Add Flower Datasets tests as GitHub workflow #2345
Conversation
adam-narozniak
requested review from
danieljanes and
tanertopal
as code owners
.github/workflows/datasets.yml
Outdated
| on: | ||
| push: | ||
| branches: | ||
| - main | ||
| paths: | ||
| - "datasets/flwr_datasets/**" | ||
| pull_request: | ||
| branches: | ||
| - main | ||
| paths: | ||
| - "datasets/flwr_datasets/**" |
There was a problem hiding this comment.
| on: | |
| push: | |
| branches: | |
| - main | |
| paths: | |
| - "datasets/flwr_datasets/**" | |
| pull_request: | |
| branches: | |
| - main | |
| paths: | |
| - "datasets/flwr_datasets/**" | |
| on: | |
| push: | |
| branches: | |
| - main | |
| pull_request: | |
| branches: | |
| - main |
There was a problem hiding this comment.
@tanertopal Why did you remove the datasets/flwr_datasets/**. I thought to run the tests only when there was any change in the library itself
There was a problem hiding this comment.
Is their no cross depend code? Meaning if anything changes in Flower does that have any influence on datasets?
There was a problem hiding this comment.
On the Flower's side, there'll be examples that install the Flower Datasets library (the dependency is added to the pyproject.toml). They are already implemented - it's just a modification to the existing partition code of our current examples. It can be merged when the library is released.
On the Flower Datasets's side, I don't think we should add such tests (there will be tests that for direct ML frameworks integration).
To sum up, there should not be any case that the change of the code in one library (without a release) affects the other library.
…datasets-ci-tests
* 'main' of github.com:adap/flower: Fix default contiguous value in IidPartitioner (#2406) Update FDS docs index (#2337) Add TensorFlow integration tests with FDS (#2350) Add paths specification to CI triggers for FDS (#2399) Add FDS how-to guides (#2332) Add Flower Datasets tests as GitHub workflow (#2345) Fix the reference API documentation (#2397) Add FDS tutorial docs (#2375) Update tutorial-series-what-is-federated-learning.ipynb (#2396) Fix missing square in proximal term of FedProx baseline (#2210)
Run standard tests for Flower Dataset based on GitHub workflow.