3bf986d08d
This PR adds support for non-streaming Zipformer CTC ASR models across multiple language bindings, WebAssembly, examples, and CI workflows. - Introduces a new OfflineZipformerCtcModelConfig in C/C++, Python, Swift, Java, Kotlin, Go, Dart, Pascal, and C# APIs - Updates initialization, freeing, and recognition logic to include Zipformer CTC in WASM and Node.js - Adds example scripts and CI steps for downloading, building, and running Zipformer CTC models Model doc is available at https://k2-fsa.github.io/sherpa/onnx/pretrained_models/offline-ctc/icefall/zipformer.html
Introduction
sherpa-onnx is one of the deployment frameworks of Next-gen Kaldi.
It supports speech-to-text, text-to-speech, speaker diarization, and VAD using onnxruntime without Internet connection.
It also supports embedded systems, Android, iOS, HarmonyOS, Raspberry Pi, RISC-V, x86_64 servers, websocket server/client, C/C++, Python, Kotlin, C#, Go, NodeJS, Java, Swift, Dart, JavaScript, Flutter, Object Pascal, Lazarus, Rust, etc.
Installation
To use sherpa-onnx in your project, please either use
ohpm install sherpa_onnx
or update your oh-package.json5 to include the following:
"dependencies": {
"sherpa_onnx": "1.12.3",
},
Note that we recommend always using the latest version.
Examples
| Demo | URL | Description |
|---|---|---|
| SherpaOnnxStreamingAsr | Address | On-device real-time/streaming speech recognition with Next-gen Kaldi |
| SherpaOnnxVadAsr | Address | It shows how to use VAD with a non-streaming ASR model for on-device speech recognition without accessing the network |
| SherpaOnnxTts | Address | It shows how to use Next-gen Kaldi for on-device text-to-speech (TTS, i.e., speech synthesis) |
| SherpaOnnxSpeakerDiarization | Address | On-device speaker diarization with Next-gen Kaldi |
| SherpaOnnxSpeakerIdentification | Address | On-device speaker identification with Next-gen Kaldi |
Documentation
If you have any issues, please either look at our doc at https://k2-fsa.github.io/sherpa/onnx/ or create an issue at https://github.com/k2-fsa/sherpa-onnx/issues