@ -0,0 +1,3 @@ | |||
| # Copyright (C) 2020-2023 Intel Corporation | |||
There was a problem hiding this comment.
nit - 2024 (noticed a couple of similar occurrences across the PR)
| # Copyright (C) 2020-2024 Intel Corporation | ||
| # SPDX-License-Identifier: Apache-2.0 | ||
| """openfl-contrib version information.""" | ||
| __version__ = '1.6' |
There was a problem hiding this comment.
same here - 0.1.0 maybe for openfl-contrib specifically?
| # Copyright (C) 2020-2023 Intel Corporation | ||
| # SPDX-License-Identifier: Apache-2.0 | ||
|
|
||
| """Aggregation functions package.""" |
There was a problem hiding this comment.
I'd suggest to keep the aggregation_functions package right under openfl_contrib. I think the intermediate interface package is a bit confusing. Essentially, the openfl_contrib package could look like this, additionally highlighting the types of contrib-s (pun intended 😅) suitable for this repo:
└── openfl-contrib
└── aggregation_functions
└── (compression_)pipelines
└── ...
|
|
||
| __all__ = [ | ||
| 'CustomWeightedAverage' | ||
| ] |
There was a problem hiding this comment.
nit - the indentation looks a bit off here
| @@ -0,0 +1,3 @@ | |||
| # Copyright (C) 2020-2023 Intel Corporation | |||
| # SPDX-License-Identifier: Apache-2.0 | |||
| """Tests package.""" | |||
There was a problem hiding this comment.
I remember seeing a similar package structure somewhere under openfl, but I don't understand very well why we need a separate github package (supposedly for tests running in the CI?). I suggest to simplify, and move the the test_task_runner.py module right under tests/openfl_contrib.
This PR adds:
setup.pyto enable installation ofopenfl-contribextension on top of latest baseopenflopenfl_contribfolder structure to mirroropenfl, containing:openfl_contrib.interface.aggregation_functionswith a custom weighted aggregation function (custom implementation of OpenFL's weighted average)openfl_contrib.pipelineswith an SKC compression pipeline (borrowed directly from base OpenFL)openfl_contrib_workspaceto demonstrate how to incorporate custom plan with new agg functions and compression pipelines. Will also serve as the workspace for Task Runner test.github/workflowsto run lint and teststestsdirectory containing:githubfor CI testing of Task Runner API with custom plan andCustomWeightedAverageandSKCPipelineopenfl-contribfor individual unit tests forCustomWeightedAverageandSKCPipeline