v0.8.0

[sammlapp]

OpenSoundscape 0.8.0 Release

This release represent a substantial update to opensoundscape, containing new features, bug fixes, and breaking changes to some parts of the API.

New Features:

The notebook New in OpenSoundscape 0.8.0 summarizes many of the new features, with code examples. New features include:

• logging progress of model training and prediction to Weights and Biases
• writing mp3 files (thanks to soundfile 1.1.0)
• training CNNs on long audio files without splitting into clips
• updated Audio metadata reading and writing
• Audio classmethods noise and silence
• Audio method normalize
• top level imports of most used classes. Eg, from opensoundscape import Audio, Spectrogram, CNN
• debug preprocessing by asking train() and predict() by passing raise_errors=True to raise instead of catch Preprocessing errors
• option to keep all intermediate versions of preprocessed sample with preprocessor.forward(...,trace=True)
• new augmentations in actions module: audio_random_gain and audio_add_noise
• added support for EfficientNet architecture
• new Audio object properties for .rms and .dBFS

Breaking changes:

• Audio.duration is now a property (use .duration not .duration()), as are several Spectrogram attributes that were previously methods
• CNN.predict() now only returns one argument, the dataframe of scores. Optionally, return a second argument with invalid samples list. Use predict_multi_target_labels or predict_single_target_labels from the metrics module to generate 0/1 outputs from continuous scores.
• a few arguments to functions have changed names and order

Merged Pull Requests

• update docs to clarify loading old models by @sammlapp in update docs to clarify loading old models #572
• update docs to clarify loading old models by @sammlapp in update docs to clarify loading old models #571
• Handles file with zero annotations by @jatinkhilnani in Handles file with zero annotations #593
• Updated BoxedAnnotations docstring by @jatinkhilnani in Updated BoxedAnnotations docstring #594
• Minor documentation updates by @sammlapp in Minor documentation updates #608
• resolves clip_times argument not clearly documented in preprocessors or dataset #591 by @sammlapp in resolves #591 #612
• Adds Python 3.9 version to CI and a macos runner. by @louisfh in Adds Python 3.9 version to CI and a macos runner. #595
• Inconsistent use of the term 'model' #552 Standardize references to 'model' object by @louisfh in #552 Standardize references to 'model' object #587
• fixes Spectrogram methods implementation #558 with named args for Spectrogram methods by @louisfh in fixes #558 with named args for Spectrogram methods #585
• Support writing mp3 (soundfile 0.11) by @sammlapp in Support writing mp3 (soundfile 0.11) #613
• resolves load_channels_as_audio fails when file has 1 channel #614 - avoid bug in load channels of mono by @sammlapp in resolves #614 - avoid bug in load channels of mono #616
• Add WandB logging and Support training on a dataframe of clip labels by @sammlapp in Add WandB logging and Support training on a dataframe of clip labels #609
• log top-scoring samples to wandb (resolves wandb: log top scoring samples #575) by @sammlapp in log top-scoring samples to wandb (resolves #575) #619
• Feat 478 efficientnet and refactor cnn_architectures by @sammlapp in Feat 478 efficientnet and refactor cnn_architectures #620
• Bump dependencies by @sammlapp in Bump dependencies #621
• Feat 569 audio by @sammlapp in Feat 569 audio #622
• Resolves Training with GPU on M1 error #563 mps logit not implemented by @sammlapp in Resolves #563 mps logit not implemented #626
• Add mix and concat by @sammlapp in Add mix and concat #623
• explicit arguments for scipy.spectrogram by @sammlapp in explicit arguments for scipy.spectrogram #627
• use explicit arguments for softmax() by @sammlapp in use explicit arguments for softmax() #629
• add json comment-string metadata support by @sammlapp in add json comment-string metadata support #632
• Issue561 - Flat Module Structure in Docs by @syunkova in Issue561 - Flat Module Structure in Docs #638
• add audio_random_gain in actions by @sammlapp in add audio_random_gain in actions #637
• modify predict to only return score df by @sammlapp in modify predict to only return score df #628
• Audio add noise augmentation by @sammlapp in Audio add noise augmentation #639
• Issue 617 clipdf by @sammlapp in Issue 617 clipdf #631
• Issue 636 one output node for single-class classification by @sammlapp in Issue 636 one output node for single-class classification #641
• Change default args from lists to tuple by @sammlapp in Change default args from lists to tuple #644
• Issue543_v2: Use_pretrained is deprecated by @syunkova in Issue543_v2: Use_pretrained is deprecated #645
• Fixed - preprocess - actions by @jatinkhilnani in Fixed - preprocess - actions #570
• Issue 562 preprocessing raise errors by @louisfh in Issue 562 preprocessing raise errors #646
• Issue_510_annotations by @sammlapp in Issue_510_annotations #647
• Debug/trace feature by @jatinkhilnani in Debug/trace feature #596
• version 0.8.0 release by @sammlapp in version 0.8.0 release #648

Full Changelog: v0.7.1...v0.8.0

---

This discussion was created from the release v0.8.0.