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
-
./SherpaOnnxHar It is for building
sherpa_onnx.har. If you don't need to change the C++ or Typescript code of sherpa-onnx, then you can download pre-builtsherpa_onnx.harfrom us. Just runohpm install sherpa_onnx. Please refer to our doc if you want to buildsherpa-onnxfrom source. -
./SherpaOnnxSpeakerDiarization It shows how to run on-device speaker diarization.
-
./SherpaOnnxSpeakerIdentification It shows how to use speaker embedding models for on-device speaker identification.
-
./SherpaOnnxStreamingAsr It shows how to use streaming ASR models for real-time on-device speech recognition.
-
./SherpaOnnxTts It shows how to run on-device text-to-speech. Please see the doc at https://k2-fsa.github.io/sherpa/onnx/harmony-os/tts.html
-
./SherpaOnnxVadAsr It shows how to use VAD + Non-streaming ASR for speech recognition. Please see the doc at https://k2-fsa.github.io/sherpa/onnx/harmony-os/vad-asr.html