0.1.5

codex-team/editor.js0.1.5Apr 28, 2026by zhu-han

AI Summary

OmniVoice training and audio processing improvements.

Key Highlights

Added support for SDPA training.
Switched to torchaudio resampling for consistency.

New Features

Training support with SDPA
Torchaudio resampling implementation

Full Release Notes

## What's Changed

* support training with SDPA a65149e000fcc6515f1d96eabcbcf99837875dc8
* switch to torchaudio resampling for consistency with training 6a3f23df5b23c93210fcb756f3c229ad40eb84db
* fix unmask schedule d8ab65f93a69b71eff3e917d70e78618f17f3420
* fix model loading without internet e7ef814012a6ebdd86f954916fa6aff2599229ab
* set load_asr=True in colab notebook f7f5040646f4fb8eef84e011adebec0c375ea6ae
* add issue templates afef76b6a2c5e5a0f875e615dc68d34ed213720a
* docs: add more tips and refine docs 7a68a5cffa71b904a862f4870b246966deebadf7 2755be9b5e54f926f5360a8ce88b6e4925117e80 0ca42732e980be398a5fd5b16b8efb0a923cc3c0 4a4b2295d822c9ab96556c83fce467860519ea27 e6664a37ac71279c892fa9c142e88fc4cbcbf298
* docs: add omnivoice-rs to community projects by @FerrisMind in https://github.com/k2-fsa/OmniVoice/pull/94
* docs: add omnivoice-trtllm to community projects by @tlitech in https://github.com/k2-fsa/OmniVoice/pull/110

## New Contributors
* @FerrisMind made their first contribution in https://github.com/k2-fsa/OmniVoice/pull/94
* @tlitech made their first contribution in https://github.com/k2-fsa/OmniVoice/pull/110

**Full Changelog**: https://github.com/k2-fsa/OmniVoice/compare/0.1.4...0.1.5