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
This folder contains examples about using sherpa-onnx's object pascal APIs with non-streaming models for speech recognition.
| File | Description |
|---|---|
| run-dolphin-ctc.sh | Use a non-streaming Dolphin CTC model for speech recognition |
| run-nemo-ctc.sh | Use a non-streaming NeMo CTC model for speech recognition |
| run-nemo-transducer.sh | Use a non-streaming NeMo transducer model for speech recognition |
| run-paraformer-itn.sh | Use a non-streaming Paraformer model for speech recognition with inverse text normalization for numbers |
| run-paraformer.sh | Use a non-streaming Paraformer model for speech recognition |
| run-sense-voice.sh | Use a non-streaming SenseVoice model for speech recognition |
| run-telespeech-ctc.sh | Use a non-streaming TeleSpeech CTC model for speech recognition |
| run-whisper.sh | Use a Whisper model for speech recognition |
| run-zipformer-transducer.sh | Use a non-streaming Zipformer transducer model for speech recognition |