Issues

• dtype float16 does not work in Audio.from_file

  blocked
  kitzeslab/opensoundscape#1287 · Filter by author sammlapp opened on Sep 23, 2026
  1

• enable passing a generic logger to inference/training methods

  kitzeslab/opensoundscape#1281 · Filter by author sammlapp opened on Aug 28, 2026

• Don't skip localization for detections at the start or end of a file

  feature requestmodule:localization
  kitzeslab/opensoundscape#1278 · Filter by author louisfh opened on Aug 24, 2026

• allow streamed output writing to parquet/csv rather than aggregating predictions/embeddings in memory

  module:ml
  kitzeslab/opensoundscape#1277 · Filter by author sammlapp opened on Aug 19, 2026

• consider migrating from pandas DF to Polar

  kitzeslab/opensoundscape#1268 · Filter by author sammlapp opened on Jul 22, 2026

• separate public, advanced, and private APIs

  kitzeslab/opensoundscape#1260 · Filter by author sammlapp opened on Jun 27, 2026

• implement Auto-ML for shallow classification

  kitzeslab/opensoundscape#1247 · Filter by author sammlapp opened on May 11, 2026

• localized event filtering

  kitzeslab/opensoundscape#1225 · Filter by author sammlapp opened on Feb 17, 2026

• Make it easy to save training curves locally

  kitzeslab/opensoundscape#1218 · Filter by author louisfh opened on Jan 29, 2026
  1

• Whombat label ingestion

  kitzeslab/opensoundscape#1204 · Filter by author sammlapp opened on Dec 22, 2025

• Audacity label ingestion

  kitzeslab/opensoundscape#1203 · Filter by author sammlapp opened on Dec 22, 2025

• Labeled data ingestion for train/eval

  kitzeslab/opensoundscape#1201 · Filter by author sammlapp opened on Dec 22, 2025

• Clarify pytz localization steps in documentation

  kitzeslab/opensoundscape#1199 · Filter by author rhine3 opened on Dec 17, 2025

• During localization, clarify need for trimming and add check/warning

  kitzeslab/opensoundscape#1198 · Filter by author rhine3 opened on Dec 17, 2025

• Enable localization pipeline to be run without localized datetimes

  kitzeslab/opensoundscape#1197 · Filter by author rhine3 opened on Dec 17, 2025

• Enable audio speed change at inference time

  kitzeslab/opensoundscape#1195 · Filter by author sammlapp opened on Dec 17, 2025

• switch training paradigm from epochs to steps/updates

  in_progress
  kitzeslab/opensoundscape#1190 · Filter by author sammlapp opened on Dec 6, 2025
  1

  

• add model evaluation tutorial notebook

  kitzeslab/opensoundscape#1187 · Filter by author sammlapp opened on Dec 3, 2025

• More intuitive API

  in_progress
  kitzeslab/opensoundscape#1183 · Filter by author sammlapp opened on Nov 26, 2025
  1

• Add .classes attribute to BoxedAnnotations

  kitzeslab/opensoundscape#1182 · Filter by author sammlapp opened on Nov 20, 2025

• allow Spectrogram(Audio) initialization

  kitzeslab/opensoundscape#1177 · Filter by author sammlapp opened on Oct 30, 2025

• take advantage of batched GPU Spectrogram operations

  kitzeslab/opensoundscape#1176 · Filter by author sammlapp opened on Oct 30, 2025

• integrate / use or provide examples of using torch-audiomentations

  kitzeslab/opensoundscape#1157 · Filter by author sammlapp opened on Sep 19, 2025

• Fast failing for model training with improper preprocessing

  module:ml
  kitzeslab/opensoundscape#1150 · Filter by author sammlapp opened on Aug 4, 2025

• Allow files with no annotation column when creating BoxedAnnotations

  kitzeslab/opensoundscape#1135 · Filter by author syunkova opened on May 27, 2025
  2