Inverse text normalization API for other programming languages (#1019)

This commit is contained in:
Fangjun Kuang
2024-06-17 17:02:39 +08:00
committed by GitHub
parent b0f7ed3ee3
commit 6e09933d99
39 changed files with 669 additions and 104 deletions

View File

@@ -0,0 +1,63 @@
// Copyright (c) 2024 Xiaomi Corporation
import 'dart:io';
import 'dart:typed_data';
import 'package:args/args.dart';
import 'package:sherpa_onnx/sherpa_onnx.dart' as sherpa_onnx;
import './init.dart';
void main(List<String> arguments) async {
await initSherpaOnnx();
final parser = ArgParser()
..addOption('model', help: 'Path to the paraformer model')
..addOption('tokens', help: 'Path to tokens.txt')
..addOption('rule-fsts',
help: 'Path to rule fsts for inverse text normalization')
..addOption('input-wav', help: 'Path to input.wav to transcribe');
final res = parser.parse(arguments);
if (res['model'] == null ||
res['tokens'] == null ||
res['rule-fsts'] == null ||
res['input-wav'] == null) {
print(parser.usage);
exit(1);
}
final model = res['model'] as String;
final tokens = res['tokens'] as String;
final ruleFsts = res['rule-fsts'] as String;
final inputWav = res['input-wav'] as String;
final paraformer = sherpa_onnx.OfflineParaformerModelConfig(
model: model,
);
final modelConfig = sherpa_onnx.OfflineModelConfig(
paraformer: paraformer,
tokens: tokens,
debug: true,
numThreads: 1,
modelType: 'paraformer',
);
final config = sherpa_onnx.OfflineRecognizerConfig(
model: modelConfig,
ruleFsts: ruleFsts,
);
final recognizer = sherpa_onnx.OfflineRecognizer(config);
final waveData = sherpa_onnx.readWave(inputWav);
final stream = recognizer.createStream();
stream.acceptWaveform(
samples: waveData.samples, sampleRate: waveData.sampleRate);
recognizer.decode(stream);
final result = recognizer.getResult(stream);
print(result.text);
stream.free();
recognizer.free();
}

View File

@@ -0,0 +1,27 @@
#!/usr/bin/env bash
set -ex
dart pub get
if [ ! -f ./sherpa-onnx-paraformer-zh-2023-03-28/tokens.txt ]; then
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-paraformer-zh-2023-03-28.tar.bz2
tar xvf sherpa-onnx-paraformer-zh-2023-03-28.tar.bz2
rm sherpa-onnx-paraformer-zh-2023-03-28.tar.bz2
fi
if [ ! -f ./itn-zh-number.wav ]; then
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/itn-zh-number.wav
fi
if [ ! -f ./itn_zh_number.fst ]; then
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/itn_zh_number.fst
fi
dart run \
./bin/paraformer-itn.dart \
--model ./sherpa-onnx-paraformer-zh-2023-03-28/model.int8.onnx \
--tokens ./sherpa-onnx-paraformer-zh-2023-03-28/tokens.txt \
--rule-fsts ./itn_zh_number.fst \
--input-wav ./itn-zh-number.wav