Niels Bantilan
Niels Bantilan
This is a stub issue for keeping track of different tutorial ideas to showcase unionml functionality, see here: https://unionml.readthedocs.io/en/latest/tutorials_intro.html
Use case: As a data scientist, I want to define a custom function hook after the `model.predictor` function has been called so that I can track metrics and aggregates about...
- extend the integration test suite to include a xgboost-based app: https://github.com/unionai-oss/unionml/tree/main/tests/integration - update the documentation index.md to include a new `xgboost` tab: https://github.com/unionai-oss/unionml/blob/main/docs/source/index.md
- extend the integration test suite to include a tensorflow-based app: https://github.com/unionai-oss/unionml/tree/main/tests/integration - update the documentation index.md to include a new `tensorflow` tab: https://github.com/unionai-oss/unionml/blob/main/docs/source/index.md
Consider a `model.callbacks.*` API for defining additional behavior that’s meant to occur at some point during the training/prediction. The development should be use-case driven, e.g. ```python @model.trainer(pre_callback=..., post_callback=...) def trainer(...):...
Similar to the [model.predict](https://unionml.readthedocs.io/en/latest/generated_api_reference/unionml.model.Model.html#unionml.model.Model.predict) method, document the functions that are invoked inside the `model.train` method.
https://www.revela.io/
https://docs.konan.ai/
Add instructions on how to install unionml using conda to https://unionml.readthedocs.io/en/latest/contributing_guide.html and README.md
Pull out unionml task and workflows and use them in Flyte workflows