* add Usage * LookupCustomMetadataMap() is deprecated in 1.13.1, so change to 1.12.1 Co-authored-by: weiliang <weiliang.chong@day-care.cn>
28 lines
824 B
Markdown
28 lines
824 B
Markdown
# Usage
|
||
|
||
Most of AI engineers are not familiar with Android development, this is a simple ‘how to’.
|
||
|
||
1. Train your model with your data
|
||
|
||
2. Export pytorch model to onnx model
|
||
|
||
3. Convert onnx model for mobile deployment
|
||
|
||
```bash
|
||
python -m onnxruntime.tools.convert_onnx_models_to_ort your-model.onnx
|
||
```
|
||
you will get `your-model.ort` and `your-model.with_runtime_opt.ort`
|
||
|
||
|
||
4. Install Android Studio and open path of wekws/runtime/android and build
|
||
|
||
*NOTE:* The default feature_dim in code is 40, if your model’s is 80, change it here `./app/src/main/cpp/wekws.cc`
|
||
|
||
```C++
|
||
feature_config = std::make_shared<wenet::FeaturePipelineConfig>(40, 16000); // 40 -> 80
|
||
```
|
||
|
||
It’s also can be built on Linux by runing `bash ./gradlew build`
|
||
|
||
5. Install `app/build/outputs/apk/debug/app-debug.apk` to your phone and try it.
|