---
title: k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801
canonical_url: "https://www.modelscope.cn/models/manyeyes/k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801"
md_url: "https://www.modelscope.cn/models/manyeyes/k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801.md"
repository: manyeyes/k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801
chinese_name: "zipformer-非流式语音识别-日语-onnx模型-reazonspeech-2024-08-01-Japanese"
last_updated: 2025-04-26
license: "Apache License 2.0"
pipeline_tag: auto-speech-recognition
tasks:
  - auto-speech-recognition
library_name:
  - onnx
  - pytorch
frameworks:
  - PyTorch
downloads: 25
stars: 0
language:
  - ja
tags:
  - automatic-speech-recognition
---

# k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801

> k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801 - manyeyes 在 ModelScope 开源的模型。zipformer-非流式语音识别-日语-onnx模型-reazonspeech-2024-08-01-Japanese

manyeyes/k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801 是 ModelScope 魔搭社区上的auto-speech-recognition模型，采用 Apache License 2.0 许可。

- **Repository**: manyeyes/k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801
- **License**: Apache License 2.0
- **Tasks**: auto-speech-recognition
- **Downloads**: 25
- **Stars**: 0
- **Last updated**: 2025-04-26

Source: https://www.modelscope.cn/models/manyeyes/k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801

---

# reazonspeech-k2-v2

> This repo is forked from https://huggingface.co/reazon-research/reazonspeech-k2-v2

`reazonspeech-k2-v2` is an automatic speech recognition (ASR) model
trained on [ReazonSpeech v2.0 corpus](https://huggingface.co/datasets/reazon-research/reazonspeech).

This model provides end-to-end Japanese speech recognition based on
[Next-gen Kaldi](https://k2-fsa.org/).

## Model Architecture

* Character-based RNN-T model. The total parameter count is 159.34M.

* This model utilizes an enhanced Transformer architecture called
  [Zipformer](https://arxiv.org/abs/2310.11230).

* The training recipe is available on
  [k2-fsa/icefall](https://github.com/k2-fsa/icefall/tree/master/egs/reazonspeech/ASR).

Note that this model can process Japanese audio clips up to ~30 seconds.

## Usage

We recommend to use this model through our
[reazonspeech](https://github.com/reazon-research/reazonspeech)
library.

```
from reazonspeech.k2.asr import load_model, transcribe, audio_from_path

audio = audio_from_path("speech.wav")
model = load_model()
ret = transcribe(model, audio)
print(ret.text)
```

## License

[Apaceh Licence 2.0](https://choosealicense.com/licenses/apache-2.0/)


### 当前模型的贡献者未提供更加详细的模型介绍。模型文件和权重，可浏览“模型文件”页面获取。
#### 您可以通过如下git clone命令，或者ModelScope SDK来下载模型

SDK下载
```bash
#安装ModelScope
pip install modelscope
```
```python
#SDK模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('manyeyes/k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801')
```
Git下载
```
#Git模型下载
git clone https://www.modelscope.cn/manyeyes/k2transducer-zipformer-ja-onnx-offline-reazonspeech-20240801.git
```

<p style="color: lightgrey;">如果您是本模型的贡献者，我们邀请您根据<a href="https://modelscope.cn/docs/ModelScope%E6%A8%A1%E5%9E%8B%E6%8E%A5%E5%85%A5%E6%B5%81%E7%A8%8B%E6%A6%82%E8%A7%88" style="color: lightgrey; text-decoration: underline;">模型贡献文档</a>，及时完善模型卡片内容。</p>