Starstruk Logo MarkStarstruk

v1.12.14

k2-fsa/sherpa-onnxv1.12.14Sep 18, 2025by github-actions[bot]

AI Summary

Version 1.12.14 focuses on stability and new language bindings, adding Dart API for Spoken Language Identification and providing pre-compiled wheels for CUDA 12.x. It also fixes TDT decoding and adds a simulated streaming ASR example.

Key Highlights

  • Added Dart API for Spoken Language Identification.
  • Provided pre-compiled binaries/wheels for CUDA 12.x and ONNX Runtime 1.22.0.
  • Added a C++ example for simulated streaming ASR.
  • Fixed TDT decoding for NeMo TDT transducers.

New Features

  • Dart SLI API
  • CUDA 12.x wheels
  • Simulated streaming ASR C++ example
  • TDT decode fix

Full Release Notes

## What's Changed
* Fix setting rknn core mask by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2594
* Add Dart API for spoken language identification by @kibsoft in https://github.com/k2-fsa/sherpa-onnx/pull/2596
* Add CI tests for dart spoken language identifcation example by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2598
* Provide pre-compiled shepra-onnx libs/binaries for CUDA 12.x + onnxruntime 1.22.0 by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2599
* Provide pre-compiled whls for cuda 12.x on Linux x64 and Windows x64 by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2601
* Fix TDT decoding for NeMo TDT transducers by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2606
* Add a C++ example for simulated streaming ASR by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2607
* Release v1.12.14 by @csukuangfj in https://github.com/k2-fsa/sherpa-onnx/pull/2608

## New Contributors
* @kibsoft made their first contribution in https://github.com/k2-fsa/sherpa-onnx/pull/2596

**Full Changelog**: https://github.com/k2-fsa/sherpa-onnx/compare/v1.12.13...v1.12.14