Releases: icecube/dnn_reco
Releases · icecube/dnn_reco
Version 1.0.2
Release Notes
- Support for tensorflow version 2.16.1
- Udate setup and CI
- Setting of random seeds
- API changes of dependent libraries:
ruamel.yaml
Added Functionality:
- Added option to limit value range of direction vector to stabilize training
Commits and PRs
Full Changelog: v1.0.1...v1.0.2
Version 1.0.1
Release Notes:
- Minor bug fixes and clean up
Added Functionality:
- Option to define keys to load in default label loader module
Changes:
- Supress NaturalNameWarning from tables package
- make iteration order over internal dicts deterministic
- Cleaned up old model configs
Bug Fixes:
- add
__del__()
method to properly terminate multiprocessing jobs in DataHandler
Version 1.0.0
Release Notes:
- DNN reco now supports TF2
- TF1 still works in compatibility mode
- Documentation also updated to python3 and TF2
Added Functionality:
- Support for learning rate schedulers
- option to only apply input data filter on specific PDG particle types
Changes:
- Progress bars via tqdm for create trafo model and counting
- more detailed output when NaNs are found in input data
- add
is_neutrino
andis_numu
labels toevent_selection_labels.upgoing_tracks
Bug Fixes:
- add dependencies
tables
,uncertainties
,tqdm
- fix git origin url for version control
- pickle protocol fixed to 2: ensures models can be used in python 2 and 3
- Azimuth and Zenith of I3Particle are now transformed to proper range in I3Module
- fixed python3 multiprocessing bug for data input pipeline
Version 0.2.0
Changes:
- Reduced memory footprint
- DOM exclusions are now handled directly in DNN I3Module
- Additional configuration checks when applying an exported DNN model
Added Functionality:
- Define event weights in loss function
- Label smoothing for fuzzy labels
- GNN comparison architecture and config
- Option to load FilterMask in general misc data loader
- File weighting for files from different datasets
- Gradient clipping and NaN replacement in gradients
- DNN LLh classes to obtain directional error contours
- Biased sampling of training events based on misc data or NN prediction
Added Models and Configs:
- NuGen L2 event selection
- Various starting event selection models
- Muon classification
- Track reconstruction at L2 (track length, direction, ..)
- HESE Spice3.2 reconstruction (used for cascade real-time alert stream)
- Muon multiplicity
- Muon scattering event selection
- Starting event selection model based on Cascade-based input data
Bug Fixes:
- Fix naming of I3Modules: allows inclusion of multiple DNN reco segments
- Throw error when pulses are not in frame
- Stabilize loss and architectures
- Fix global time offset shift
- Fix python3 compatibility
- Automatically find packages in setup.py
Initial Release
Initial Release of DNN reco Project.