v2021.0.6

What's Changed

• Updating README and other links in the repo to reflect the organization change from neuralaudio to hearbenchmark
• Update Organization Name by @jorshi in #21
• Version bump to 2021.0.6 by @jorshi in #22

Full Changelog: v2021.0.5-release...v2021.0.6

v2021.0.5

• Added 32KHz audio validation.

v2021.0.3

• Update python version support: 3.7 & 3.8 supported, 3.9 not officially supported
• Check timestamp embeddings with a few different lengths of audio
• Added warning regarding common timestamp bug involving rounding errors causing drifts
• Timestamp units are more clear now -- milliseconds
• Added checks that ensures timestamps start and end at the correct time in relation to the input audio

v2021.0.2

Initial release of the hear-validator package.

• A command-line util to verify an audio embedding module and model against the HEAR 2021 API
• PyTorch and TensorFlow model validation