v0.2

What's Changed

• Adding dev_mode = false to fibad_default_config.toml by @drewoldag in #107
• Quick Bug Fixes in fibad_default_config.toml by @aritraghsh09 in #112
• Bump pre-commit-ci/lite-action from 1.0.3 to 1.1.0 by @dependabot in #110
• Initial version of filter_catalog feature. by @mtauraso in #113
• Updating documentation after demo session this morning. by @drewoldag in #117
• Hsc data set updates by @mtauraso in #120
• Rebuild a fits manifest from an HSC data directory. by @mtauraso in #115
• Removing dimensions print statement by @aritraghsh09 in #121
• TensorBoard and training/validation during training by @drewoldag in #119
• Allow definition of Loss and Optimizer in config file. by @drewoldag in #132
• Bump codecov/codecov-action from 4 to 5 by @dependabot in #122
• Bump dawidd6/action-download-artifact from 6 to 7 by @dependabot in #123
• Adding guards in HSCDataSet for the manifest representation of un-downloaded files by @mtauraso in #130
• Adding a simple HSC Autoencoder as an example by @aritraghsh09 in #136
• Breaking setup_model_and_dataset into two functions. by @drewoldag in #139
• Renaming verb predict to infer by @drewoldag in #141
• Fix assorted issues encountered while experimenting with vectordb by @drewoldag in #146
• Updating type hint by @drewoldag in #147
• Is having no validation split valid? by @mtauraso in #154
• User Controlled Caching by @aritraghsh09 in #155
• Type checking by @mtauraso in #153
• Default to using the whole dataset for inference unless overriden by @mtauraso in #156
• Adding credentials file support by @mtauraso in #157
• Holoview based prototype of FIBAD visualizations by @mtauraso in #149
• Retaining comments in persisted config files by @drewoldag in #148
• Cleaning up default config formatting, values and comments by @drewoldag in #162
• Optimizer and criterion subtables by @drewoldag in #163
• Adding the Capability for any numpy function to be passed as a transform variable by @aritraghsh09 in #164
• Fixing tensorboard instructions for running on a remote server by @aritraghsh09 in #166
• WIP Initial commit to get some things logged to mlflow. by @drewoldag in #170
• Using experiment_name from config to set mlflow experiment name. by @drewoldag in #174
• Prep for similarity search: adding chromadb to infer by @mtauraso in #159
• Adding scaffolding for documentation pages. by @drewoldag in #172
• Remove config.get, use in line if/else for "rebuild_manifest". by @drewoldag in #176
• Pass along the HSCDataset id for populating vectordb by @drewoldag in #178
• Being more explicit with the import of importlib.util. by @drewoldag in #181
• Add lookup verb and verbs-taking-arguments infrastructure by @mtauraso in #184
• Demo notebook prep by @drewoldag in #180
• Bump dawidd6/action-download-artifact from 7 to 8 by @dependabot in #192
• Umap verb by @mtauraso in #190
• Update simple training example notebook by @drewoldag in #196
• Add default of None to umap verb by @drewoldag in #198
• Protect against 0-length validation splits in cifar dataset by @drewoldag in #199
• Fix github pre-commit mypy and import checks by @drewoldag in #201
• Adding a UMAP section to the notebook demo. by @drewoldag in #202
• Configuration help in notebook by @drewoldag in #194
• Visualize verb implementing Holoviews scatter plot and table by @mtauraso in #206
• Added interactive visualization to the notebook. by @drewoldag in #208
• UMAP Enchancements by @aritraghsh09 in #200

Full Changelog: v0.1.2...v0.2

v0.1.2

What's Changed

• Set default models to training mode in the train_step. by @drewoldag in #103
• Include all immutable dc.Rect fields in manifest so Resume works in all configs. by @mtauraso in #99
• Handle externally defined default configuration files by @drewoldag in #100
• Splits for HSCDataSet by @mtauraso in #105

Full Changelog: v0.1.1...v0.1.2

v0.1.1

What's Changed

• Implement checkpointing during training by @drewoldag in #90
• Fix error when using fibad with multiple gpus by @drewoldag in #94
• Move config parameters to a different table so that we can define a dataloader easier by @drewoldag in #91

Full Changelog: v0.1...v0.1.1

v0.1

What's Changed

• Bump asv from 0.6.1 to 0.6.3 by @dependabot in #2
• Bump dawidd6/action-download-artifact from 3 to 6 by @dependabot in #1
• CLI scaffolding by @drewoldag in #4
• Downloader prototype and CLI rework by @mtauraso in #7
• Scaffolding for config file handling and model selection by @drewoldag in #6
• Adding bandwith measurement to downloadCutout by @mtauraso in #9
• Organize pyproject.toml, don't accidentally include data/ in commit by @drewoldag in #21
• Retry and resume functionality for downloader by @mtauraso in #17
• Scaffolding for first data loader by @drewoldag in #18
• Setting file names to start with object_id by @mtauraso in #26
• Moving default config, adding it to the docs as well by @drewoldag in #28
• Scaffolding for Fibad library instance object by @mtauraso in #30
• Skip failed chunks by @mtauraso in #27
• Initial version of logging for fibad. by @mtauraso in #31
• Adding configs for retries, retrywait, timeout, and chunksize by @mtauraso in #41
• Make fibad train work end-to-end with HSC cutouts by @mtauraso in #33
• Fix divide by zero in stats module bandwidth calculation by @mtauraso in #46
• Getting Started notebook for downloader and Logging fixup by @mtauraso in #43
• Adding user-specified cutout crop size. by @mtauraso in #49
• Remove Michael's password from default config file by @mtauraso in #52
• Add a configurable filter list to HSC data set by @mtauraso in #53
• Multithreaded manifest-based downloader by @mtauraso in #57
• Bump asv from 0.6.3 to 0.6.4 by @dependabot in #55
• Proper multithreaded stats by @mtauraso in #60
• Various Fixups for klone download by @mtauraso in #62
• First implementation with pytorch ignite to support cpu->multi-gpu scaling by @drewoldag in #54
• Add a fibad top-level method to get the dimensions of downloaded files by @mtauraso in #66
• Centralize config defaults, add safety for Config updates by @mtauraso in #68
• Initial commit to consolidate name and external_cls. by @drewoldag in #71
• Predict and Output scaffolding by @mtauraso in #70
• Switching to Dataset based modularity by @mtauraso in #76
• Bump pre-commit-ci/lite-action from 1.0.2 to 1.0.3 by @dependabot in #78
• Updating to latest PPT version by @drewoldag in #79
• Adding/Updating Readme by @aritraghsh09 in #81
• Reduce pr template size by @drewoldag in #85
• Fixing issue with ConfigDict. by @mtauraso in #87
• Fixing CNN shape issue by @drewoldag in #84

New Contributors

• @dependabot made their first contribution in #2
• @drewoldag made their first contribution in #4
• @mtauraso made their first contribution in #7
• @aritraghsh09 made their first contribution in #81

Full Changelog: https://github.com/lincc-frameworks/fibad/commits/v0.1