Releases: NxNiki/brain_decoding
twilight!
working version!
This is the first working version with sleep data as test set.
Add plots on predictions.
Note: config has a parameter to use test_mode which will only read 10000 predictions.
What's Changed
- Dev-Add sleep data to inference dataset by @NxNiki in #5
- check how to lost whole sleep data. by @NxNiki in #11
Full Changelog: v0.0.4-beta...v0.0.5
Bug fix on config for test and validation
In the previous update, due to limited epochs (5) and large validation_step (25), validation and test is never run and activation plots are not created. This release fix this and debug config.
organize all parameters in one config data model.
This update organizes all parameters (i.e model, data, and experiment) into a single data model (build with pydantic).
Reduce duplicated code to set parameters and use a single script to set parameters for different models and experiments. Save config as a yaml file which is loaded in main.py
first working version
Add a Pydantic data model for patients' information.
Fix some bugs running on sample data.
What's Changed
New Contributors
Full Changelog: v0.0.1-beta...v0.0.2-beta
backup original code
backup yiyuang's original code with minor changes:
reorganize directories
add peotry to setup environments
add pre-commit with isort, black to format codes.