Do you have plans to open source reazonspeech-zipformer-large model?

[jiangzhuo]

I noticed this commit. It sounds like a zipformer model trained on the reazonspeech corpus? It will be very useful.

[fujimotos]

It sounds like a zipformer model trained on the reazonspeech corpus?

@jiangzhuo Right, we are currently preparing a release. I think we can release v2.1 this week or so.

I'll let you know once we have done the release. WFM.

[fujimotos]

ReazonSpeech v2.1 is now public. Check out the following blog post for more details:

https://research.reazon.jp/blog/2024-08-01-ReazonSpeech.html

[csukuangfj]

FYI: We have added it to sherpa-onnx.

Please see https://k2-fsa.github.io/sherpa/onnx/pretrained_models/offline-transducer/zipformer-transducer-models.html#sherpa-onnx-zipformer-ja-reazonspeech-2024-08-01-japanese

---

We have also made an Android demo for it with silero-vad, which can be found at https://www.bilibili.com/video/BV1BRvve1ESV

You can find the APK at https://k2-fsa.github.io/sherpa/onnx/vad/apk-asr.html

e.g., https://huggingface.co/csukuangfj/sherpa-onnx-apk/resolve/main/vad-asr/1.10.20/sherpa-onnx-1.10.20-arm64-v8a-vad_asr-ja-zipformer_reazonspeech.apk

---

Hope that a streaming model would be available soon

[jiangzhuo]

Thank you for your work, looking forward to future progress, such as streaming ASR capabilities.

[fujimotos]

FYI: We have added it to sherpa-onnx. ... We have also made an Android demo for it with silero-vad, which can be found at

@csukuangfj I just shared your post with our team, and they were both throughly impressed and thrilled by your demo. Thank you!