sense-voice-with-hr-cxx-api.cc
3.1 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
// cxx-api-examples/sense-voice-with-hr-cxx-api.cc
//
// Copyright (c) 2024-2025 Xiaomi Corporation
//
// This file demonstrates how to use sense voice with sherpa-onnx's C++ API.
//
// clang-format off
//
// wget https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17.tar.bz2
// tar xvf sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17.tar.bz2
// rm sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17.tar.bz2
//
// wget https://github.com/k2-fsa/sherpa-onnx/releases/download/hr-files/dict.tar.bz2
// tar xf dict.tar.bz2
//
// wget https://github.com/k2-fsa/sherpa-onnx/releases/download/hr-files/replace.fst
// wget https://github.com/k2-fsa/sherpa-onnx/releases/download/hr-files/test-hr.wav
// wget https://github.com/k2-fsa/sherpa-onnx/releases/download/hr-files/lexicon.txt
//
// clang-format on
#include <chrono> // NOLINT
#include <iostream>
#include <string>
#include "sherpa-onnx/c-api/cxx-api.h"
int32_t main() {
using namespace sherpa_onnx::cxx; // NOLINT
OfflineRecognizerConfig config;
config.model_config.sense_voice.model =
"./sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/model.int8.onnx";
config.model_config.sense_voice.use_itn = true;
config.model_config.sense_voice.language = "auto";
config.model_config.tokens =
"./sherpa-onnx-sense-voice-zh-en-ja-ko-yue-2024-07-17/tokens.txt";
config.hr.dict_dir = "./dict";
config.hr.lexicon = "./lexicon.txt";
// Please see
// https://colab.research.google.com/drive/1jEaS3s8FbRJIcVQJv2EQx19EM_mnuARi?usp=sharing
// for how to generate your own replace.fst
config.hr.rule_fsts = "./replace.fst";
config.model_config.num_threads = 1;
std::cout << "Loading model\n";
OfflineRecognizer recognizer = OfflineRecognizer::Create(config);
if (!recognizer.Get()) {
std::cerr << "Please check your config\n";
return -1;
}
std::cout << "Loading model done\n";
std::string wave_filename = "./test-hr.wav";
Wave wave = ReadWave(wave_filename);
if (wave.samples.empty()) {
std::cerr << "Failed to read: '" << wave_filename << "'\n";
return -1;
}
std::cout << "Start recognition\n";
const auto begin = std::chrono::steady_clock::now();
OfflineStream stream = recognizer.CreateStream();
stream.AcceptWaveform(wave.sample_rate, wave.samples.data(),
wave.samples.size());
recognizer.Decode(&stream);
OfflineRecognizerResult result = recognizer.GetResult(&stream);
const auto end = std::chrono::steady_clock::now();
const float elapsed_seconds =
std::chrono::duration_cast<std::chrono::milliseconds>(end - begin)
.count() /
1000.;
float duration = wave.samples.size() / static_cast<float>(wave.sample_rate);
float rtf = elapsed_seconds / duration;
std::cout << "text: " << result.text << "\n";
printf("Number of threads: %d\n", config.model_config.num_threads);
printf("Duration: %.3fs\n", duration);
printf("Elapsed seconds: %.3fs\n", elapsed_seconds);
printf("(Real time factor) RTF = %.3f / %.3f = %.3f\n", elapsed_seconds,
duration, rtf);
return 0;
}