xuning / sherpaonnx

转到一个项目
Toggle navigation Toggle navigation pinning
Fangjun Kuang
Committed by GitHub
7969cf44 1 parent 0bc571f6

Refactor C# code and support building nuget packages for cross-platforms (#144)

内嵌 并排对比
正在显示 39 个修改的文件 包含 2041 行增加 和 2302 行删除
name: dot-net
on:
push:
branches:
- dot-net
tags:
- '*'
concurrency:
group: dot-net-${{ github.ref }}
cancel-in-progress: true
jobs:
build-libs:
name: dot-net for ${{ matrix.os }}
runs-on: ${{ matrix.os }}
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, windows-latest, macos-latest]
steps:
- uses: actions/checkout@v2
# see https://cibuildwheel.readthedocs.io/en/stable/changelog/
# for a list of versions
- name: Build wheels
uses: pypa/cibuildwheel@v2.11.4
env:
CIBW_BEFORE_BUILD: "pip install -U cmake numpy"
CIBW_BUILD: "cp38-*64"
CIBW_SKIP: "cp27-* cp35-* cp36-* *-win32 pp* *-musllinux* *-manylinux_i686"
CIBW_BUILD_VERBOSITY: 3
CIBW_ENVIRONMENT_LINUX: LD_LIBRARY_PATH='/project/build/bdist.linux-x86_64/wheel/sherpa_onnx/lib'
CIBW_REPAIR_WHEEL_COMMAND_MACOS: ""
- name: Display wheels
shell: bash
run: |
ls -lh ./wheelhouse/*.whl
unzip -l ./wheelhouse/*.whl
- uses: actions/upload-artifact@v2
with:
name: ${{ matrix.os }}-wheels
path: ./wheelhouse/*.whl
build-nuget-packages:
name: build-nuget-packages
runs-on: ubuntu-latest
needs: build-libs
steps:
- uses: actions/checkout@v2
- name: Retrieve artifact from ubuntu-latest
uses: actions/download-artifact@v2
with:
name: ubuntu-latest-wheels
path: ./linux
- name: Retrieve artifact from macos-latest
uses: actions/download-artifact@v2
with:
name: macos-latest-wheels
path: ./macos
- name: Retrieve artifact from windows-latest
uses: actions/download-artifact@v2
with:
name: windows-latest-wheels
path: ./windows
- name: Display wheels
shell: bash
run: |
tree .
- name: Unzip Ubuntu wheels
shell: bash
run: |
cd linux
unzip ./*.whl
tree .
- name: Unzip macOS wheels
shell: bash
run: |
cd macos
unzip ./*.whl
tree .
- name: Unzip Windows wheels
shell: bash
run: |
cd windows
unzip ./*.whl
cp -v ./*.dll sherpa_onnx/lib/
tree .
- name: Setup .NET Core 3.1
uses: actions/setup-dotnet@v1
with:
dotnet-version: 3.1.x
- name: Setup .NET 7.0
uses: actions/setup-dotnet@v1
with:
dotnet-version: 7.0.x
- name: Check dotnet
run: dotnet --info
- name: build nuget packages
shell: bash
run: |
cd scripts/dotnet
./run.sh
ls -lh packages
- uses: actions/upload-artifact@v2
name: upload nuget packages
with:
name: nuget-packages
path: scripts/dotnet/packages/*.nupkg
- name: publish .Net packages to nuget.org
if: github.repository == 'csukuangfj/sherpa-onnx' || github.repository == 'k2-fsa/sherpa-onnx'
shell: bash
env:
API_KEY: ${{ secrets.NUGET_API_KEY }}
run: |
# API_KEY is valid until 2024.05.02
cd scripts/dotnet/packages
dotnet nuget push ./org.k2fsa.sherpa.onnx.*.nupkg --skip-duplicate --api-key $API_KEY --source https://api.nuget.org/v3/index.json
... ...
name: test-dot-net
on:
push:
branches:
- master
paths:
- '.github/workflows/test-dot-net'
- 'dotnet-examples/**'
pull_request:
branches:
- master
paths:
- '.github/workflows/test-dot-net'
- 'dotnet-examples/**'
schedule:
# minute (0-59)
# hour (0-23)
# day of the month (1-31)
# month (1-12)
# day of the week (0-6)
# nightly build at 23:50 UTC time every day
- cron: "50 23 * * *"
concurrency:
group: test-dot-net
cancel-in-progress: true
permissions:
contents: read
jobs:
test-dot-net:
runs-on: ${{ matrix.os }}
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, macos-latest, windows-latest]
steps:
- uses: actions/checkout@v2
with:
fetch-depth: 0
- name: Setup .NET Core 3.1
uses: actions/setup-dotnet@v1
with:
dotnet-version: 3.1.x
- name: Setup .NET 6.0
uses: actions/setup-dotnet@v1
with:
dotnet-version: 6.0.x
- name: Check dotnet
run: dotnet --info
- name: Decode a file
shell: bash
run: |
cd dotnet-examples/
cd online-decode-files
./run.sh
cd ../offline-decode-files
./run-nemo-ctc.sh
./run-paraformer.sh
./run-zipformer.sh
... ...
... ... @@ -57,3 +57,4 @@ sherpa-onnx-nemo-ctc-en-citrinet-512
run-offline-decode-files-nemo-ctc.sh
*.jar
sherpa-onnx-nemo-ctc-*
*.wav
... ...
cmake_minimum_required(VERSION 3.13 FATAL_ERROR)
project(sherpa-onnx)
set(SHERPA_ONNX_VERSION "1.4.1")
set(SHERPA_ONNX_VERSION "1.4.2")
# Disable warning about
#
... ... @@ -37,16 +37,12 @@ endif()
set(CMAKE_INSTALL_RPATH ${SHERPA_ONNX_RPATH_ORIGIN})
set(CMAKE_BUILD_RPATH ${SHERPA_ONNX_RPATH_ORIGIN})
if(BUILD_SHARED_LIBS AND MSVC)
set(CMAKE_WINDOWS_EXPORT_ALL_SYMBOLS ON)
endif()
if(NOT CMAKE_BUILD_TYPE)
message(STATUS "No CMAKE_BUILD_TYPE given, default to Release")
set(CMAKE_BUILD_TYPE Release)
endif()
if(DEFINED ANDROID_ABI)
if(DEFINED ANDROID_ABI AND NOT SHERPA_ONNX_ENABLE_JNI)
message(STATUS "Set SHERPA_ONNX_ENABLE_JNI to ON for Android")
set(SHERPA_ONNX_ENABLE_JNI ON CACHE BOOL "" FORCE)
endif()
... ... @@ -61,6 +57,10 @@ if(SHERPA_ONNX_ENABLE_JNI AND NOT BUILD_SHARED_LIBS)
set(BUILD_SHARED_LIBS ON CACHE BOOL "" FORCE)
endif()
if(BUILD_SHARED_LIBS AND MSVC)
set(CMAKE_WINDOWS_EXPORT_ALL_SYMBOLS ON)
endif()
message(STATUS "CMAKE_BUILD_TYPE: ${CMAKE_BUILD_TYPE}")
message(STATUS "CMAKE_INSTALL_PREFIX: ${CMAKE_INSTALL_PREFIX}")
message(STATUS "BUILD_SHARED_LIBS ${BUILD_SHARED_LIBS}")
... ...
... ... @@ -41,7 +41,6 @@ try:
# -linux_x86_64.whl
self.root_is_pure = False
except ImportError:
bdist_wheel = None
... ... @@ -78,7 +77,6 @@ class BuildExtension(build_ext):
extra_cmake_args += " -DSHERPA_ONNX_ENABLE_CHECK=OFF "
extra_cmake_args += " -DSHERPA_ONNX_ENABLE_PYTHON=ON "
extra_cmake_args += " -DSHERPA_ONNX_ENABLE_PORTAUDIO=ON "
extra_cmake_args += " -DSHERPA_ONNX_ENABLE_C_API=OFF "
extra_cmake_args += " -DSHERPA_ONNX_ENABLE_WEBSOCKET=ON "
if "PYTHON_EXECUTABLE" not in cmake_args:
... ...
// See https://aka.ms/new-console-template for more information
// Copyright (c) 2023 by manyeyes
using SherpaOnnx;
/// Please refer to
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
/// to download pre-trained models. That is, you can find encoder-xxx.onnx
/// decoder-xxx.onnx, joiner-xxx.onnx, and tokens.txt for this struct
/// from there.
/// download model eg:
/// (The directory where the application runs)
/// [/path/to]=System.AppDomain.CurrentDomain.BaseDirectory
/// cd /path/to
/// git clone https://huggingface.co/csukuangfj/sherpa-onnx-zipformer-en-2023-04-01
/// git clone https://huggingface.co/csukuangfj/paraformer-onnxruntime-python-example
/// git clone https://huggingface.co/csukuangfj/sherpa-onnx-nemo-ctc-en-citrinet-512
/// NuGet for sherpa-onnx
/// PM > Install-Package NAudio -version 2.1.0 -Project sherpa-onnx
/// PM > Install-Package SherpaOnnxCsharp -Project sherpa-onnx
// transducer Usage:
/*
.\SherpaOnnx.Examples.exe `
--tokens=./all_models/sherpa-onnx-conformer-en-2023-03-18/tokens.txt `
--encoder=./all_models/sherpa-onnx-conformer-en-2023-03-18/encoder-epoch-99-avg-1.onnx `
--decoder=./all_models/sherpa-onnx-conformer-en-2023-03-18/decoder-epoch-99-avg-1.onnx `
--joiner=./all_models/sherpa-onnx-conformer-en-2023-03-18/joiner-epoch-99-avg-1.onnx `
--num-threads=2 `
--decoding-method=greedy_search `
--debug=false `
./all_models/sherpa-onnx-conformer-en-2023-03-18/test_wavs/0.wav
*/
// paraformer Usage:
/*
.\SherpaOnnx.Examples.exe `
--tokens=./all_models/paraformer-onnxruntime-python-example/tokens.txt `
--paraformer=./all_models/paraformer-onnxruntime-python-example/model.onnx `
--num-threads=2 `
--decoding-method=greedy_search `
--debug=false `
./all_models/paraformer-onnxruntime-python-example/test_wavs/0.wav
*/
// paraformer Usage:
/*
.\SherpaOnnx.Examples.exe `
--tokens=./all_models/paraformer-onnxruntime-python-example/tokens.txt `
--paraformer=./all_models/paraformer-onnxruntime-python-example/model.onnx `
--num-threads=2 `
--decoding-method=greedy_search `
--debug=false `
./all_models/paraformer-onnxruntime-python-example/test_wavs/0.wav
*/
internal class OfflineDecodeFiles
{
static void Main(string[] args)
{
string usage = @"
-----------------------------
transducer Usage:
--tokens=./all_models/sherpa-onnx-conformer-en-2023-03-18/tokens.txt `
--encoder=./all_models/sherpa-onnx-conformer-en-2023-03-18/encoder-epoch-99-avg-1.onnx `
--decoder=./all_models/sherpa-onnx-conformer-en-2023-03-18/decoder-epoch-99-avg-1.onnx `
--joiner=./all_models/sherpa-onnx-conformer-en-2023-03-18/joiner-epoch-99-avg-1.onnx `
--num-threads=2 `
--decoding-method=greedy_search `
--debug=false `
./all_models/sherpa-onnx-conformer-en-2023-03-18/test_wavs/0.wav
paraformer Usage:
--tokens=./all_models/paraformer-onnxruntime-python-example/tokens.txt `
--paraformer=./all_models/paraformer-onnxruntime-python-example/model.onnx `
--num-threads=2 `
--decoding-method=greedy_search `
--debug=false `
./all_models/paraformer-onnxruntime-python-example/test_wavs/0.wav
nemo Usage:
--tokens=./all_models/sherpa-onnx-nemo-ctc-en-citrinet-512/tokens.txt `
--nemo_ctc=./all_models/sherpa-onnx-nemo-ctc-en-citrinet-512/model.onnx `
--num-threads=2 `
--decoding-method=greedy_search `
--debug=false `
./all_models/sherpa-onnx-nemo-ctc-en-citrinet-512/test_wavs/0.wav
-----------------------------
";
if (args.Length == 0)
{
System.Console.WriteLine("Please enter the correct parameters:");
System.Console.WriteLine(usage);
System.Text.StringBuilder sb = new System.Text.StringBuilder();
//args = Console.ReadLine().Split(" ");
while (true)
{
string input = Console.ReadLine();
sb.AppendLine(input);
if (Console.ReadKey().Key == ConsoleKey.Enter)
break;
}
args = sb.ToString().Split("\r\n");
}
Console.WriteLine("Started!\n");
string? applicationBase = System.AppDomain.CurrentDomain.BaseDirectory;
List<string> wavFiles = new List<string>();
Dictionary<string, string> argsDict = GetDict(args, applicationBase, ref wavFiles);
string decoder = argsDict.ContainsKey("decoder") ? Path.Combine(applicationBase, argsDict["decoder"]) : "";
string encoder = argsDict.ContainsKey("encoder") ? Path.Combine(applicationBase, argsDict["encoder"]) : "";
string joiner = argsDict.ContainsKey("joiner") ? Path.Combine(applicationBase, argsDict["joiner"]) : "";
string paraformer = argsDict.ContainsKey("paraformer") ? Path.Combine(applicationBase, argsDict["paraformer"]) : "";
string nemo_ctc = argsDict.ContainsKey("nemo_ctc") ? Path.Combine(applicationBase, argsDict["nemo_ctc"]) : "";
string tokens = argsDict.ContainsKey("tokens") ? Path.Combine(applicationBase, argsDict["tokens"]) : "";
string num_threads = argsDict.ContainsKey("num_threads") ? argsDict["num_threads"] : "";
string decoding_method = argsDict.ContainsKey("decoding_method") ? argsDict["decoding_method"] : "";
string debug = argsDict.ContainsKey("debug") ? argsDict["debug"] : "";
OfflineTransducer offlineTransducer = new OfflineTransducer();
offlineTransducer.EncoderFilename = encoder;
offlineTransducer.DecoderFilename = decoder;
offlineTransducer.JoinerFilename = joiner;
OfflineParaformer offlineParaformer = new OfflineParaformer();
offlineParaformer.Model = paraformer;
OfflineNemoEncDecCtc offlineNemoEncDecCtc = new OfflineNemoEncDecCtc();
offlineNemoEncDecCtc.Model = nemo_ctc;
int numThreads = 0;
int.TryParse(num_threads, out numThreads);
bool isDebug = false;
bool.TryParse(debug, out isDebug);
string decodingMethod = string.IsNullOrEmpty(decoding_method) ? "" : decoding_method;
if ((string.IsNullOrEmpty(encoder) || string.IsNullOrEmpty(decoder) || string.IsNullOrEmpty(joiner))
&& string.IsNullOrEmpty(paraformer)
&& string.IsNullOrEmpty(nemo_ctc))
{
Console.WriteLine("Please specify at least one model");
Console.WriteLine(usage);
}
// batch decode
TimeSpan total_duration = TimeSpan.Zero;
TimeSpan start_time = TimeSpan.Zero;
TimeSpan end_time = TimeSpan.Zero;
List<OfflineRecognizerResultEntity> results = new List<OfflineRecognizerResultEntity>();
if (!(string.IsNullOrEmpty(encoder) || string.IsNullOrEmpty(decoder) || string.IsNullOrEmpty(joiner)))
{
OfflineRecognizer<OfflineTransducer> offlineRecognizer = new OfflineRecognizer<OfflineTransducer>(
offlineTransducer,
tokens,
num_threads: numThreads,
debug: isDebug,
decoding_method: decodingMethod);
List<float[]> samplesList = new List<float[]>();
foreach (string wavFile in wavFiles)
{
TimeSpan duration = TimeSpan.Zero;
float[] samples = AudioHelper.GetFileSamples(wavFile, ref duration);
samplesList.Add(samples);
total_duration += duration;
}
OfflineStream[] streams = offlineRecognizer.CreateOfflineStream(samplesList);
start_time = new TimeSpan(DateTime.Now.Ticks);
offlineRecognizer.DecodeMultipleOfflineStreams(streams);
results = offlineRecognizer.GetResults(streams);
end_time = new TimeSpan(DateTime.Now.Ticks);
}
else if (!string.IsNullOrEmpty(paraformer))
{
OfflineRecognizer<OfflineParaformer> offlineRecognizer = new OfflineRecognizer<OfflineParaformer>(
offlineParaformer,
tokens,
num_threads: numThreads,
debug: isDebug,
decoding_method: decodingMethod);
List<float[]> samplesList = new List<float[]>();
foreach (string wavFile in wavFiles)
{
TimeSpan duration = TimeSpan.Zero;
float[] samples = AudioHelper.GetFileSamples(wavFile, ref duration);
samplesList.Add(samples);
total_duration += duration;
}
OfflineStream[] streams = offlineRecognizer.CreateOfflineStream(samplesList);
start_time = new TimeSpan(DateTime.Now.Ticks);
offlineRecognizer.DecodeMultipleOfflineStreams(streams);
results = offlineRecognizer.GetResults(streams);
end_time = new TimeSpan(DateTime.Now.Ticks);
}
else if (!string.IsNullOrEmpty(nemo_ctc))
{
OfflineRecognizer<OfflineNemoEncDecCtc> offlineRecognizer = new OfflineRecognizer<OfflineNemoEncDecCtc>(
offlineNemoEncDecCtc,
tokens,
num_threads: numThreads,
debug: isDebug,
decoding_method: decodingMethod);
List<float[]> samplesList = new List<float[]>();
foreach (string wavFile in wavFiles)
{
TimeSpan duration = TimeSpan.Zero;
float[] samples = AudioHelper.GetFileSamples(wavFile, ref duration);
samplesList.Add(samples);
total_duration += duration;
}
OfflineStream[] streams = offlineRecognizer.CreateOfflineStream(samplesList);
start_time = new TimeSpan(DateTime.Now.Ticks);
offlineRecognizer.DecodeMultipleOfflineStreams(streams);
results = offlineRecognizer.GetResults(streams);
end_time = new TimeSpan(DateTime.Now.Ticks);
}
foreach (var item in results.Zip<OfflineRecognizerResultEntity, string>(wavFiles))
{
Console.WriteLine("wavFile:{0}", item.Second);
Console.WriteLine("text:{0}", item.First.text.ToLower());
Console.WriteLine("text_len:{0}\n", item.First.text_len.ToString());
}
double elapsed_milliseconds = end_time.TotalMilliseconds - start_time.TotalMilliseconds;
double rtf = elapsed_milliseconds / total_duration.TotalMilliseconds;
Console.WriteLine("num_threads:{0}", num_threads);
Console.WriteLine("decoding_method:{0}", decodingMethod);
Console.WriteLine("elapsed_milliseconds:{0}", elapsed_milliseconds.ToString());
Console.WriteLine("wave total_duration_milliseconds:{0}", total_duration.TotalMilliseconds.ToString());
Console.WriteLine("Real time factor (RTF):{0}", rtf.ToString());
Console.WriteLine("End!");
}
static Dictionary<string, string> GetDict(string[] args, string applicationBase, ref List<string> wavFiles)
{
Dictionary<string, string> argsDict = new Dictionary<string, string>();
foreach (string input in args)
{
string[] ss = input.Split("=");
if (ss.Length == 1)
{
if (!string.IsNullOrEmpty(ss[0]))
{
wavFiles.Add(Path.Combine(applicationBase, ss[0].Trim(new char[] { '-', '`', ' ' })));
}
}
else
{
argsDict.Add(ss[0].Trim(new char[] { '-', '`', ' ' }).Replace("-", "_"), ss[1].Trim(new char[] { '-', '`', ' ' }));
}
}
return argsDict;
}
}
\ No newline at end of file
// See https://aka.ms/new-console-template for more information
// Copyright (c) 2023 by manyeyes
using SherpaOnnx;
/// Please refer to
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
/// to download pre-trained models. That is, you can find encoder-xxx.onnx
/// decoder-xxx.onnx, joiner-xxx.onnx, and tokens.txt for this struct
/// from there.
/// download model eg:
/// (The directory where the application runs)
/// [/path/to]=System.AppDomain.CurrentDomain.BaseDirectory
/// cd /path/to
/// git clone https://huggingface.co/csukuangfj/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20
/// NuGet for sherpa-onnx
/// PM > Install-Package NAudio -version 2.1.0 -Project sherpa-onnx
/// PM > Install-Package SherpaOnnxCsharp -Project sherpa-onnx
// transducer Usage:
/*
.\SherpaOnnx.Examples.exe `
--tokens=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/tokens.txt `
--encoder=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/encoder-epoch-99-avg-1.onnx `
--decoder=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/decoder-epoch-99-avg-1.onnx `
--joiner=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/joiner-epoch-99-avg-1.onnx `
--num-threads=2 `
--decoding-method=modified_beam_search `
--debug=false `
./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/test_wavs/0.wav
*/
internal class OnlineDecodeFile
{
static void Main(string[] args)
{
string usage = @"
-----------------------------
transducer Usage:
--tokens=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/tokens.txt `
--encoder=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/encoder-epoch-99-avg-1.onnx `
--decoder=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/decoder-epoch-99-avg-1.onnx `
--joiner=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/joiner-epoch-99-avg-1.onnx `
--num-threads=2 `
--decoding-method=modified_beam_search `
--debug=false `
./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/test_wavs/0.wav
-----------------------------
";
if (args.Length == 0)
{
System.Console.WriteLine("Please enter the correct parameters:");
System.Console.WriteLine(usage);
System.Text.StringBuilder sb = new System.Text.StringBuilder();
//args = Console.ReadLine().Split(" ");
while (true)
{
string input = Console.ReadLine();
sb.AppendLine(input);
if (Console.ReadKey().Key == ConsoleKey.Enter)
break;
}
args = sb.ToString().Split("\r\n");
}
Console.WriteLine("Started!\n");
string? applicationBase = System.AppDomain.CurrentDomain.BaseDirectory;
List<string> wavFiles = new List<string>();
Dictionary<string, string> argsDict = GetDict(args, applicationBase, ref wavFiles);
string decoder = argsDict.ContainsKey("decoder") ? Path.Combine(applicationBase, argsDict["decoder"]) : "";
string encoder = argsDict.ContainsKey("encoder") ? Path.Combine(applicationBase, argsDict["encoder"]) : "";
string joiner = argsDict.ContainsKey("joiner") ? Path.Combine(applicationBase, argsDict["joiner"]) : "";
string paraformer = argsDict.ContainsKey("paraformer") ? Path.Combine(applicationBase, argsDict["paraformer"]) : "";
string nemo_ctc = argsDict.ContainsKey("nemo_ctc") ? Path.Combine(applicationBase, argsDict["nemo_ctc"]) : "";
string tokens = argsDict.ContainsKey("tokens") ? Path.Combine(applicationBase, argsDict["tokens"]) : "";
string num_threads = argsDict.ContainsKey("num_threads") ? argsDict["num_threads"] : "";
string decoding_method = argsDict.ContainsKey("decoding_method") ? argsDict["decoding_method"] : "";
string debug = argsDict.ContainsKey("debug") ? argsDict["debug"] : "";
OfflineTransducer offlineTransducer = new OfflineTransducer();
offlineTransducer.EncoderFilename = encoder;
offlineTransducer.DecoderFilename = decoder;
offlineTransducer.JoinerFilename = joiner;
OfflineParaformer offlineParaformer = new OfflineParaformer();
offlineParaformer.Model = paraformer;
OfflineNemoEncDecCtc offlineNemoEncDecCtc = new OfflineNemoEncDecCtc();
offlineNemoEncDecCtc.Model = nemo_ctc;
int numThreads = 0;
int.TryParse(num_threads, out numThreads);
bool isDebug = false;
bool.TryParse(debug, out isDebug);
string decodingMethod = string.IsNullOrEmpty(decoding_method) ? "" : decoding_method;
if ((string.IsNullOrEmpty(encoder) || string.IsNullOrEmpty(decoder) || string.IsNullOrEmpty(joiner))
&& string.IsNullOrEmpty(paraformer)
&& string.IsNullOrEmpty(nemo_ctc))
{
Console.WriteLine("Please specify at least one model");
Console.WriteLine(usage);
}
// batch decode
TimeSpan total_duration = TimeSpan.Zero;
TimeSpan start_time = TimeSpan.Zero;
TimeSpan end_time = TimeSpan.Zero;
List<OfflineRecognizerResultEntity> results = new List<OfflineRecognizerResultEntity>();
if (!(string.IsNullOrEmpty(encoder) || string.IsNullOrEmpty(decoder) || string.IsNullOrEmpty(joiner)))
{
OnlineTransducer onlineTransducer = new OnlineTransducer();
onlineTransducer.EncoderFilename = encoder;
onlineTransducer.DecoderFilename = decoder;
onlineTransducer.JoinerFilename = joiner;
//test online
OnlineRecognizer<OnlineTransducer> onlineRecognizer = new OnlineRecognizer<OnlineTransducer>(
onlineTransducer,
tokens,
num_threads: numThreads,
debug: isDebug,
decoding_method: decodingMethod);
foreach (string wavFile in wavFiles)
{
TimeSpan duration = TimeSpan.Zero;
List<float[]> samplesList = AudioHelper.GetChunkSamplesList(wavFile, ref duration);
OnlineStream stream = onlineRecognizer.CreateStream();
start_time = new TimeSpan(DateTime.Now.Ticks);
for (int i = 0; i < samplesList.Count; i++)
{
onlineRecognizer.AcceptWaveForm(stream, 16000, samplesList[i]);
onlineRecognizer.DecodeStream(stream);
OnlineRecognizerResultEntity result_on = onlineRecognizer.GetResult(stream);
Console.WriteLine(result_on.text);
}
total_duration += duration;
}
end_time = new TimeSpan(DateTime.Now.Ticks);
}
double elapsed_milliseconds = end_time.TotalMilliseconds - start_time.TotalMilliseconds;
double rtf = elapsed_milliseconds / total_duration.TotalMilliseconds;
Console.WriteLine("num_threads:{0}", num_threads);
Console.WriteLine("decoding_method:{0}", decodingMethod);
Console.WriteLine("elapsed_milliseconds:{0}", elapsed_milliseconds.ToString());
Console.WriteLine("wave total_duration_milliseconds:{0}", total_duration.TotalMilliseconds.ToString());
Console.WriteLine("Real time factor (RTF):{0}", rtf.ToString());
Console.WriteLine("End!");
}
static Dictionary<string, string> GetDict(string[] args, string applicationBase, ref List<string> wavFiles)
{
Dictionary<string, string> argsDict = new Dictionary<string, string>();
foreach (string input in args)
{
string[] ss = input.Split("=");
if (ss.Length == 1)
{
if (!string.IsNullOrEmpty(ss[0]))
{
wavFiles.Add(Path.Combine(applicationBase, ss[0].Trim(new char[] { '-', '`', ' ' })));
}
}
else
{
argsDict.Add(ss[0].Trim(new char[] { '-', '`', ' ' }).Replace("-", "_"), ss[1].Trim(new char[] { '-', '`', ' ' }));
}
}
return argsDict;
}
}
\ No newline at end of file
// See https://aka.ms/new-console-template for more information
// Copyright (c) 2023 by manyeyes
using SherpaOnnx;
/// Please refer to
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
/// to download pre-trained models. That is, you can find encoder-xxx.onnx
/// decoder-xxx.onnx, joiner-xxx.onnx, and tokens.txt for this struct
/// from there.
/// download model eg:
/// (The directory where the application runs)
/// [/path/to]=System.AppDomain.CurrentDomain.BaseDirectory
/// cd /path/to
/// git clone https://huggingface.co/csukuangfj/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20
/// NuGet for sherpa-onnx
/// PM > Install-Package NAudio -version 2.1.0 -Project sherpa-onnx
/// PM > Install-Package SherpaOnnxCsharp -Project sherpa-onnx
// transducer Usage:
/*
.\SherpaOnnx.Examples.exe `
--tokens=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/tokens.txt `
--encoder=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/encoder-epoch-99-avg-1.onnx `
--decoder=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/decoder-epoch-99-avg-1.onnx `
--joiner=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/joiner-epoch-99-avg-1.onnx `
--num-threads=2 `
--decoding-method=modified_beam_search `
--debug=false `
./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/test_wavs/0.wav `
./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/test_wavs/1.wav
*/
internal class OnlineDecodeFiles
{
static void Main(string[] args)
{
string usage = @"
-----------------------------
transducer Usage:
--tokens=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/tokens.txt `
--encoder=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/encoder-epoch-99-avg-1.onnx `
--decoder=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/decoder-epoch-99-avg-1.onnx `
--joiner=./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/joiner-epoch-99-avg-1.onnx `
--num-threads=2 `
--decoding-method=modified_beam_search `
--debug=false `
./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/test_wavs/0.wav `
./all_models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/test_wavs/1.wav
-----------------------------
";
if (args.Length == 0)
{
System.Console.WriteLine("Please enter the correct parameters:");
System.Console.WriteLine(usage);
System.Text.StringBuilder sb = new System.Text.StringBuilder();
//args = Console.ReadLine().Split(" ");
while (true)
{
string input = Console.ReadLine();
sb.AppendLine(input);
if (Console.ReadKey().Key == ConsoleKey.Enter)
break;
}
args = sb.ToString().Split("\r\n");
}
Console.WriteLine("Started!\n");
string? applicationBase = System.AppDomain.CurrentDomain.BaseDirectory;
List<string> wavFiles = new List<string>();
Dictionary<string, string> argsDict = GetDict(args, applicationBase, ref wavFiles);
string decoder = argsDict.ContainsKey("decoder") ? Path.Combine(applicationBase, argsDict["decoder"]) : "";
string encoder = argsDict.ContainsKey("encoder") ? Path.Combine(applicationBase, argsDict["encoder"]) : "";
string joiner = argsDict.ContainsKey("joiner") ? Path.Combine(applicationBase, argsDict["joiner"]) : "";
string paraformer = argsDict.ContainsKey("paraformer") ? Path.Combine(applicationBase, argsDict["paraformer"]) : "";
string nemo_ctc = argsDict.ContainsKey("nemo_ctc") ? Path.Combine(applicationBase, argsDict["nemo_ctc"]) : "";
string tokens = argsDict.ContainsKey("tokens") ? Path.Combine(applicationBase, argsDict["tokens"]) : "";
string num_threads = argsDict.ContainsKey("num_threads") ? argsDict["num_threads"] : "";
string decoding_method = argsDict.ContainsKey("decoding_method") ? argsDict["decoding_method"] : "";
string debug = argsDict.ContainsKey("debug") ? argsDict["debug"] : "";
OfflineTransducer offlineTransducer = new OfflineTransducer();
offlineTransducer.EncoderFilename = encoder;
offlineTransducer.DecoderFilename = decoder;
offlineTransducer.JoinerFilename = joiner;
OfflineParaformer offlineParaformer = new OfflineParaformer();
offlineParaformer.Model = paraformer;
OfflineNemoEncDecCtc offlineNemoEncDecCtc = new OfflineNemoEncDecCtc();
offlineNemoEncDecCtc.Model = nemo_ctc;
int numThreads = 0;
int.TryParse(num_threads, out numThreads);
bool isDebug = false;
bool.TryParse(debug, out isDebug);
string decodingMethod = string.IsNullOrEmpty(decoding_method) ? "" : decoding_method;
if ((string.IsNullOrEmpty(encoder) || string.IsNullOrEmpty(decoder) || string.IsNullOrEmpty(joiner))
&& string.IsNullOrEmpty(paraformer)
&& string.IsNullOrEmpty(nemo_ctc))
{
Console.WriteLine("Please specify at least one model");
Console.WriteLine(usage);
}
// batch decode
TimeSpan total_duration = TimeSpan.Zero;
TimeSpan start_time = TimeSpan.Zero;
TimeSpan end_time = TimeSpan.Zero;
List<OnlineRecognizerResultEntity> results = new List<OnlineRecognizerResultEntity>();
if (!(string.IsNullOrEmpty(encoder) || string.IsNullOrEmpty(decoder) || string.IsNullOrEmpty(joiner)))
{
OnlineTransducer onlineTransducer = new OnlineTransducer();
onlineTransducer.EncoderFilename = encoder;
onlineTransducer.DecoderFilename = decoder;
onlineTransducer.JoinerFilename = joiner;
//test online
OnlineRecognizer<OnlineTransducer> onlineRecognizer = new OnlineRecognizer<OnlineTransducer>(
onlineTransducer,
tokens,
num_threads: numThreads,
debug: isDebug,
decoding_method: decodingMethod);
List<float[]> samplesList = new List<float[]>();
foreach (string wavFile in wavFiles)
{
TimeSpan duration = TimeSpan.Zero;
float[] samples = AudioHelper.GetFileSamples(wavFile, ref duration);
samplesList.Add(samples);
total_duration += duration;
}
start_time = new TimeSpan(DateTime.Now.Ticks);
List<OnlineStream> streams = new List<OnlineStream>();
foreach (float[] samples in samplesList)
{
OnlineStream stream = onlineRecognizer.CreateStream();
onlineRecognizer.AcceptWaveForm(stream, 16000, samples);
streams.Add(stream);
onlineRecognizer.InputFinished(stream);
}
onlineRecognizer.DecodeMultipleStreams(streams);
results = onlineRecognizer.GetResults(streams);
foreach (OnlineRecognizerResultEntity result in results)
{
Console.WriteLine(result.text);
}
end_time = new TimeSpan(DateTime.Now.Ticks);
}
foreach (var item in results.Zip<OnlineRecognizerResultEntity, string>(wavFiles))
{
Console.WriteLine("wavFile:{0}", item.Second);
Console.WriteLine("text:{0}", item.First.text.ToLower());
Console.WriteLine("text_len:{0}\n", item.First.text_len.ToString());
}
double elapsed_milliseconds = end_time.TotalMilliseconds - start_time.TotalMilliseconds;
double rtf = elapsed_milliseconds / total_duration.TotalMilliseconds;
Console.WriteLine("num_threads:{0}", num_threads);
Console.WriteLine("decoding_method:{0}", decodingMethod);
Console.WriteLine("elapsed_milliseconds:{0}", elapsed_milliseconds.ToString());
Console.WriteLine("wave total_duration_milliseconds:{0}", total_duration.TotalMilliseconds.ToString());
Console.WriteLine("Real time factor (RTF):{0}", rtf.ToString());
Console.WriteLine("End!");
}
public void AnotherWayOfDecodeFiles(string encoder, string decoder, string joiner, string tokens, int numThreads, bool isDebug, string decodingMethod, List<string> wavFiles, ref TimeSpan total_duration)
{
OnlineTransducer onlineTransducer = new OnlineTransducer();
onlineTransducer.EncoderFilename = encoder;
onlineTransducer.DecoderFilename = decoder;
onlineTransducer.JoinerFilename = joiner;
//test online
OnlineRecognizer<OnlineTransducer> onlineRecognizer = new OnlineRecognizer<OnlineTransducer>(
onlineTransducer,
tokens,
num_threads: numThreads,
debug: isDebug,
decoding_method: decodingMethod);
List<float[]> samplesList = new List<float[]>();
foreach (string wavFile in wavFiles)
{
TimeSpan duration = TimeSpan.Zero;
float[] samples = AudioHelper.GetFileSamples(wavFile, ref duration);
samplesList.Add(samples);
total_duration += duration;
}
TimeSpan start_time = new TimeSpan(DateTime.Now.Ticks);
List<OnlineStream> streams = onlineRecognizer.CreateStreams(samplesList);
onlineRecognizer.DecodeMultipleStreams(streams);
List<OnlineRecognizerResultEntity> results = onlineRecognizer.GetResults(streams);
foreach (OnlineRecognizerResultEntity result in results)
{
Console.WriteLine(result.text);
}
TimeSpan end_time = new TimeSpan(DateTime.Now.Ticks);
}
static Dictionary<string, string> GetDict(string[] args, string applicationBase, ref List<string> wavFiles)
{
Dictionary<string, string> argsDict = new Dictionary<string, string>();
foreach (string input in args)
{
string[] ss = input.Split("=");
if (ss.Length == 1)
{
if (!string.IsNullOrEmpty(ss[0]))
{
wavFiles.Add(Path.Combine(applicationBase, ss[0].Trim(new char[] { '-', '`', ' ' })));
}
}
else
{
argsDict.Add(ss[0].Trim(new char[] { '-', '`', ' ' }).Replace("-", "_"), ss[1].Trim(new char[] { '-', '`', ' ' }));
}
}
return argsDict;
}
}
\ No newline at end of file
#ProjectReference csharp-api
`<ProjectReference Include="..\SherpaOnnx\SherpaOnnx.csproj" />`
The location of the 'SherpaOnnx' file is ../sherpa-onnx/csharp-api.
This C # API is cross platform and you can compile it yourself in Windows, Mac OS, and Linux environments.
------------
Alternatively, install sherpaonnx through nuget.
#NuGet for sherpa-onnx
PM > Install-Package SherpaOnnxCsharp -Project sherpa-onnx
\ No newline at end of file
using NAudio.Wave;
using System;
using System.Collections.Generic;
using System.Diagnostics;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
/// <summary>
/// audio processing
/// Copyright (c) 2023 by manyeyes
/// </summary>
public class AudioHelper
{
public static float[] GetFileSamples(string wavFilePath, ref TimeSpan duration)
{
if (!File.Exists(wavFilePath))
{
Trace.Assert(File.Exists(wavFilePath), "file does not exist:" + wavFilePath);
return new float[1];
}
AudioFileReader _audioFileReader = new AudioFileReader(wavFilePath);
byte[] datas = new byte[_audioFileReader.Length];
_audioFileReader.Read(datas, 0, datas.Length);
duration = _audioFileReader.TotalTime;
float[] wavdata = new float[datas.Length / sizeof(float)];
Buffer.BlockCopy(datas, 0, wavdata, 0, datas.Length);
return wavdata;
}
public static List<float[]> GetChunkSamplesList(string wavFilePath, ref TimeSpan duration)
{
List<float[]> wavdatas = new List<float[]>();
if (!File.Exists(wavFilePath))
{
Trace.Assert(File.Exists(wavFilePath), "file does not exist:" + wavFilePath);
wavdatas.Add(new float[1]);
return wavdatas;
}
AudioFileReader _audioFileReader = new AudioFileReader(wavFilePath);
byte[] datas = new byte[_audioFileReader.Length];
int chunkSize = 16000;// datas.Length / sizeof(float);
int chunkNum = (int)Math.Ceiling((double)datas.Length / chunkSize);
for (int i = 0; i < chunkNum; i++)
{
int offset = 0;
int dataCount = 0;
if (Math.Abs(datas.Length - i * chunkSize) > chunkSize)
{
offset = i * chunkSize;
dataCount = chunkSize;
}
else
{
offset = i * chunkSize;
dataCount = datas.Length - i * chunkSize;
}
_audioFileReader.Read(datas, offset, dataCount);
duration += _audioFileReader.TotalTime;
float[] wavdata = new float[chunkSize / sizeof(float)];
Buffer.BlockCopy(datas, offset, wavdata, 0, dataCount);
wavdatas.Add(wavdata);
}
return wavdatas;
}
}
# top-most EditorConfig file
root = true
# Don't use tabs for indentation.
[*]
indent_style = space
# Code files
[*.{cs,csx,vb,vbx}]
indent_size = 2
insert_final_newline = true
charset = utf-8-bom
end_of_line = crlf
... ...
// Copyright (c) 2023 Xiaomi Corporation
// Copyright (c) 2023 by manyeyes
//
// This file shows how to use a non-streaming model to decode files
// Please refer to
// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
// to download non-streaming models
using CommandLine.Text;
using CommandLine;
using SherpaOnnx;
using System.Collections.Generic;
using System;
class OfflineDecodeFiles
{
class Options
{
[Option(Required = false, HelpText = "Path to tokens.txt")]
public string Tokens { get; set; }
[Option(Required = false, HelpText = "Path to encoder.onnx. Used only for transducer models")]
public string Encoder { get; set; }
[Option(Required = false, HelpText = "Path to decoder.onnx. Used only for transducer models")]
public string Decoder { get; set; }
[Option(Required = false, HelpText = "Path to joiner.onnx. Used only for transducer models")]
public string Joiner { get; set; }
[Option(Required = false, HelpText = "Path to model.onnx. Used only for paraformer models")]
public string Paraformer { get; set; }
[Option("nemo-ctc", Required = false, HelpText = "Path to model.onnx. Used only for NeMo CTC models")]
public string NeMoCtc { get; set; }
[Option("num-threads", Required = false, Default = 1, HelpText = "Number of threads for computation")]
public int NumThreads { get; set; }
[Option("decoding-method", Required = false, Default = "greedy_search",
HelpText = "Valid decoding methods are: greedy_search, modified_beam_search")]
public string DecodingMethod { get; set; }
[Option("max-active-paths", Required = false, Default = 4,
HelpText = @"Used only when --decoding--method is modified_beam_search.
It specifies number of active paths to keep during the search")]
public int MaxActivePaths { get; set; }
[Option("files", Required = true, HelpText = "Audio files for decoding")]
public IEnumerable<string> Files { get; set; }
}
static void Main(string[] args)
{
var parser = new CommandLine.Parser(with => with.HelpWriter = null);
var parserResult = parser.ParseArguments<Options>(args);
parserResult
.WithParsed<Options>(options => Run(options))
.WithNotParsed(errs => DisplayHelp(parserResult, errs));
}
private static void DisplayHelp<T>(ParserResult<T> result, IEnumerable<Error> errs)
{
string usage = @"
# Zipformer
dotnet run \
--tokens=./sherpa-onnx-zipformer-en-2023-04-01/tokens.txt \
--encoder=./sherpa-onnx-zipformer-en-2023-04-01/encoder-epoch-99-avg-1.onnx \
--decoder=./sherpa-onnx-zipformer-en-2023-04-01/decoder-epoch-99-avg-1.onnx \
--joiner=./sherpa-onnx-zipformer-en-2023-04-01/joiner-epoch-99-avg-1.onnx \
--files ./sherpa-onnx-zipformer-en-2023-04-01/test_wavs/0.wav \
./sherpa-onnx-zipformer-en-2023-04-01/test_wavs/1.wav \
./sherpa-onnx-zipformer-en-2023-04-01/test_wavs/8k.wav
Please refer to
https://k2-fsa.github.io/sherpa/onnx/pretrained_models/offline-transducer/index.html
to download pre-trained non-streaming zipformer models.
# Paraformer
dotnet run \
--tokens=./sherpa-onnx-paraformer-zh-2023-03-28/tokens.txt \
--paraformer=./sherpa-onnx-paraformer-zh-2023-03-28/model.onnx \
--files ./sherpa-onnx-zipformer-en-2023-04-01/test_wavs/0.wav \
./sherpa-onnx-paraformer-zh-2023-03-28/test_wavs/0.wav \
./sherpa-onnx-paraformer-zh-2023-03-28/test_wavs/1.wav \
./sherpa-onnx-paraformer-zh-2023-03-28/test_wavs/2.wav \
./sherpa-onnx-paraformer-zh-2023-03-28/test_wavs/8k.wav
Please refer to
https://k2-fsa.github.io/sherpa/onnx/pretrained_models/offline-paraformer/index.html
to download pre-trained paraformer models
# NeMo CTC
dotnet run \
--tokens=./sherpa-onnx-nemo-ctc-en-conformer-medium/tokens.txt \
--nemo-ctc=./sherpa-onnx-nemo-ctc-en-conformer-medium/model.onnx \
--num-threads=1 \
--files ./sherpa-onnx-nemo-ctc-en-conformer-medium/test_wavs/0.wav \
./sherpa-onnx-nemo-ctc-en-conformer-medium/test_wavs/1.wav \
./sherpa-onnx-nemo-ctc-en-conformer-medium/test_wavs/8k.wav
Please refer to
https://k2-fsa.github.io/sherpa/onnx/pretrained_models/offline-ctc/index.html
to download pre-trained paraformer models
";
var helpText = HelpText.AutoBuild(result, h =>
{
h.AdditionalNewLineAfterOption = false;
h.Heading = usage;
h.Copyright = "Copyright (c) 2023 Xiaomi Corporation";
return HelpText.DefaultParsingErrorsHandler(result, h);
}, e => e);
Console.WriteLine(helpText);
}
private static void Run(Options options)
{
OfflineRecognizerConfig config = new OfflineRecognizerConfig();
config.ModelConfig.Tokens = options.Tokens;
if (!String.IsNullOrEmpty(options.Encoder))
{
// this is a transducer model
config.ModelConfig.Transducer.Encoder = options.Encoder;
config.ModelConfig.Transducer.Decoder = options.Decoder;
config.ModelConfig.Transducer.Joiner = options.Joiner;
}
else if (!String.IsNullOrEmpty(options.Paraformer))
{
config.ModelConfig.Paraformer.Model = options.Paraformer;
}
else if (!String.IsNullOrEmpty(options.NeMoCtc))
{
config.ModelConfig.NeMoCtc.Model = options.NeMoCtc;
}
else
{
Console.WriteLine("Please provide a model");
return;
}
config.DecodingMethod = options.DecodingMethod;
config.MaxActivePaths = options.MaxActivePaths;
config.ModelConfig.Debug = 0;
OfflineRecognizer recognizer = new OfflineRecognizer(config);
string[] files = options.Files.ToArray();
// We create a separate stream for each file
List<OfflineStream> streams = new List<OfflineStream>();
streams.EnsureCapacity(files.Length);
for (int i = 0; i != files.Length; ++i)
{
OfflineStream s = recognizer.CreateStream();
WaveReader waveReader = new WaveReader(files[i]);
s.AcceptWaveform(waveReader.SampleRate, waveReader.Samples);
streams.Add(s);
}
recognizer.Decode(streams);
// display results
for (int i = 0; i != files.Length; ++i)
{
var text = streams[i].Result.Text;
Console.WriteLine("--------------------");
Console.WriteLine(files[i]);
Console.WriteLine(text);
}
Console.WriteLine("--------------------");
}
}
... ...
../online-decode-files/WaveReader.cs
\ No newline at end of file
... ...
<Project Sdk="Microsoft.NET.Sdk">
<PropertyGroup>
<OutputType>Exe</OutputType>
<TargetFramework>net6.0</TargetFramework>
<RootNamespace>sherpa_onnx</RootNamespace>
<ImplicitUsings>enable</ImplicitUsings>
<Nullable>enable</Nullable>
<StartupObject>OnlineDecodeFiles</StartupObject>
</PropertyGroup>
<ItemGroup>
<PackageReference Include="NAudio" Version="2.1.0" />
</ItemGroup>
<ItemGroup>
<ProjectReference Include="..\SherpaOnnx\SherpaOnnx.csproj" />
</ItemGroup>
</Project>
<Project Sdk="Microsoft.NET.Sdk">
<PropertyGroup>
<OutputType>Exe</OutputType>
<TargetFramework>net6.0</TargetFramework>
<RootNamespace>offline_decode_files</RootNamespace>
<ImplicitUsings>enable</ImplicitUsings>
<Nullable>enable</Nullable>
</PropertyGroup>
<ItemGroup>
<PackageReference Include="CommandLineParser" Version="2.9.1" />
<PackageReference Include="org.k2fsa.sherpa.onnx" Version="*" />
</ItemGroup>
</Project>
... ...
#!/usr/bin/env bash
if [ ! -d ./sherpa-onnx-nemo-ctc-en-conformer-medium ]; then
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/csukuangfj/sherpa-onnx-nemo-ctc-en-conformer-medium
cd sherpa-onnx-nemo-ctc-en-conformer-medium
git lfs pull --include "*.onnx"
cd ..
fi
dotnet run \
--tokens=./sherpa-onnx-nemo-ctc-en-conformer-medium/tokens.txt \
--nemo-ctc=./sherpa-onnx-nemo-ctc-en-conformer-medium/model.onnx \
--num-threads=1 \
--files ./sherpa-onnx-nemo-ctc-en-conformer-medium/test_wavs/0.wav \
./sherpa-onnx-nemo-ctc-en-conformer-medium/test_wavs/1.wav \
./sherpa-onnx-nemo-ctc-en-conformer-medium/test_wavs/8k.wav
... ...
#!/usr/bin/env bash
if [ ! -d ./sherpa-onnx-paraformer-zh-2023-03-28 ]; then
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/csukuangfj/sherpa-onnx-paraformer-zh-2023-03-28
cd sherpa-onnx-paraformer-zh-2023-03-28
git lfs pull --include "*.onnx"
cd ..
fi
dotnet run \
--tokens=./sherpa-onnx-paraformer-zh-2023-03-28/tokens.txt \
--paraformer=./sherpa-onnx-paraformer-zh-2023-03-28/model.onnx \
--num-threads=2 \
--files ./sherpa-onnx-paraformer-zh-2023-03-28/test_wavs/0.wav \
./sherpa-onnx-paraformer-zh-2023-03-28/test_wavs/1.wav \
./sherpa-onnx-paraformer-zh-2023-03-28/test_wavs/2.wav \
./sherpa-onnx-paraformer-zh-2023-03-28/test_wavs/8k.wav
... ...
#!/usr/bin/env bash
#
if [ ! -d ./sherpa-onnx-zipformer-en-2023-04-01 ]; then
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/csukuangfj/sherpa-onnx-zipformer-en-2023-04-01
cd sherpa-onnx-zipformer-en-2023-04-01
git lfs pull --include "*.onnx"
cd ..
fi
dotnet run \
--tokens=./sherpa-onnx-zipformer-en-2023-04-01/tokens.txt \
--encoder=./sherpa-onnx-zipformer-en-2023-04-01/encoder-epoch-99-avg-1.onnx \
--decoder=./sherpa-onnx-zipformer-en-2023-04-01/decoder-epoch-99-avg-1.onnx \
--joiner=./sherpa-onnx-zipformer-en-2023-04-01/joiner-epoch-99-avg-1.onnx \
--num-threads=2 \
--decoding-method=modified_beam_search \
--files ./sherpa-onnx-zipformer-en-2023-04-01/test_wavs/0.wav \
./sherpa-onnx-zipformer-en-2023-04-01/test_wavs/1.wav \
./sherpa-onnx-zipformer-en-2023-04-01/test_wavs/8k.wav
... ...
// Copyright (c) 2023 Xiaomi Corporation
// Copyright (c) 2023 by manyeyes
//
// This file shows how to use a streaming model to decode files
// Please refer to
// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/online-transducer/zipformer-transducer-models.html
// to download streaming models
using CommandLine.Text;
using CommandLine;
using SherpaOnnx;
using System.Collections.Generic;
using System.Linq;
using System;
class OnlineDecodeFiles
{
class Options
{
[Option(Required = true, HelpText = "Path to tokens.txt")]
public string Tokens { get; set; }
[Option(Required = true, HelpText = "Path to encoder.onnx")]
public string Encoder { get; set; }
[Option(Required = true, HelpText = "Path to decoder.onnx")]
public string Decoder { get; set; }
[Option(Required = true, HelpText = "Path to joiner.onnx")]
public string Joiner { get; set; }
[Option("num-threads", Required = false, Default = 1, HelpText = "Number of threads for computation")]
public int NumThreads { get; set; }
[Option("decoding-method", Required = false, Default = "greedy_search",
HelpText = "Valid decoding methods are: greedy_search, modified_beam_search")]
public string DecodingMethod { get; set; }
[Option(Required = false, Default = false, HelpText = "True to show model info during loading")]
public bool Debug { get; set; }
[Option("sample-rate", Required = false, Default = 16000, HelpText = "Sample rate of the data used to train the model")]
public int SampleRate { get; set; }
[Option("max-active-paths", Required = false, Default = 4,
HelpText = @"Used only when --decoding--method is modified_beam_search.
It specifies number of active paths to keep during the search")]
public int MaxActivePaths { get; set; }
[Option("enable-endpoint", Required = false, Default = false,
HelpText = "True to enable endpoint detection.")]
public bool EnableEndpoint { get; set; }
[Option("rule1-min-trailing-silence", Required = false, Default = 2.4F,
HelpText = @"An endpoint is detected if trailing silence in seconds is
larger than this value even if nothing has been decoded. Used only when --enable-endpoint is true.")]
public float Rule1MinTrailingSilence { get; set; }
[Option("rule2-min-trailing-silence", Required = false, Default = 1.2F,
HelpText = @"An endpoint is detected if trailing silence in seconds is
larger than this value after something that is not blank has been decoded. Used
only when --enable-endpoint is true.")]
public float Rule2MinTrailingSilence { get; set; }
[Option("rule3-min-utterance-length", Required = false, Default = 20.0F,
HelpText = @"An endpoint is detected if the utterance in seconds is
larger than this value. Used only when --enable-endpoint is true.")]
public float Rule3MinUtteranceLength { get; set; }
[Option("files", Required = true, HelpText = "Audio files for decoding")]
public IEnumerable<string> Files { get; set; }
}
static void Main(string[] args)
{
var parser = new CommandLine.Parser(with => with.HelpWriter = null);
var parserResult = parser.ParseArguments<Options>(args);
parserResult
.WithParsed<Options>(options => Run(options))
.WithNotParsed(errs => DisplayHelp(parserResult, errs));
}
private static void DisplayHelp<T>(ParserResult<T> result, IEnumerable<Error> errs)
{
string usage = @"
dotnet run \
--tokens=./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/tokens.txt \
--encoder=./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/encoder-epoch-99-avg-1.onnx \
--decoder=./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/decoder-epoch-99-avg-1.onnx \
--joiner=./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/joiner-epoch-99-avg-1.onnx \
--num-threads=2 \
--decoding-method=modified_beam_search \
--debug=false \
./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/test_wavs/0.wav \
./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/test_wavs/1.wav
Please refer to
https://k2-fsa.github.io/sherpa/onnx/pretrained_models/online-transducer/index.html
to download pre-trained streaming models.
";
var helpText = HelpText.AutoBuild(result, h =>
{
h.AdditionalNewLineAfterOption = false;
h.Heading = usage;
h.Copyright = "Copyright (c) 2023 Xiaomi Corporation";
return HelpText.DefaultParsingErrorsHandler(result, h);
}, e => e);
Console.WriteLine(helpText);
}
private static void Run(Options options)
{
OnlineRecognizerConfig config = new OnlineRecognizerConfig();
config.FeatConfig.SampleRate = options.SampleRate;
// All models from icefall using feature dim 80.
// You can change it if your model has a different feature dim.
config.FeatConfig.FeatureDim = 80;
config.TransducerModelConfig.Encoder = options.Encoder;
config.TransducerModelConfig.Decoder = options.Decoder;
config.TransducerModelConfig.Joiner = options.Joiner;
config.TransducerModelConfig.Tokens = options.Tokens;
config.TransducerModelConfig.NumThreads = options.NumThreads;
config.TransducerModelConfig.Debug = options.Debug ? 1 : 0;
config.DecodingMethod = options.DecodingMethod;
config.MaxActivePaths = options.MaxActivePaths;
config.EnableEndpoint = options.EnableEndpoint ? 1 : 0;
config.Rule1MinTrailingSilence = options.Rule1MinTrailingSilence;
config.Rule2MinTrailingSilence = options.Rule2MinTrailingSilence;
config.Rule3MinUtteranceLength = options.Rule3MinUtteranceLength;
OnlineRecognizer recognizer = new OnlineRecognizer(config);
string[] files = options.Files.ToArray();
// We create a separate stream for each file
List<OnlineStream> streams = new List<OnlineStream>();
streams.EnsureCapacity(files.Length);
for (int i = 0; i != files.Length; ++i)
{
OnlineStream s = recognizer.CreateStream();
WaveReader waveReader = new WaveReader(files[i]);
s.AcceptWaveform(waveReader.SampleRate, waveReader.Samples);
float[] tailPadding = new float[(int)(waveReader.SampleRate * 0.3)];
s.AcceptWaveform(waveReader.SampleRate, tailPadding);
s.InputFinished();
streams.Add(s);
}
while (true)
{
var readyStreams = streams.Where(s => recognizer.IsReady(s));
if (!readyStreams.Any())
{
break;
}
recognizer.Decode(readyStreams);
}
// display results
for (int i = 0; i != files.Length; ++i)
{
var text = recognizer.GetResult(streams[i]).Text;
Console.WriteLine("--------------------");
Console.WriteLine(files[i]);
Console.WriteLine(text);
}
Console.WriteLine("--------------------");
}
}
... ...
// Copyright (c) 2023 Xiaomi Corporation (authors: Fangjun Kuang)
using System;
using System.IO;
using System.Runtime.InteropServices;
namespace SherpaOnnx
{
[StructLayout(LayoutKind.Sequential)]
public struct WaveHeader
{
public Int32 ChunkID;
public Int32 ChunkSize;
public Int32 Format;
public Int32 SubChunk1ID;
public Int32 SubChunk1Size;
public Int16 AudioFormat;
public Int16 NumChannels;
public Int32 SampleRate;
public Int32 ByteRate;
public Int16 BlockAlign;
public Int16 BitsPerSample;
public Int32 SubChunk2ID;
public Int32 SubChunk2Size;
public bool Validate()
{
if (ChunkID != 0x46464952)
{
Console.WriteLine($"Invalid chunk ID: 0x{ChunkID:X}. Expect 0x46464952");
return false;
}
// E V A W
if (Format != 0x45564157)
{
Console.WriteLine($"Invalid format: 0x{Format:X}. Expect 0x45564157");
return false;
}
// t m f
if (SubChunk1ID != 0x20746d66)
{
Console.WriteLine($"Invalid SubChunk1ID: 0x{SubChunk1ID:X}. Expect 0x20746d66");
return false;
}
if (SubChunk1Size != 16)
{
Console.WriteLine($"Invalid SubChunk1Size: {SubChunk1Size}. Expect 16");
return false;
}
if (AudioFormat != 1)
{
Console.WriteLine($"Invalid AudioFormat: {AudioFormat}. Expect 1");
return false;
}
if (NumChannels != 1)
{
Console.WriteLine($"Invalid NumChannels: {NumChannels}. Expect 1");
return false;
}
if (ByteRate != (SampleRate * NumChannels * BitsPerSample / 8))
{
Console.WriteLine($"Invalid byte rate: {ByteRate}.");
return false;
}
if (BlockAlign != (NumChannels * BitsPerSample / 8))
{
Console.WriteLine($"Invalid block align: {ByteRate}.");
return false;
}
if (BitsPerSample != 16)
{ // we support only 16 bits per sample
Console.WriteLine($"Invalid bits per sample: {BitsPerSample}. Expect 16");
return false;
}
return true;
}
}
// It supports only 16-bit, single channel WAVE format.
// The sample rate can be any value.
public class WaveReader
{
public WaveReader(String fileName)
{
if (!File.Exists(fileName))
{
throw new ApplicationException($"{fileName} does not exist!");
}
using (var stream = File.Open(fileName, FileMode.Open))
{
using (var reader = new BinaryReader(stream))
{
_header = ReadHeader(reader);
if (!_header.Validate())
{
throw new ApplicationException($"Invalid wave file ${fileName}");
}
SkipMetaData(reader);
// now read samples
// _header.SubChunk2Size contains number of bytes in total.
// we assume each sample is of type int16
byte[] buffer = reader.ReadBytes(_header.SubChunk2Size);
short[] samples_int16 = new short[_header.SubChunk2Size / 2];
Buffer.BlockCopy(buffer, 0, samples_int16, 0, buffer.Length);
_samples = new float[samples_int16.Length];
for (var i = 0; i < samples_int16.Length; ++i)
{
_samples[i] = samples_int16[i] / 32768.0F;
}
}
}
}
private static WaveHeader ReadHeader(BinaryReader reader)
{
byte[] bytes = reader.ReadBytes(Marshal.SizeOf(typeof(WaveHeader)));
GCHandle handle = GCHandle.Alloc(bytes, GCHandleType.Pinned);
WaveHeader header = (WaveHeader)Marshal.PtrToStructure(handle.AddrOfPinnedObject(), typeof(WaveHeader))!;
handle.Free();
return header;
}
private void SkipMetaData(BinaryReader reader)
{
var bs = reader.BaseStream;
Int32 subChunk2ID = _header.SubChunk2ID;
Int32 subChunk2Size = _header.SubChunk2Size;
while (bs.Position != bs.Length && subChunk2ID != 0x61746164)
{
bs.Seek(subChunk2Size, SeekOrigin.Current);
subChunk2ID = reader.ReadInt32();
subChunk2Size = reader.ReadInt32();
}
_header.SubChunk2ID = subChunk2ID;
_header.SubChunk2Size = subChunk2Size;
}
private WaveHeader _header;
// Samples are normalized to the range [-1, 1]
private float[] _samples;
public int SampleRate => _header.SampleRate;
public float[] Samples => _samples;
public static void Test(String fileName)
{
WaveReader reader = new WaveReader(fileName);
Console.WriteLine($"samples length: {reader.Samples.Length}");
Console.WriteLine($"samples rate: {reader.SampleRate}");
}
}
}
... ...
<Project Sdk="Microsoft.NET.Sdk">
<PropertyGroup>
<OutputType>Exe</OutputType>
<TargetFramework>net6.0</TargetFramework>
<RootNamespace>online_decode_files</RootNamespace>
<ImplicitUsings>enable</ImplicitUsings>
<Nullable>enable</Nullable>
<AllowUnsafeBlocks>true</AllowUnsafeBlocks>
</PropertyGroup>
<ItemGroup>
<PackageReference Include="CommandLineParser" Version="2.9.1" />
<PackageReference Include="org.k2fsa.sherpa.onnx" Version="*" />
</ItemGroup>
</Project>
... ...
#!/usr/bin/env bash
# Please refer to
# https://k2-fsa.github.io/sherpa/onnx/pretrained_models/online-transducer/zipformer-transducer-models.html#csukuangfj-sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20-bilingual-chinese-english
# to download the model files
if [ ! -d ./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20 ]; then
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/csukuangfj/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20
cd sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20
git lfs pull --include "*.onnx"
cd ..
fi
dotnet run -c Release \
--tokens ./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/tokens.txt \
--encoder ./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/encoder-epoch-99-avg-1.int8.onnx \
--decoder ./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/decoder-epoch-99-avg-1.int8.onnx \
--joiner ./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/joiner-epoch-99-avg-1.int8.onnx \
--decoding-method greedy_search \
--files ./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/test_wavs/1.wav \
./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/test_wavs/0.wav \
... ...

Microsoft Visual Studio Solution File, Format Version 12.00
# Visual Studio Version 17
VisualStudioVersion = 17.0.31903.59
MinimumVisualStudioVersion = 10.0.40219.1
Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "online-decode-files", "online-decode-files\online-decode-files.csproj", "{45307474-BECB-4ABE-9388-D01D55A1A9BE}"
EndProject
Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "offline-decode-files", "offline-decode-files\offline-decode-files.csproj", "{2DAB152C-9E24-47A0-9DB0-781297ECE458}"
EndProject
Global
GlobalSection(SolutionConfigurationPlatforms) = preSolution
Debug|Any CPU = Debug|Any CPU
Release|Any CPU = Release|Any CPU
EndGlobalSection
GlobalSection(SolutionProperties) = preSolution
HideSolutionNode = FALSE
EndGlobalSection
GlobalSection(ProjectConfigurationPlatforms) = postSolution
{45307474-BECB-4ABE-9388-D01D55A1A9BE}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
{45307474-BECB-4ABE-9388-D01D55A1A9BE}.Debug|Any CPU.Build.0 = Debug|Any CPU
{45307474-BECB-4ABE-9388-D01D55A1A9BE}.Release|Any CPU.ActiveCfg = Release|Any CPU
{45307474-BECB-4ABE-9388-D01D55A1A9BE}.Release|Any CPU.Build.0 = Release|Any CPU
{2DAB152C-9E24-47A0-9DB0-781297ECE458}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
{2DAB152C-9E24-47A0-9DB0-781297ECE458}.Debug|Any CPU.Build.0 = Debug|Any CPU
{2DAB152C-9E24-47A0-9DB0-781297ECE458}.Release|Any CPU.ActiveCfg = Release|Any CPU
{2DAB152C-9E24-47A0-9DB0-781297ECE458}.Release|Any CPU.Build.0 = Release|Any CPU
EndGlobalSection
EndGlobal
... ...
all
macos
linux
windows
packages
... ...
# Introduction
[sherpa-onnx](https://github.com/k2-fsa/sherpa-onnx) is an open-source
real-time speech recognition toolkit developed
by the Next-gen Kaldi team.
It supports streaming recognition on a variety of
platforms such as Android, iOS, Raspberry, Linux, Windows, macOS, etc.
It does not require Internet connection during recognition.
See the documentation https://k2-fsa.github.io/sherpa/onnx/index.html
for details.
Please see
https://github.com/k2-fsa/sherpa-onnx/tree/dot-net/dotnet-examples
for how to use C# APIs of this package.
... ...
#!/usr/bin/env python3
# Copyright (c) 2023 Xiaomi Corporation
import argparse
import re
from pathlib import Path
import jinja2
SHERPA_ONNX_DIR = Path(__file__).resolve().parent.parent.parent
def get_version():
cmake_file = SHERPA_ONNX_DIR / "CMakeLists.txt"
with open(cmake_file) as f:
content = f.read()
version = re.search(r"set\(SHERPA_ONNX_VERSION (.*)\)", content).group(1)
return version.strip('"')
def read_proj_file(filename):
with open(filename) as f:
return f.read()
def get_dict():
version = get_version()
return {
"version": get_version(),
}
def process_linux(s):
libs = [
"libkaldi-native-fbank-core.so",
"libonnxruntime.so.1.14.0",
"libsherpa-onnx-c-api.so",
"libsherpa-onnx-core.so",
]
prefix = f"{SHERPA_ONNX_DIR}/linux/sherpa_onnx/lib/"
libs = [prefix + lib for lib in libs]
libs = "\n ;".join(libs)
d = get_dict()
d["dotnet_rid"] = "linux-x64"
d["libs"] = libs
environment = jinja2.Environment()
template = environment.from_string(s)
s = template.render(**d)
with open("./linux/sherpa-onnx.runtime.csproj", "w") as f:
f.write(s)
def process_macos(s):
libs = [
"libkaldi-native-fbank-core.dylib",
"libonnxruntime.1.14.0.dylib",
"libsherpa-onnx-c-api.dylib",
"libsherpa-onnx-core.dylib",
]
prefix = f"{SHERPA_ONNX_DIR}/macos/sherpa_onnx/lib/"
libs = [prefix + lib for lib in libs]
libs = "\n ;".join(libs)
d = get_dict()
d["dotnet_rid"] = "osx-x64"
d["libs"] = libs
environment = jinja2.Environment()
template = environment.from_string(s)
s = template.render(**d)
with open("./macos/sherpa-onnx.runtime.csproj", "w") as f:
f.write(s)
def process_windows(s):
libs = [
"kaldi-native-fbank-core.dll",
"onnxruntime.dll",
"sherpa-onnx-c-api.dll",
"sherpa-onnx-core.dll",
]
prefix = f"{SHERPA_ONNX_DIR}/windows/sherpa_onnx/lib/"
libs = [prefix + lib for lib in libs]
libs = "\n ;".join(libs)
d = get_dict()
d["dotnet_rid"] = "win-x64"
d["libs"] = libs
environment = jinja2.Environment()
template = environment.from_string(s)
s = template.render(**d)
with open("./windows/sherpa-onnx.runtime.csproj", "w") as f:
f.write(s)
def main():
s = read_proj_file("./sherpa-onnx.csproj.runtime.in")
process_macos(s)
process_linux(s)
process_windows(s)
s = read_proj_file("./sherpa-onnx.csproj.in")
d = get_dict()
d["packages_dir"] = str(SHERPA_ONNX_DIR / "scripts/dotnet/packages")
environment = jinja2.Environment()
template = environment.from_string(s)
s = template.render(**d)
with open("./all/sherpa-onnx.csproj", "w") as f:
f.write(s)
if __name__ == "__main__":
main()
... ...
/// Copyright (c) 2023 Xiaomi Corporation (authors: Fangjun Kuang)
/// Copyright (c) 2023 by manyeyes
using System.Linq;
using System.Collections.Generic;
using System.Runtime.InteropServices;
using System;
namespace SherpaOnnx
{
[StructLayout(LayoutKind.Sequential)]
public struct OfflineTransducerModelConfig
{
public OfflineTransducerModelConfig()
{
Encoder = "";
Decoder = "";
Joiner = "";
}
[MarshalAs(UnmanagedType.LPStr)]
public string Encoder;
[MarshalAs(UnmanagedType.LPStr)]
public string Decoder;
[MarshalAs(UnmanagedType.LPStr)]
public string Joiner;
}
[StructLayout(LayoutKind.Sequential)]
public struct OfflineParaformerModelConfig
{
public OfflineParaformerModelConfig()
{
Model = "";
}
[MarshalAs(UnmanagedType.LPStr)]
public string Model;
}
[StructLayout(LayoutKind.Sequential)]
public struct OfflineNemoEncDecCtcModelConfig
{
public OfflineNemoEncDecCtcModelConfig()
{
Model = "";
}
[MarshalAs(UnmanagedType.LPStr)]
public string Model;
}
[StructLayout(LayoutKind.Sequential)]
public struct OfflineLMConfig
{
public OfflineLMConfig()
{
Model = "";
Scale = 0.5F;
}
[MarshalAs(UnmanagedType.LPStr)]
public string Model;
public float Scale;
}
[StructLayout(LayoutKind.Sequential)]
public struct OfflineModelConfig
{
public OfflineModelConfig()
{
Transducer = new OfflineTransducerModelConfig();
Paraformer = new OfflineParaformerModelConfig();
NeMoCtc = new OfflineNemoEncDecCtcModelConfig();
Tokens = "";
NumThreads = 1;
Debug = 0;
}
public OfflineTransducerModelConfig Transducer;
public OfflineParaformerModelConfig Paraformer;
public OfflineNemoEncDecCtcModelConfig NeMoCtc;
[MarshalAs(UnmanagedType.LPStr)]
public string Tokens;
public int NumThreads;
public int Debug;
}
[StructLayout(LayoutKind.Sequential)]
public struct OfflineRecognizerConfig
{
public OfflineRecognizerConfig()
{
FeatConfig = new FeatureConfig();
ModelConfig = new OfflineModelConfig();
LmConfig = new OfflineLMConfig();
DecodingMethod = "greedy_search";
MaxActivePaths = 4;
}
public FeatureConfig FeatConfig;
public OfflineModelConfig ModelConfig;
public OfflineLMConfig LmConfig;
[MarshalAs(UnmanagedType.LPStr)]
public string DecodingMethod;
public int MaxActivePaths;
}
public class OfflineRecognizerResult
{
public OfflineRecognizerResult(IntPtr handle)
{
Impl impl = (Impl)Marshal.PtrToStructure(handle, typeof(Impl));
_text = Marshal.PtrToStringUTF8(impl.Text);
}
[StructLayout(LayoutKind.Sequential)]
struct Impl
{
public IntPtr Text;
}
private String _text;
public String Text => _text;
}
public class OfflineStream : IDisposable
{
public OfflineStream(IntPtr p)
{
_handle = new HandleRef(this, p);
}
public void AcceptWaveform(int sampleRate, float[] samples)
{
AcceptWaveform(Handle, sampleRate, samples, samples.Length);
}
public OfflineRecognizerResult Result
{
get
{
IntPtr h = GetResult(_handle.Handle);
OfflineRecognizerResult result = new OfflineRecognizerResult(h);
DestroyResult(h);
return result;
}
}
~OfflineStream()
{
Cleanup();
}
public void Dispose()
{
Cleanup();
// Prevent the object from being placed on the
// finalization queue
System.GC.SuppressFinalize(this);
}
private void Cleanup()
{
DestroyOfflineStream(Handle);
// Don't permit the handle to be used again.
_handle = new HandleRef(this, IntPtr.Zero);
}
private HandleRef _handle;
public IntPtr Handle => _handle.Handle;
[DllImport(Dll.Filename)]
private static extern void DestroyOfflineStream(IntPtr handle);
[DllImport(Dll.Filename, EntryPoint = "AcceptWaveformOffline")]
private static extern void AcceptWaveform(IntPtr handle, int sampleRate, float[] samples, int n);
[DllImport(Dll.Filename, EntryPoint = "GetOfflineStreamResult")]
private static extern IntPtr GetResult(IntPtr handle);
[DllImport(Dll.Filename, EntryPoint = "DestroyOfflineRecognizerResult")]
private static extern void DestroyResult(IntPtr handle);
}
public class OfflineRecognizer : IDisposable
{
public OfflineRecognizer(OfflineRecognizerConfig config)
{
IntPtr h = CreateOfflineRecognizer(ref config);
_handle = new HandleRef(this, h);
}
public OfflineStream CreateStream()
{
IntPtr p = CreateOfflineStream(_handle.Handle);
return new OfflineStream(p);
}
/// You have to ensure that IsReady(stream) returns true before
/// you call this method
public void Decode(OfflineStream stream)
{
Decode(_handle.Handle, stream.Handle);
}
// The caller should ensure all passed streams are ready for decoding.
public void Decode(IEnumerable<OfflineStream> streams)
{
IntPtr[] ptrs = streams.Select(s => s.Handle).ToArray();
Decode(_handle.Handle, ptrs, ptrs.Length);
}
public void Dispose()
{
Cleanup();
// Prevent the object from being placed on the
// finalization queue
System.GC.SuppressFinalize(this);
}
~OfflineRecognizer()
{
Cleanup();
}
private void Cleanup()
{
DestroyOfflineRecognizer(_handle.Handle);
// Don't permit the handle to be used again.
_handle = new HandleRef(this, IntPtr.Zero);
}
private HandleRef _handle;
[DllImport(Dll.Filename)]
private static extern IntPtr CreateOfflineRecognizer(ref OfflineRecognizerConfig config);
[DllImport(Dll.Filename)]
private static extern void DestroyOfflineRecognizer(IntPtr handle);
[DllImport(Dll.Filename)]
private static extern IntPtr CreateOfflineStream(IntPtr handle);
[DllImport(Dll.Filename, EntryPoint = "DecodeOfflineStream")]
private static extern void Decode(IntPtr handle, IntPtr stream);
[DllImport(Dll.Filename, EntryPoint = "DecodeMultipleOfflineStreams")]
private static extern void Decode(IntPtr handle, IntPtr[] streams, int n);
}
}
... ...
/// Copyright (c) 2023 Xiaomi Corporation (authors: Fangjun Kuang)
/// Copyright (c) 2023 by manyeyes
using System.Linq;
using System.Collections.Generic;
using System.Runtime.InteropServices;
using System;
namespace SherpaOnnx
{
internal static class Dll
{
public const string Filename = "sherpa-onnx-c-api";
}
[StructLayout(LayoutKind.Sequential)]
public struct OnlineTransducerModelConfig
{
public OnlineTransducerModelConfig()
{
Encoder = "";
Decoder = "";
Joiner = "";
Tokens = "";
NumThreads = 1;
Debug = 0;
}
[MarshalAs(UnmanagedType.LPStr)]
public string Encoder;
[MarshalAs(UnmanagedType.LPStr)]
public string Decoder;
[MarshalAs(UnmanagedType.LPStr)]
public string Joiner;
[MarshalAs(UnmanagedType.LPStr)]
public string Tokens;
/// Number of threads used to run the neural network model
public int NumThreads;
/// true to print debug information of the model
public int Debug;
}
/// It expects 16 kHz 16-bit single channel wave format.
[StructLayout(LayoutKind.Sequential)]
public struct FeatureConfig
{
public FeatureConfig()
{
SampleRate = 16000;
FeatureDim = 80;
}
/// Sample rate of the input data. MUST match the one expected
/// by the model. For instance, it should be 16000 for models provided
/// by us.
public int SampleRate;
/// Feature dimension of the model.
/// For instance, it should be 80 for models provided by us.
public int FeatureDim;
}
[StructLayout(LayoutKind.Sequential)]
public struct OnlineRecognizerConfig
{
public OnlineRecognizerConfig()
{
FeatConfig = new FeatureConfig();
TransducerModelConfig = new OnlineTransducerModelConfig();
DecodingMethod = "greedy_search";
MaxActivePaths = 4;
EnableEndpoint = 0;
Rule1MinTrailingSilence = 1.2F;
Rule2MinTrailingSilence = 2.4F;
Rule3MinUtteranceLength = 20.0F;
}
public FeatureConfig FeatConfig;
public OnlineTransducerModelConfig TransducerModelConfig;
[MarshalAs(UnmanagedType.LPStr)]
public string DecodingMethod;
/// Used only when decoding_method is modified_beam_search
/// Example value: 4
public int MaxActivePaths;
/// 0 to disable endpoint detection.
/// A non-zero value to enable endpoint detection.
public int EnableEndpoint;
/// An endpoint is detected if trailing silence in seconds is larger than
/// this value even if nothing has been decoded.
/// Used only when enable_endpoint is not 0.
public float Rule1MinTrailingSilence;
/// An endpoint is detected if trailing silence in seconds is larger than
/// this value after something that is not blank has been decoded.
/// Used only when enable_endpoint is not 0.
public float Rule2MinTrailingSilence;
/// An endpoint is detected if the utterance in seconds is larger than
/// this value.
/// Used only when enable_endpoint is not 0.
public float Rule3MinUtteranceLength;
}
public class OnlineRecognizerResult
{
public OnlineRecognizerResult(IntPtr handle)
{
Impl impl = (Impl)Marshal.PtrToStructure(handle, typeof(Impl));
_text = Marshal.PtrToStringUTF8(impl.Text);
}
[StructLayout(LayoutKind.Sequential)]
struct Impl
{
public IntPtr Text;
}
private String _text;
public String Text => _text;
}
public class OnlineStream : IDisposable
{
public OnlineStream(IntPtr p)
{
_handle = new HandleRef(this, p);
}
public void AcceptWaveform(int sampleRate, float[] samples)
{
AcceptWaveform(Handle, sampleRate, samples, samples.Length);
}
public void InputFinished()
{
InputFinished(Handle);
}
~OnlineStream()
{
Cleanup();
}
public void Dispose()
{
Cleanup();
// Prevent the object from being placed on the
// finalization queue
System.GC.SuppressFinalize(this);
}
private void Cleanup()
{
DestroyOnlineStream(Handle);
// Don't permit the handle to be used again.
_handle = new HandleRef(this, IntPtr.Zero);
}
private HandleRef _handle;
public IntPtr Handle => _handle.Handle;
[DllImport(Dll.Filename)]
private static extern void DestroyOnlineStream(IntPtr handle);
[DllImport(Dll.Filename)]
private static extern void AcceptWaveform(IntPtr handle, int sampleRate, float[] samples, int n);
[DllImport(Dll.Filename)]
private static extern void InputFinished(IntPtr handle);
}
// please see
// https://www.mono-project.com/docs/advanced/pinvoke/#gc-safe-pinvoke-code
// https://www.mono-project.com/docs/advanced/pinvoke/#properly-disposing-of-resources
public class OnlineRecognizer : IDisposable
{
public OnlineRecognizer(OnlineRecognizerConfig config)
{
IntPtr h = CreateOnlineRecognizer(ref config);
_handle = new HandleRef(this, h);
}
public OnlineStream CreateStream()
{
IntPtr p = CreateOnlineStream(_handle.Handle);
return new OnlineStream(p);
}
/// Return true if the passed stream is ready for decoding.
public bool IsReady(OnlineStream stream)
{
return IsReady(_handle.Handle, stream.Handle) != 0;
}
/// Return true if an endpoint is detected for this stream.
/// You probably need to invoke Reset(stream) when this method returns
/// true.
public bool IsEndpoint(OnlineStream stream)
{
return IsEndpoint(_handle.Handle, stream.Handle) != 0;
}
/// You have to ensure that IsReady(stream) returns true before
/// you call this method
public void Decode(OnlineStream stream)
{
Decode(_handle.Handle, stream.Handle);
}
// The caller should ensure all passed streams are ready for decoding.
public void Decode(IEnumerable<OnlineStream> streams)
{
IntPtr[] ptrs = streams.Select(s => s.Handle).ToArray();
Decode(_handle.Handle, ptrs, ptrs.Length);
}
public OnlineRecognizerResult GetResult(OnlineStream stream)
{
IntPtr h = GetResult(_handle.Handle, stream.Handle);
OnlineRecognizerResult result = new OnlineRecognizerResult(h);
DestroyResult(h);
return result;
}
/// When this method returns, IsEndpoint(stream) will return false.
public void Reset(OnlineStream stream)
{
Reset(_handle.Handle, stream.Handle);
}
public void Dispose()
{
Cleanup();
// Prevent the object from being placed on the
// finalization queue
System.GC.SuppressFinalize(this);
}
~OnlineRecognizer()
{
Cleanup();
}
private void Cleanup()
{
DestroyOnlineRecognizer(_handle.Handle);
// Don't permit the handle to be used again.
_handle = new HandleRef(this, IntPtr.Zero);
}
private HandleRef _handle;
[DllImport(Dll.Filename)]
private static extern IntPtr CreateOnlineRecognizer(ref OnlineRecognizerConfig config);
[DllImport(Dll.Filename)]
private static extern void DestroyOnlineRecognizer(IntPtr handle);
[DllImport(Dll.Filename)]
private static extern IntPtr CreateOnlineStream(IntPtr handle);
[DllImport(Dll.Filename, EntryPoint = "IsOnlineStreamReady")]
private static extern int IsReady(IntPtr handle, IntPtr stream);
[DllImport(Dll.Filename, EntryPoint = "DecodeOnlineStream")]
private static extern void Decode(IntPtr handle, IntPtr stream);
[DllImport(Dll.Filename, EntryPoint = "DecodeMultipleOnlineStreams")]
private static extern void Decode(IntPtr handle, IntPtr[] streams, int n);
[DllImport(Dll.Filename, EntryPoint = "GetOnlineStreamResult")]
private static extern IntPtr GetResult(IntPtr handle, IntPtr stream);
[DllImport(Dll.Filename, EntryPoint = "DestroyOnlineRecognizerResult")]
private static extern void DestroyResult(IntPtr result);
[DllImport(Dll.Filename)]
private static extern void Reset(IntPtr handle, IntPtr stream);
[DllImport(Dll.Filename)]
private static extern int IsEndpoint(IntPtr handle, IntPtr stream);
}
}
... ...
#!/usr/bin/env bash
# Copyright (c) 2023 Xiaomi Corporation
set -ex
mkdir -p macos linux windows all
cp ./online.cs all
cp ./offline.cs all
./generate.py
pushd linux
dotnet build -c Release
dotnet pack -c Release -o ../packages
popd
pushd macos
dotnet build -c Release
dotnet pack -c Release -o ../packages
popd
pushd windows
dotnet build -c Release
dotnet pack -c Release -o ../packages
popd
pushd all
dotnet build -c Release
dotnet pack -c Release -o ../packages
popd
ls -lh packages
... ...
<Project Sdk="Microsoft.NET.Sdk">
<PropertyGroup>
<PackageLicenseExpression>Apache-2.0</PackageLicenseExpression>
<PackageReadmeFile>README.md</PackageReadmeFile>
<OutputType>Library</OutputType>
<LangVersion>10.0</LangVersion>
<TargetFrameworks>netstandard2.1;netcoreapp3.1;net6.0;net7.0</TargetFrameworks>
<RuntimeIdentifiers>linux-x64;osx-x64;win-x64</RuntimeIdentifiers>
<AllowUnsafeBlocks>true</AllowUnsafeBlocks>
<AssemblyName>sherpa-onnx</AssemblyName>
<Version>{{ version }}</Version>
<PackageProjectUrl>https://github.com/k2-fsa/sherpa-onnx</PackageProjectUrl>
<RepositoryUrl>https://github.com/k2-fsa/sherpa-onnx</RepositoryUrl>
<PackageTags>speech recognition voice audio stt asr speech-to-text AI offline
privacy open-sourced next-gen-kaldi k2 kaldi2 sherpa-onnx</PackageTags>
<Authors>The Next-gen Kaldi development team</Authors>
<Owners>The Next-gen Kaldi development team</Owners>
<Company>Xiaomi Corporation</Company>
<Copyright>Copyright 2019-2023 Xiaomi Corporation</Copyright>
<Description>sherpa-onnx is an open-source real-time speech recognition toolkit developed
by the Next-gen Kaldi team. It supports streaming recognition on a variety of
platforms such as Android, iOS, Raspberry, Linux, Windows, macOS, etc.
It does not require Internet connection during recognition.
See the documentation https://k2-fsa.github.io/sherpa/onnx/index.html
for details.
</Description>
<!-- Pack Option -->
<Title>sherpa-onnx v{{ version }}</Title>
<PackageId>org.k2fsa.sherpa.onnx</PackageId>
<!-- Signing -->
<SignAssembly>false</SignAssembly>
<PublicSign>false</PublicSign>
<DelaySign>false</DelaySign>
</PropertyGroup>
<PropertyGroup>
<RestoreSources>{{ packages_dir }};$(RestoreSources);https://api.nuget.org/v3/index.json</RestoreSources>
</PropertyGroup>
<ItemGroup>
<None Include="../README.md" Pack="true" PackagePath="/"/>
</ItemGroup>
<ItemGroup>
<PackageReference Include="org.k2fsa.sherpa.onnx.runtime.linux-x64" Version="{{ version }}" />
<PackageReference Include="org.k2fsa.sherpa.onnx.runtime.osx-x64" Version="{{ version }}" />
<PackageReference Include="org.k2fsa.sherpa.onnx.runtime.win-x64" Version="{{ version }}" />
</ItemGroup>
</Project>
... ...
<Project Sdk="Microsoft.NET.Sdk">
<PropertyGroup>
<PackageLicenseExpression>Apache-2.0</PackageLicenseExpression>
<PackageReadmeFile>README.md</PackageReadmeFile>
<OutputType>Library</OutputType>
<TargetFrameworks>netstandard2.0;netcoreapp3.1;net6.0</TargetFrameworks>
<RuntimeIdentifier>{{ dotnet_rid }}</RuntimeIdentifier>
<AssemblyName>sherpa-onnx</AssemblyName>
<Version>{{ version }}</Version>
<PackageProjectUrl>https://github.com/k2-fsa/sherpa-onnx</PackageProjectUrl>
<RepositoryUrl>https://github.com/k2-fsa/sherpa-onnx</RepositoryUrl>
<PackageTags>speech recognition voice audio stt asr speech-to-text AI offline
privacy open-sourced next-gen-kaldi k2 kaldi2 sherpa-onnx</PackageTags>
<!-- Nuget Properties -->
<Description>.NET native {{ dotnet_rid }} wrapper for the sherpa-onnx project.
In general, you don't need to use this package directly.
Please use https://www.nuget.org/packages/org.k2fsa.sherpa.onnx instead
</Description>
<IncludeBuildOutput>false</IncludeBuildOutput>
<!-- Pack Option -->
<Title>sherpa-onnx {{ dotnet_rid }} v{{ version }}</Title>
<PackageId>org.k2fsa.sherpa.onnx.runtime.{{ dotnet_rid }}</PackageId>
<!-- Signing -->
<SignAssembly>false</SignAssembly>
<PublicSign>false</PublicSign>
<DelaySign>false</DelaySign>
</PropertyGroup>
<ItemGroup>
<None Include="../README.md" Pack="true" PackagePath="/"/>
</ItemGroup>
<ItemGroup>
<!-- Native library must be in native directory... -->
<!-- If project is built as a STATIC_LIBRARY (e.g. Windows) then we don't have to include it -->
<Content Include="
{{ libs }}
">
<PackagePath>runtimes/{{ dotnet_rid }}/native/%(Filename)%(Extension)</PackagePath>
<Pack>true</Pack>
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</Content>
</ItemGroup>
</Project>
... ...
... ... @@ -2,6 +2,11 @@ include_directories(${CMAKE_SOURCE_DIR})
add_library(sherpa-onnx-c-api c-api.cc)
target_link_libraries(sherpa-onnx-c-api sherpa-onnx-core)
if(BUILD_SHARED_LIBS)
target_compile_definitions(sherpa-onnx-c-api PRIVATE SHERPA_ONNX_BUILD_SHARED_LIBS=1)
target_compile_definitions(sherpa-onnx-c-api PRIVATE SHERPA_ONNX_BUILD_MAIN_LIB=1)
endif()
install(TARGETS sherpa-onnx-c-api DESTINATION lib)
install(FILES c-api.h
... ...
... ... @@ -10,10 +10,11 @@
#include <vector>
#include "sherpa-onnx/csrc/display.h"
#include "sherpa-onnx/csrc/offline-recognizer.h"
#include "sherpa-onnx/csrc/online-recognizer.h"
struct SherpaOnnxOnlineRecognizer {
sherpa_onnx::OnlineRecognizer *impl;
std::unique_ptr<sherpa_onnx::OnlineRecognizer> impl;
};
struct SherpaOnnxOnlineStream {
... ... @@ -56,14 +57,19 @@ SherpaOnnxOnlineRecognizer *CreateOnlineRecognizer(
recognizer_config.endpoint_config.rule3.min_utterance_length =
config->rule3_min_utterance_length;
if (config->model_config.debug) {
fprintf(stderr, "%s\n", recognizer_config.ToString().c_str());
}
SherpaOnnxOnlineRecognizer *recognizer = new SherpaOnnxOnlineRecognizer;
recognizer->impl = new sherpa_onnx::OnlineRecognizer(recognizer_config);
recognizer->impl =
std::make_unique<sherpa_onnx::OnlineRecognizer>(recognizer_config);
return recognizer;
}
void DestroyOnlineRecognizer(SherpaOnnxOnlineRecognizer *recognizer) {
delete recognizer->impl;
delete recognizer;
}
... ... @@ -144,3 +150,116 @@ void DestroyDisplay(SherpaOnnxDisplay *display) { delete display; }
void SherpaOnnxPrint(SherpaOnnxDisplay *display, int32_t idx, const char *s) {
display->impl->Print(idx, s);
}
// ============================================================
// For offline ASR (i.e., non-streaming ASR)
// ============================================================
//
struct SherpaOnnxOfflineRecognizer {
std::unique_ptr<sherpa_onnx::OfflineRecognizer> impl;
};
struct SherpaOnnxOfflineStream {
std::unique_ptr<sherpa_onnx::OfflineStream> impl;
explicit SherpaOnnxOfflineStream(
std::unique_ptr<sherpa_onnx::OfflineStream> p)
: impl(std::move(p)) {}
};
SherpaOnnxOfflineRecognizer *CreateOfflineRecognizer(
const SherpaOnnxOfflineRecognizerConfig *config) {
sherpa_onnx::OfflineRecognizerConfig recognizer_config;
recognizer_config.feat_config.sampling_rate = config->feat_config.sample_rate;
recognizer_config.feat_config.feature_dim = config->feat_config.feature_dim;
recognizer_config.model_config.transducer.encoder_filename =
config->model_config.transducer.encoder;
recognizer_config.model_config.transducer.decoder_filename =
config->model_config.transducer.decoder;
recognizer_config.model_config.transducer.joiner_filename =
config->model_config.transducer.joiner;
recognizer_config.model_config.paraformer.model =
config->model_config.paraformer.model;
recognizer_config.model_config.nemo_ctc.model =
config->model_config.nemo_ctc.model;
recognizer_config.model_config.tokens = config->model_config.tokens;
recognizer_config.model_config.num_threads = config->model_config.num_threads;
recognizer_config.model_config.debug = config->model_config.debug;
recognizer_config.lm_config.model = config->lm_config.model;
recognizer_config.lm_config.scale = config->lm_config.scale;
recognizer_config.decoding_method = config->decoding_method;
recognizer_config.max_active_paths = config->max_active_paths;
if (config->model_config.debug) {
fprintf(stderr, "%s\n", recognizer_config.ToString().c_str());
}
SherpaOnnxOfflineRecognizer *recognizer = new SherpaOnnxOfflineRecognizer;
recognizer->impl =
std::make_unique<sherpa_onnx::OfflineRecognizer>(recognizer_config);
return recognizer;
}
void DestroyOfflineRecognizer(SherpaOnnxOfflineRecognizer *recognizer) {
delete recognizer;
}
SherpaOnnxOfflineStream *CreateOfflineStream(
const SherpaOnnxOfflineRecognizer *recognizer) {
SherpaOnnxOfflineStream *stream =
new SherpaOnnxOfflineStream(recognizer->impl->CreateStream());
return stream;
}
void DestoryOfflineStream(SherpaOnnxOfflineStream *stream) { delete stream; }
void AcceptWaveformOffline(SherpaOnnxOfflineStream *stream, int32_t sample_rate,
const float *samples, int32_t n) {
stream->impl->AcceptWaveform(sample_rate, samples, n);
}
void DecodeOfflineStream(SherpaOnnxOfflineRecognizer *recognizer,
SherpaOnnxOfflineStream *stream) {
recognizer->impl->DecodeStream(stream->impl.get());
}
void DecodeMultipleOfflineStreams(SherpaOnnxOfflineRecognizer *recognizer,
SherpaOnnxOfflineStream **streams,
int32_t n) {
std::vector<sherpa_onnx::OfflineStream *> ss(n);
for (int32_t i = 0; i != n; ++i) {
ss[i] = streams[i]->impl.get();
}
recognizer->impl->DecodeStreams(ss.data(), n);
}
SherpaOnnxOfflineRecognizerResult *GetOfflineStreamResult(
SherpaOnnxOfflineStream *stream) {
const sherpa_onnx::OfflineRecognitionResult &result =
stream->impl->GetResult();
const auto &text = result.text;
auto r = new SherpaOnnxOfflineRecognizerResult;
r->text = new char[text.size() + 1];
std::copy(text.begin(), text.end(), const_cast<char *>(r->text));
const_cast<char *>(r->text)[text.size()] = 0;
return r;
}
void DestroyOfflineRecognizerResult(
const SherpaOnnxOfflineRecognizerResult *r) {
delete[] r->text;
delete r;
}
... ...
... ... @@ -18,12 +18,35 @@
extern "C" {
#endif
// See https://github.com/pytorch/pytorch/blob/main/c10/macros/Export.h
// We will set SHERPA_ONNX_BUILD_SHARED_LIBS and SHERPA_ONNX_BUILD_MAIN_LIB in
// CMakeLists.txt
#if defined(_WIN32)
#if defined(SHERPA_ONNX_BUILD_SHARED_LIBS)
#define SHERPA_ONNX_EXPORT __declspec(dllexport)
#define SHERPA_ONNX_IMPORT __declspec(dllimport)
#else
#define SHERPA_ONNX_EXPORT
#define SHERPA_ONNX_IMPORT
#endif
#else // WIN32
#define SHERPA_ONNX_EXPORT __attribute__((__visibility__("default")))
#define SHERPA_ONNX_IMPORT SHERPA_ONNX_EXPORT
#endif
#if defined(SHERPA_ONNX_BUILD_MAIN_LIB)
#define SHERPA_ONNX_API SHERPA_ONNX_EXPORT
#else
#define SHERPA_ONNX_API SHERPA_ONNX_IMPORT
#endif
/// Please refer to
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
/// to download pre-trained models. That is, you can find encoder-xxx.onnx
/// decoder-xxx.onnx, joiner-xxx.onnx, and tokens.txt for this struct
/// from there.
typedef struct SherpaOnnxOnlineTransducerModelConfig {
SHERPA_ONNX_API typedef struct SherpaOnnxOnlineTransducerModelConfig {
const char *encoder;
const char *decoder;
const char *joiner;
... ... @@ -33,7 +56,7 @@ typedef struct SherpaOnnxOnlineTransducerModelConfig {
} SherpaOnnxOnlineTransducerModelConfig;
/// It expects 16 kHz 16-bit single channel wave format.
typedef struct SherpaOnnxFeatureConfig {
SHERPA_ONNX_API typedef struct SherpaOnnxFeatureConfig {
/// Sample rate of the input data. MUST match the one expected
/// by the model. For instance, it should be 16000 for models provided
/// by us.
... ... @@ -44,7 +67,7 @@ typedef struct SherpaOnnxFeatureConfig {
int32_t feature_dim;
} SherpaOnnxFeatureConfig;
typedef struct SherpaOnnxOnlineRecognizerConfig {
SHERPA_ONNX_API typedef struct SherpaOnnxOnlineRecognizerConfig {
SherpaOnnxFeatureConfig feat_config;
SherpaOnnxOnlineTransducerModelConfig model_config;
... ... @@ -75,7 +98,7 @@ typedef struct SherpaOnnxOnlineRecognizerConfig {
float rule3_min_utterance_length;
} SherpaOnnxOnlineRecognizerConfig;
typedef struct SherpaOnnxOnlineRecognizerResult {
SHERPA_ONNX_API typedef struct SherpaOnnxOnlineRecognizerResult {
const char *text;
// TODO(fangjun): Add more fields
} SherpaOnnxOnlineRecognizerResult;
... ... @@ -83,32 +106,34 @@ typedef struct SherpaOnnxOnlineRecognizerResult {
/// Note: OnlineRecognizer here means StreamingRecognizer.
/// It does not need to access the Internet during recognition.
/// Everything is run locally.
typedef struct SherpaOnnxOnlineRecognizer SherpaOnnxOnlineRecognizer;
typedef struct SherpaOnnxOnlineStream SherpaOnnxOnlineStream;
SHERPA_ONNX_API typedef struct SherpaOnnxOnlineRecognizer
SherpaOnnxOnlineRecognizer;
SHERPA_ONNX_API typedef struct SherpaOnnxOnlineStream SherpaOnnxOnlineStream;
/// @param config Config for the recongizer.
/// @param config Config for the recognizer.
/// @return Return a pointer to the recognizer. The user has to invoke
// DestroyOnlineRecognizer() to free it to avoid memory leak.
SherpaOnnxOnlineRecognizer *CreateOnlineRecognizer(
SHERPA_ONNX_API SherpaOnnxOnlineRecognizer *CreateOnlineRecognizer(
const SherpaOnnxOnlineRecognizerConfig *config);
/// Free a pointer returned by CreateOnlineRecognizer()
///
/// @param p A pointer returned by CreateOnlineRecognizer()
void DestroyOnlineRecognizer(SherpaOnnxOnlineRecognizer *recognizer);
SHERPA_ONNX_API void DestroyOnlineRecognizer(
SherpaOnnxOnlineRecognizer *recognizer);
/// Create an online stream for accepting wave samples.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer()
/// @return Return a pointer to an OnlineStream. The user has to invoke
/// DestoryOnlineStream() to free it to avoid memory leak.
SherpaOnnxOnlineStream *CreateOnlineStream(
SHERPA_ONNX_API SherpaOnnxOnlineStream *CreateOnlineStream(
const SherpaOnnxOnlineRecognizer *recognizer);
/// Destory an online stream.
/// Destroy an online stream.
///
/// @param stream A pointer returned by CreateOnlineStream()
void DestoryOnlineStream(SherpaOnnxOnlineStream *stream);
SHERPA_ONNX_API void DestoryOnlineStream(SherpaOnnxOnlineStream *stream);
/// Accept input audio samples and compute the features.
/// The user has to invoke DecodeOnlineStream() to run the neural network and
... ... @@ -121,16 +146,17 @@ void DestoryOnlineStream(SherpaOnnxOnlineStream *stream);
/// @param samples A pointer to a 1-D array containing audio samples.
/// The range of samples has to be normalized to [-1, 1].
/// @param n Number of elements in the samples array.
void AcceptWaveform(SherpaOnnxOnlineStream *stream, int32_t sample_rate,
const float *samples, int32_t n);
SHERPA_ONNX_API void AcceptWaveform(SherpaOnnxOnlineStream *stream,
int32_t sample_rate, const float *samples,
int32_t n);
/// Return 1 if there are enough number of feature frames for decoding.
/// Return 0 otherwise.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer
/// @param stream A pointer returned by CreateOnlineStream
int32_t IsOnlineStreamReady(SherpaOnnxOnlineRecognizer *recognizer,
SherpaOnnxOnlineStream *stream);
SHERPA_ONNX_API int32_t IsOnlineStreamReady(
SherpaOnnxOnlineRecognizer *recognizer, SherpaOnnxOnlineStream *stream);
/// Call this function to run the neural network model and decoding.
//
... ... @@ -142,8 +168,8 @@ int32_t IsOnlineStreamReady(SherpaOnnxOnlineRecognizer *recognizer,
/// DecodeOnlineStream(recognizer, stream);
/// }
///
void DecodeOnlineStream(SherpaOnnxOnlineRecognizer *recognizer,
SherpaOnnxOnlineStream *stream);
SHERPA_ONNX_API void DecodeOnlineStream(SherpaOnnxOnlineRecognizer *recognizer,
SherpaOnnxOnlineStream *stream);
/// This function is similar to DecodeOnlineStream(). It decodes multiple
/// OnlineStream in parallel.
... ... @@ -155,8 +181,9 @@ void DecodeOnlineStream(SherpaOnnxOnlineRecognizer *recognizer,
/// @param streams A pointer array containing pointers returned by
/// CreateOnlineRecognizer()
/// @param n Number of elements in the given streams array.
void DecodeMultipleOnlineStreams(SherpaOnnxOnlineRecognizer *recognizer,
SherpaOnnxOnlineStream **streams, int32_t n);
SHERPA_ONNX_API void DecodeMultipleOnlineStreams(
SherpaOnnxOnlineRecognizer *recognizer, SherpaOnnxOnlineStream **streams,
int32_t n);
/// Get the decoding results so far for an OnlineStream.
///
... ... @@ -165,47 +192,188 @@ void DecodeMultipleOnlineStreams(SherpaOnnxOnlineRecognizer *recognizer,
/// @return A pointer containing the result. The user has to invoke
/// DestroyOnlineRecognizerResult() to free the returned pointer to
/// avoid memory leak.
SherpaOnnxOnlineRecognizerResult *GetOnlineStreamResult(
SHERPA_ONNX_API SherpaOnnxOnlineRecognizerResult *GetOnlineStreamResult(
SherpaOnnxOnlineRecognizer *recognizer, SherpaOnnxOnlineStream *stream);
/// Destroy the pointer returned by GetOnlineStreamResult().
///
/// @param r A pointer returned by GetOnlineStreamResult()
void DestroyOnlineRecognizerResult(const SherpaOnnxOnlineRecognizerResult *r);
SHERPA_ONNX_API void DestroyOnlineRecognizerResult(
const SherpaOnnxOnlineRecognizerResult *r);
/// Reset an OnlineStream , which clears the neural network model state
/// and the state for decoding.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer().
/// @param stream A pointer returned by CreateOnlineStream
void Reset(SherpaOnnxOnlineRecognizer *recognizer,
SherpaOnnxOnlineStream *stream);
SHERPA_ONNX_API void Reset(SherpaOnnxOnlineRecognizer *recognizer,
SherpaOnnxOnlineStream *stream);
/// Signal that no more audio samples would be available.
/// After this call, you cannot call AcceptWaveform() any more.
///
/// @param stream A pointer returned by CreateOnlineStream()
void InputFinished(SherpaOnnxOnlineStream *stream);
SHERPA_ONNX_API void InputFinished(SherpaOnnxOnlineStream *stream);
/// Return 1 if an endpoint has been detected.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer()
/// @param stream A pointer returned by CreateOnlineStream()
/// @return Return 1 if an endpoint is detected. Return 0 otherwise.
int32_t IsEndpoint(SherpaOnnxOnlineRecognizer *recognizer,
SherpaOnnxOnlineStream *stream);
SHERPA_ONNX_API int32_t IsEndpoint(SherpaOnnxOnlineRecognizer *recognizer,
SherpaOnnxOnlineStream *stream);
// for displaying results on Linux/macOS.
typedef struct SherpaOnnxDisplay SherpaOnnxDisplay;
SHERPA_ONNX_API typedef struct SherpaOnnxDisplay SherpaOnnxDisplay;
/// Create a display object. Must be freed using DestroyDisplay to avoid
/// memory leak.
SherpaOnnxDisplay *CreateDisplay(int32_t max_word_per_line);
SHERPA_ONNX_API SherpaOnnxDisplay *CreateDisplay(int32_t max_word_per_line);
void DestroyDisplay(SherpaOnnxDisplay *display);
SHERPA_ONNX_API void DestroyDisplay(SherpaOnnxDisplay *display);
/// Print the result.
void SherpaOnnxPrint(SherpaOnnxDisplay *display, int32_t idx, const char *s);
SHERPA_ONNX_API void SherpaOnnxPrint(SherpaOnnxDisplay *display, int32_t idx,
const char *s);
// ============================================================
// For offline ASR (i.e., non-streaming ASR)
// ============================================================
/// Please refer to
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
/// to download pre-trained models. That is, you can find encoder-xxx.onnx
/// decoder-xxx.onnx, and joiner-xxx.onnx for this struct
/// from there.
SHERPA_ONNX_API typedef struct SherpaOnnxOfflineTransducerModelConfig {
const char *encoder;
const char *decoder;
const char *joiner;
} SherpaOnnxOfflineTransducerModelConfig;
SHERPA_ONNX_API typedef struct SherpaOnnxOfflineParaformerModelConfig {
const char *model;
} SherpaOnnxOfflineParaformerModelConfig;
SHERPA_ONNX_API typedef struct SherpaOnnxOfflineNemoEncDecCtcModelConfig {
const char *model;
} SherpaOnnxOfflineNemoEncDecCtcModelConfig;
SHERPA_ONNX_API typedef struct SherpaOnnxOfflineLMConfig {
const char *model;
float scale;
} SherpaOnnxOfflineLMConfig;
SHERPA_ONNX_API typedef struct SherpaOnnxOfflineModelConfig {
SherpaOnnxOfflineTransducerModelConfig transducer;
SherpaOnnxOfflineParaformerModelConfig paraformer;
SherpaOnnxOfflineNemoEncDecCtcModelConfig nemo_ctc;
const char *tokens;
int32_t num_threads;
int32_t debug;
} SherpaOnnxOfflineModelConfig;
SHERPA_ONNX_API typedef struct SherpaOnnxOfflineRecognizerConfig {
SherpaOnnxFeatureConfig feat_config;
SherpaOnnxOfflineModelConfig model_config;
SherpaOnnxOfflineLMConfig lm_config;
const char *decoding_method;
int32_t max_active_paths;
} SherpaOnnxOfflineRecognizerConfig;
SHERPA_ONNX_API typedef struct SherpaOnnxOfflineRecognizer
SherpaOnnxOfflineRecognizer;
SHERPA_ONNX_API typedef struct SherpaOnnxOfflineStream SherpaOnnxOfflineStream;
/// @param config Config for the recognizer.
/// @return Return a pointer to the recognizer. The user has to invoke
// DestroyOfflineRecognizer() to free it to avoid memory leak.
SHERPA_ONNX_API SherpaOnnxOfflineRecognizer *CreateOfflineRecognizer(
const SherpaOnnxOfflineRecognizerConfig *config);
/// Free a pointer returned by CreateOfflineRecognizer()
///
/// @param p A pointer returned by CreateOfflineRecognizer()
SHERPA_ONNX_API void DestroyOfflineRecognizer(
SherpaOnnxOfflineRecognizer *recognizer);
/// Create an offline stream for accepting wave samples.
///
/// @param recognizer A pointer returned by CreateOfflineRecognizer()
/// @return Return a pointer to an OfflineStream. The user has to invoke
/// DestoryOfflineStream() to free it to avoid memory leak.
SHERPA_ONNX_API SherpaOnnxOfflineStream *CreateOfflineStream(
const SherpaOnnxOfflineRecognizer *recognizer);
/// Destroy an offline stream.
///
/// @param stream A pointer returned by CreateOfflineStream()
SHERPA_ONNX_API void DestoryOfflineStream(SherpaOnnxOfflineStream *stream);
/// Accept input audio samples and compute the features.
/// The user has to invoke DecodeOfflineStream() to run the neural network and
/// decoding.
///
/// @param stream A pointer returned by CreateOfflineStream().
/// @param sample_rate Sample rate of the input samples. If it is different
/// from config.feat_config.sample_rate, we will do
/// resampling inside sherpa-onnx.
/// @param samples A pointer to a 1-D array containing audio samples.
/// The range of samples has to be normalized to [-1, 1].
/// @param n Number of elements in the samples array.
///
/// @caution: For each offline stream, please invoke this function only once!
SHERPA_ONNX_API void AcceptWaveformOffline(SherpaOnnxOfflineStream *stream,
int32_t sample_rate,
const float *samples, int32_t n);
/// Decode an offline stream.
///
/// We assume you have invoked AcceptWaveformOffline() for the given stream
/// before calling this function.
///
/// @param recognizer A pointer returned by CreateOfflineRecognizer().
/// @param stream A pointer returned by CreateOfflineStream()
SHERPA_ONNX_API void DecodeOfflineStream(
SherpaOnnxOfflineRecognizer *recognizer, SherpaOnnxOfflineStream *stream);
/// Decode a list offline streams in parallel.
///
/// We assume you have invoked AcceptWaveformOffline() for each stream
/// before calling this function.
///
/// @param recognizer A pointer returned by CreateOfflineRecognizer().
/// @param streams A pointer pointer array containing pointers returned
/// by CreateOfflineStream().
/// @param n Number of entries in the given streams.
SHERPA_ONNX_API void DecodeMultipleOfflineStreams(
SherpaOnnxOfflineRecognizer *recognizer, SherpaOnnxOfflineStream **streams,
int32_t n);
SHERPA_ONNX_API typedef struct SherpaOnnxOfflineRecognizerResult {
const char *text;
// TODO(fangjun): Add more fields
} SherpaOnnxOfflineRecognizerResult;
/// Get the result of the offline stream.
///
/// We assume you have called DecodeOfflineStream() or
/// DecodeMultipleOfflineStreams() with the given stream before calling
/// this function.
///
/// @param stream A pointer returned by CreateOfflineStream().
/// @return Return a pointer to the result. The user has to invoke
/// DestroyOnlineRecognizerResult() to free the returned pointer to
/// avoid memory leak.
SHERPA_ONNX_API SherpaOnnxOfflineRecognizerResult *GetOfflineStreamResult(
SherpaOnnxOfflineStream *stream);
/// Destroy the pointer returned by GetOfflineStreamResult().
///
/// @param r A pointer returned by GetOfflineStreamResult()
SHERPA_ONNX_API void DestroyOfflineRecognizerResult(
const SherpaOnnxOfflineRecognizerResult *r);
#ifdef __cplusplus
} /* extern "C" */
... ...
using System.Runtime.InteropServices;
using System.Diagnostics;
namespace SherpaOnnx
{
/// <summary>
/// online recognizer package
/// Copyright (c) 2023 by manyeyes
/// </summary>
public class OnlineBase : IDisposable
{
public void Dispose()
{
Dispose(disposing: true);
GC.SuppressFinalize(this);
}
protected virtual void Dispose(bool disposing)
{
if (!disposing)
{
if (_onlineRecognizerResult != IntPtr.Zero)
{
SherpaOnnxSharp.DestroyOnlineRecognizerResult(_onlineRecognizerResult);
_onlineRecognizerResult = IntPtr.Zero;
}
if (_onlineStream.impl != IntPtr.Zero)
{
SherpaOnnxSharp.DestroyOnlineStream(_onlineStream);
_onlineStream.impl = IntPtr.Zero;
}
if (_onlineRecognizer.impl != IntPtr.Zero)
{
SherpaOnnxSharp.DestroyOnlineRecognizer(_onlineRecognizer);
_onlineRecognizer.impl = IntPtr.Zero;
}
this._disposed = true;
}
}
~OnlineBase()
{
Dispose(this._disposed);
}
internal SherpaOnnxOnlineStream _onlineStream;
internal IntPtr _onlineRecognizerResult;
internal SherpaOnnxOnlineRecognizer _onlineRecognizer;
internal bool _disposed = false;
}
public class OnlineStream : OnlineBase
{
internal OnlineStream(SherpaOnnxOnlineStream onlineStream)
{
this._onlineStream = onlineStream;
}
protected override void Dispose(bool disposing)
{
if (!disposing)
{
SherpaOnnxSharp.DestroyOnlineStream(_onlineStream);
_onlineStream.impl = IntPtr.Zero;
this._disposed = true;
base.Dispose();
}
}
}
public class OnlineRecognizerResult : OnlineBase
{
internal OnlineRecognizerResult(IntPtr onlineRecognizerResult)
{
this._onlineRecognizerResult = onlineRecognizerResult;
}
protected override void Dispose(bool disposing)
{
if (!disposing)
{
SherpaOnnxSharp.DestroyOnlineRecognizerResult(_onlineRecognizerResult);
_onlineRecognizerResult = IntPtr.Zero;
this._disposed = true;
base.Dispose(disposing);
}
}
}
public class OnlineRecognizer<T> : OnlineBase
where T : class, new()
{
public OnlineRecognizer(T t,
string tokensFilePath, string decoding_method = "greedy_search",
int sample_rate = 16000, int feature_dim = 80,
int num_threads = 2, bool debug = false, int max_active_paths = 4,
int enable_endpoint=0,int rule1_min_trailing_silence=0,
int rule2_min_trailing_silence=0,int rule3_min_utterance_length=0)
{
SherpaOnnxOnlineTransducer transducer = new SherpaOnnxOnlineTransducer();
SherpaOnnxOnlineModelConfig model_config = new SherpaOnnxOnlineModelConfig();
if (t is not null && t.GetType() == typeof(OnlineTransducer))
{
OnlineTransducer? onlineTransducer = t as OnlineTransducer;
#pragma warning disable CS8602 // 解引用可能出现空引用。
Trace.Assert(File.Exists(onlineTransducer.DecoderFilename)
&& File.Exists(onlineTransducer.EncoderFilename)
&& File.Exists(onlineTransducer.JoinerFilename), "Please provide a model");
#pragma warning restore CS8602 // 解引用可能出现空引用。
Trace.Assert(File.Exists(tokensFilePath), "Please provide a tokens");
Trace.Assert(num_threads > 0, "num_threads must be greater than 0");
transducer.encoder_filename = onlineTransducer.EncoderFilename;
transducer.decoder_filename = onlineTransducer.DecoderFilename;
transducer.joiner_filename = onlineTransducer.JoinerFilename;
}
model_config.transducer = transducer;
model_config.num_threads = num_threads;
model_config.debug = debug;
model_config.tokens = tokensFilePath;
SherpaOnnxFeatureConfig feat_config = new SherpaOnnxFeatureConfig();
feat_config.sample_rate = sample_rate;
feat_config.feature_dim = feature_dim;
SherpaOnnxOnlineRecognizerConfig sherpaOnnxOnlineRecognizerConfig;
sherpaOnnxOnlineRecognizerConfig.decoding_method = decoding_method;
sherpaOnnxOnlineRecognizerConfig.feat_config = feat_config;
sherpaOnnxOnlineRecognizerConfig.model_config = model_config;
sherpaOnnxOnlineRecognizerConfig.max_active_paths = max_active_paths;
//endpoint
sherpaOnnxOnlineRecognizerConfig.enable_endpoint = enable_endpoint;
sherpaOnnxOnlineRecognizerConfig.rule1_min_trailing_silence = rule1_min_trailing_silence;
sherpaOnnxOnlineRecognizerConfig.rule2_min_trailing_silence = rule2_min_trailing_silence;
sherpaOnnxOnlineRecognizerConfig.rule3_min_utterance_length = rule3_min_utterance_length;
_onlineRecognizer =
SherpaOnnxSharp.CreateOnlineRecognizer(sherpaOnnxOnlineRecognizerConfig);
}
internal OnlineStream CreateOnlineStream()
{
SherpaOnnxOnlineStream stream = SherpaOnnxSharp.CreateOnlineStream(_onlineRecognizer);
return new OnlineStream(stream);
}
public void InputFinished(OnlineStream stream)
{
SherpaOnnxSharp.InputFinished(stream._onlineStream);
}
public List<OnlineStream> CreateStreams(List<float[]> samplesList)
{
int batch_size = samplesList.Count;
List<OnlineStream> streams = new List<OnlineStream>();
for (int i = 0; i < batch_size; i++)
{
OnlineStream stream = CreateOnlineStream();
AcceptWaveform(stream._onlineStream, 16000, samplesList[i]);
InputFinished(stream);
streams.Add(stream);
}
return streams;
}
public OnlineStream CreateStream()
{
OnlineStream stream = CreateOnlineStream();
return stream;
}
internal void AcceptWaveform(SherpaOnnxOnlineStream stream, int sample_rate, float[] samples)
{
SherpaOnnxSharp.AcceptOnlineWaveform(stream, sample_rate, samples, samples.Length);
}
public void AcceptWaveForm(OnlineStream stream, int sample_rate, float[] samples)
{
AcceptWaveform(stream._onlineStream, sample_rate, samples);
}
internal IntPtr GetStreamsIntPtr(OnlineStream[] streams)
{
int streams_len = streams.Length;
int size = Marshal.SizeOf(typeof(SherpaOnnxOnlineStream));
IntPtr streamsIntPtr = Marshal.AllocHGlobal(size * streams_len);
unsafe
{
byte* ptrbds = (byte*)(streamsIntPtr.ToPointer());
for (int i = 0; i < streams_len; i++, ptrbds += (size))
{
IntPtr streamIntptr = new IntPtr(ptrbds);
Marshal.StructureToPtr(streams[i]._onlineStream, streamIntptr, false);
}
}
return streamsIntPtr;
}
internal bool IsReady(OnlineStream stream)
{
return SherpaOnnxSharp.IsOnlineStreamReady(_onlineRecognizer, stream._onlineStream) != 0;
}
public void DecodeMultipleStreams(List<OnlineStream> streams)
{
while (true)
{
List<OnlineStream> streamList = new List<OnlineStream>();
foreach (OnlineStream stream in streams)
{
if (IsReady(stream))
{
streamList.Add(stream);
}
}
if (streamList.Count == 0)
{
break;
}
OnlineStream[] streamsBatch = new OnlineStream[streamList.Count];
for (int i = 0; i < streamsBatch.Length; i++)
{
streamsBatch[i] = streamList[i];
}
streamList.Clear();
IntPtr streamsIntPtr = GetStreamsIntPtr(streamsBatch);
SherpaOnnxSharp.DecodeMultipleOnlineStreams(_onlineRecognizer, streamsIntPtr, streamsBatch.Length);
Marshal.FreeHGlobal(streamsIntPtr);
}
}
public void DecodeStream(OnlineStream stream)
{
while (IsReady(stream))
{
SherpaOnnxSharp.DecodeOnlineStream(_onlineRecognizer, stream._onlineStream);
}
}
internal OnlineRecognizerResultEntity GetResult(SherpaOnnxOnlineStream stream)
{
IntPtr result_ip = SherpaOnnxSharp.GetOnlineStreamResult(_onlineRecognizer, stream);
OnlineRecognizerResult onlineRecognizerResult = new OnlineRecognizerResult(result_ip);
#pragma warning disable CS8605 // 取消装箱可能为 null 的值。
SherpaOnnxOnlineRecognizerResult result =
(SherpaOnnxOnlineRecognizerResult)Marshal.PtrToStructure(
onlineRecognizerResult._onlineRecognizerResult, typeof(SherpaOnnxOnlineRecognizerResult));
#pragma warning restore CS8605 // 取消装箱可能为 null 的值。
#pragma warning disable CS8600 // 将 null 字面量或可能为 null 的值转换为非 null 类型。
string text = Marshal.PtrToStringAnsi(result.text);
#pragma warning restore CS8600 // 将 null 字面量或可能为 null 的值转换为非 null 类型。
OnlineRecognizerResultEntity onlineRecognizerResultEntity =
new OnlineRecognizerResultEntity();
onlineRecognizerResultEntity.text = text;
onlineRecognizerResultEntity.text_len = result.text_len;
return onlineRecognizerResultEntity;
}
public OnlineRecognizerResultEntity GetResult(OnlineStream stream)
{
OnlineRecognizerResultEntity result = GetResult(stream._onlineStream);
return result;
}
public List<OnlineRecognizerResultEntity> GetResults(List<OnlineStream> streams)
{
List<OnlineRecognizerResultEntity> results = new List<OnlineRecognizerResultEntity>();
foreach (OnlineStream stream in streams)
{
OnlineRecognizerResultEntity onlineRecognizerResultEntity = GetResult(stream._onlineStream);
results.Add(onlineRecognizerResultEntity);
}
return results;
}
protected override void Dispose(bool disposing)
{
if (!disposing)
{
SherpaOnnxSharp.DestroyOnlineRecognizer(_onlineRecognizer);
_onlineRecognizer.impl = IntPtr.Zero;
this._disposed = true;
base.Dispose();
}
}
}
public class OfflineBase : IDisposable
{
public void Dispose()
{
Dispose(disposing: true);
GC.SuppressFinalize(this);
}
protected virtual void Dispose(bool disposing)
{
if (!disposing)
{
if (_offlineRecognizerResult != IntPtr.Zero)
{
SherpaOnnxSharp.DestroyOfflineRecognizerResult(_offlineRecognizerResult);
_offlineRecognizerResult = IntPtr.Zero;
}
if (_offlineStream.impl != IntPtr.Zero)
{
SherpaOnnxSharp.DestroyOfflineStream(_offlineStream);
_offlineStream.impl = IntPtr.Zero;
}
if (_offlineRecognizer.impl != IntPtr.Zero)
{
SherpaOnnxSharp.DestroyOfflineRecognizer(_offlineRecognizer);
_offlineRecognizer.impl = IntPtr.Zero;
}
this._disposed = true;
}
}
~OfflineBase()
{
Dispose(this._disposed);
}
internal SherpaOnnxOfflineStream _offlineStream;
internal IntPtr _offlineRecognizerResult;
internal SherpaOnnxOfflineRecognizer _offlineRecognizer;
internal bool _disposed = false;
}
public class OfflineStream : OfflineBase
{
internal OfflineStream(SherpaOnnxOfflineStream offlineStream)
{
this._offlineStream = offlineStream;
}
protected override void Dispose(bool disposing)
{
if (!disposing)
{
SherpaOnnxSharp.DestroyOfflineStream(_offlineStream);
_offlineStream.impl = IntPtr.Zero;
this._disposed = true;
base.Dispose();
}
}
}
public class OfflineRecognizerResult : OfflineBase
{
internal OfflineRecognizerResult(IntPtr offlineRecognizerResult)
{
this._offlineRecognizerResult = offlineRecognizerResult;
}
protected override void Dispose(bool disposing)
{
if (!disposing)
{
SherpaOnnxSharp.DestroyOfflineRecognizerResult(_offlineRecognizerResult);
_offlineRecognizerResult = IntPtr.Zero;
this._disposed = true;
base.Dispose(disposing);
}
}
}
public class OfflineRecognizer<T> : OfflineBase
where T : class, new()
{
public OfflineRecognizer(T t,
string tokensFilePath, string decoding_method = "greedy_search",
int sample_rate = 16000, int feature_dim = 80,
int num_threads = 2, bool debug = false)
{
SherpaOnnxOfflineTransducer transducer = new SherpaOnnxOfflineTransducer();
SherpaOnnxOfflineParaformer paraformer = new SherpaOnnxOfflineParaformer();
SherpaOnnxOfflineNemoEncDecCtc nemo_ctc = new SherpaOnnxOfflineNemoEncDecCtc();
SherpaOnnxOfflineModelConfig model_config = new SherpaOnnxOfflineModelConfig();
if (t is not null && t.GetType() == typeof(OfflineTransducer))
{
OfflineTransducer? offlineTransducer = t as OfflineTransducer;
#pragma warning disable CS8602 // 解引用可能出现空引用。
Trace.Assert(File.Exists(offlineTransducer.DecoderFilename)
&& File.Exists(offlineTransducer.EncoderFilename)
&& File.Exists(offlineTransducer.JoinerFilename), "Please provide a model");
#pragma warning restore CS8602 // 解引用可能出现空引用。
Trace.Assert(File.Exists(tokensFilePath), "Please provide a tokens");
Trace.Assert(num_threads > 0, "num_threads must be greater than 0");
transducer.encoder_filename = offlineTransducer.EncoderFilename;
transducer.decoder_filename = offlineTransducer.DecoderFilename;
transducer.joiner_filename = offlineTransducer.JoinerFilename;
}
else if (t is not null && t.GetType() == typeof(OfflineParaformer))
{
OfflineParaformer? offlineParaformer = t as OfflineParaformer;
#pragma warning disable CS8602 // 解引用可能出现空引用。
Trace.Assert(File.Exists(offlineParaformer.Model), "Please provide a model");
#pragma warning restore CS8602 // 解引用可能出现空引用。
Trace.Assert(File.Exists(tokensFilePath), "Please provide a tokens");
Trace.Assert(num_threads > 0, "num_threads must be greater than 0");
paraformer.model = offlineParaformer.Model;
}
else if (t is not null && t.GetType() == typeof(OfflineNemoEncDecCtc))
{
OfflineNemoEncDecCtc? offlineNemoEncDecCtc = t as OfflineNemoEncDecCtc;
#pragma warning disable CS8602 // 解引用可能出现空引用。
Trace.Assert(File.Exists(offlineNemoEncDecCtc.Model), "Please provide a model");
#pragma warning restore CS8602 // 解引用可能出现空引用。
Trace.Assert(File.Exists(tokensFilePath), "Please provide a tokens");
Trace.Assert(num_threads > 0, "num_threads must be greater than 0");
nemo_ctc.model = offlineNemoEncDecCtc.Model;
}
model_config.transducer = transducer;
model_config.paraformer = paraformer;
model_config.nemo_ctc = nemo_ctc;
model_config.num_threads = num_threads;
model_config.debug = debug;
model_config.tokens = tokensFilePath;
SherpaOnnxFeatureConfig feat_config = new SherpaOnnxFeatureConfig();
feat_config.sample_rate = sample_rate;
feat_config.feature_dim = feature_dim;
SherpaOnnxOfflineRecognizerConfig sherpaOnnxOfflineRecognizerConfig;
sherpaOnnxOfflineRecognizerConfig.decoding_method = decoding_method;
sherpaOnnxOfflineRecognizerConfig.feat_config = feat_config;
sherpaOnnxOfflineRecognizerConfig.model_config = model_config;
_offlineRecognizer =
SherpaOnnxSharp.CreateOfflineRecognizer(sherpaOnnxOfflineRecognizerConfig);
}
internal OfflineStream CreateOfflineStream()
{
SherpaOnnxOfflineStream stream = SherpaOnnxSharp.CreateOfflineStream(_offlineRecognizer);
return new OfflineStream(stream);
}
public OfflineStream[] CreateOfflineStream(List<float[]> samplesList)
{
int batch_size = samplesList.Count;
OfflineStream[] streams = new OfflineStream[batch_size];
List<string> wavFiles = new List<string>();
for (int i = 0; i < batch_size; i++)
{
OfflineStream stream = CreateOfflineStream();
AcceptWaveform(stream._offlineStream, 16000, samplesList[i]);
streams[i] = stream;
}
return streams;
}
internal void AcceptWaveform(SherpaOnnxOfflineStream stream, int sample_rate, float[] samples)
{
SherpaOnnxSharp.AcceptWaveform(stream, sample_rate, samples, samples.Length);
}
internal IntPtr GetStreamsIntPtr(OfflineStream[] streams)
{
int streams_len = streams.Length;
int size = Marshal.SizeOf(typeof(SherpaOnnxOfflineStream));
IntPtr streamsIntPtr = Marshal.AllocHGlobal(size * streams_len);
unsafe
{
byte* ptrbds = (byte*)(streamsIntPtr.ToPointer());
for (int i = 0; i < streams_len; i++, ptrbds += (size))
{
IntPtr streamIntptr = new IntPtr(ptrbds);
Marshal.StructureToPtr(streams[i]._offlineStream, streamIntptr, false);
}
}
return streamsIntPtr;
}
public void DecodeMultipleOfflineStreams(OfflineStream[] streams)
{
IntPtr streamsIntPtr = GetStreamsIntPtr(streams);
SherpaOnnxSharp.DecodeMultipleOfflineStreams(_offlineRecognizer, streamsIntPtr, streams.Length);
Marshal.FreeHGlobal(streamsIntPtr);
}
internal OfflineRecognizerResultEntity GetResult(SherpaOnnxOfflineStream stream)
{
IntPtr result_ip = SherpaOnnxSharp.GetOfflineStreamResult(stream);
OfflineRecognizerResult offlineRecognizerResult = new OfflineRecognizerResult(result_ip);
#pragma warning disable CS8605 // 取消装箱可能为 null 的值。
SherpaOnnxOfflineRecognizerResult result =
(SherpaOnnxOfflineRecognizerResult)Marshal.PtrToStructure(
offlineRecognizerResult._offlineRecognizerResult, typeof(SherpaOnnxOfflineRecognizerResult));
#pragma warning restore CS8605 // 取消装箱可能为 null 的值。
#pragma warning disable CS8600 // 将 null 字面量或可能为 null 的值转换为非 null 类型。
string text = Marshal.PtrToStringAnsi(result.text);
#pragma warning restore CS8600 // 将 null 字面量或可能为 null 的值转换为非 null 类型。
OfflineRecognizerResultEntity offlineRecognizerResultEntity =
new OfflineRecognizerResultEntity();
offlineRecognizerResultEntity.text = text;
offlineRecognizerResultEntity.text_len = result.text_len;
return offlineRecognizerResultEntity;
}
public List<OfflineRecognizerResultEntity> GetResults(OfflineStream[] streams)
{
List<OfflineRecognizerResultEntity> results = new List<OfflineRecognizerResultEntity>();
foreach (OfflineStream stream in streams)
{
OfflineRecognizerResultEntity offlineRecognizerResultEntity = GetResult(stream._offlineStream);
results.Add(offlineRecognizerResultEntity);
}
return results;
}
protected override void Dispose(bool disposing)
{
if (!disposing)
{
SherpaOnnxSharp.DestroyOfflineRecognizer(_offlineRecognizer);
_offlineRecognizer.impl = IntPtr.Zero;
this._disposed = true;
base.Dispose();
}
}
}
internal static partial class SherpaOnnxSharp
{
private const string dllName = @"SherpaOnnxSharp";
[DllImport(dllName, EntryPoint = "CreateOfflineRecognizer", CharSet = CharSet.Ansi, CallingConvention = CallingConvention.Cdecl)]
internal static extern SherpaOnnxOfflineRecognizer CreateOfflineRecognizer(SherpaOnnxOfflineRecognizerConfig config);
[DllImport(dllName, EntryPoint = "CreateOfflineStream", CharSet = CharSet.Ansi, CallingConvention = CallingConvention.Cdecl)]
internal static extern SherpaOnnxOfflineStream CreateOfflineStream(SherpaOnnxOfflineRecognizer offlineRecognizer);
[DllImport(dllName, EntryPoint = "AcceptWaveform", CharSet = CharSet.Ansi, CallingConvention = CallingConvention.Cdecl)]
internal static extern void AcceptWaveform(SherpaOnnxOfflineStream stream, int sample_rate, float[] samples, int samples_size);
[DllImport(dllName, EntryPoint = "DecodeOfflineStream", CharSet = CharSet.Ansi, CallingConvention = CallingConvention.Cdecl)]
internal static extern void DecodeOfflineStream(SherpaOnnxOfflineRecognizer recognizer, SherpaOnnxOfflineStream stream);
[DllImport(dllName, EntryPoint = "DecodeMultipleOfflineStreams", CharSet = CharSet.Ansi, CallingConvention = CallingConvention.Cdecl)]
internal static extern void DecodeMultipleOfflineStreams(SherpaOnnxOfflineRecognizer recognizer, IntPtr
streams, int n);
[DllImport(dllName, EntryPoint = "GetOfflineStreamResult", CallingConvention = CallingConvention.Cdecl)]
internal static extern IntPtr GetOfflineStreamResult(SherpaOnnxOfflineStream stream);
[DllImport(dllName, EntryPoint = "DestroyOfflineRecognizerResult", CallingConvention = CallingConvention.Cdecl)]
internal static extern void DestroyOfflineRecognizerResult(IntPtr result);
[DllImport(dllName, EntryPoint = "DestroyOfflineStream", CallingConvention = CallingConvention.Cdecl)]
internal static extern void DestroyOfflineStream(SherpaOnnxOfflineStream stream);
[DllImport(dllName, EntryPoint = "DestroyOfflineRecognizer", CallingConvention = CallingConvention.Cdecl)]
internal static extern void DestroyOfflineRecognizer(SherpaOnnxOfflineRecognizer offlineRecognizer);
[DllImport(dllName, EntryPoint = "CreateOnlineRecognizer", CallingConvention = CallingConvention.Cdecl)]
internal static extern SherpaOnnxOnlineRecognizer CreateOnlineRecognizer(SherpaOnnxOnlineRecognizerConfig config);
/// Free a pointer returned by CreateOnlineRecognizer()
///
/// @param p A pointer returned by CreateOnlineRecognizer()
[DllImport(dllName, EntryPoint = "DestroyOnlineRecognizer", CallingConvention = CallingConvention.Cdecl)]
internal static extern void DestroyOnlineRecognizer(SherpaOnnxOnlineRecognizer recognizer);
/// Create an online stream for accepting wave samples.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer()
/// @return Return a pointer to an OnlineStream. The user has to invoke
/// DestroyOnlineStream() to free it to avoid memory leak.
[DllImport(dllName, EntryPoint = "CreateOnlineStream", CallingConvention = CallingConvention.Cdecl)]
internal static extern SherpaOnnxOnlineStream CreateOnlineStream(
SherpaOnnxOnlineRecognizer recognizer);
/// Destroy an online stream.
///
/// @param stream A pointer returned by CreateOnlineStream()
[DllImport(dllName, EntryPoint = "DestroyOnlineStream", CallingConvention = CallingConvention.Cdecl)]
internal static extern void DestroyOnlineStream(SherpaOnnxOnlineStream stream);
/// Accept input audio samples and compute the features.
/// The user has to invoke DecodeOnlineStream() to run the neural network and
/// decoding.
///
/// @param stream A pointer returned by CreateOnlineStream().
/// @param sample_rate Sample rate of the input samples. If it is different
/// from config.feat_config.sample_rate, we will do
/// resampling inside sherpa-onnx.
/// @param samples A pointer to a 1-D array containing audio samples.
/// The range of samples has to be normalized to [-1, 1].
/// @param n Number of elements in the samples array.
[DllImport(dllName, EntryPoint = "AcceptOnlineWaveform", CallingConvention = CallingConvention.Cdecl)]
internal static extern void AcceptOnlineWaveform(SherpaOnnxOnlineStream stream, int sample_rate,
float[] samples, int n);
/// Return 1 if there are enough number of feature frames for decoding.
/// Return 0 otherwise.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer
/// @param stream A pointer returned by CreateOnlineStream
[DllImport(dllName, EntryPoint = "IsOnlineStreamReady", CallingConvention = CallingConvention.Cdecl)]
internal static extern int IsOnlineStreamReady(SherpaOnnxOnlineRecognizer recognizer,
SherpaOnnxOnlineStream stream);
/// Call this function to run the neural network model and decoding.
//
/// Precondition for this function: IsOnlineStreamReady() MUST return 1.
///
/// Usage example:
///
/// while (IsOnlineStreamReady(recognizer, stream)) {
/// DecodeOnlineStream(recognizer, stream);
/// }
///
[DllImport(dllName, EntryPoint = "DecodeOnlineStream", CallingConvention = CallingConvention.Cdecl)]
internal static extern void DecodeOnlineStream(SherpaOnnxOnlineRecognizer recognizer,
SherpaOnnxOnlineStream stream);
/// This function is similar to DecodeOnlineStream(). It decodes multiple
/// OnlineStream in parallel.
///
/// Caution: The caller has to ensure each OnlineStream is ready, i.e.,
/// IsOnlineStreamReady() for that stream should return 1.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer()
/// @param streams A pointer array containing pointers returned by
/// CreateOnlineRecognizer()
/// @param n Number of elements in the given streams array.
[DllImport(dllName, EntryPoint = "DecodeMultipleOnlineStreams", CallingConvention = CallingConvention.Cdecl)]
internal static extern void DecodeMultipleOnlineStreams(SherpaOnnxOnlineRecognizer recognizer,
IntPtr streams, int n);
/// Get the decoding results so far for an OnlineStream.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer().
/// @param stream A pointer returned by CreateOnlineStream().
/// @return A pointer containing the result. The user has to invoke
/// DestroyOnlineRecognizerResult() to free the returned pointer to
/// avoid memory leak.
[DllImport(dllName, EntryPoint = "GetOnlineStreamResult", CallingConvention = CallingConvention.Cdecl)]
internal static extern IntPtr GetOnlineStreamResult(
SherpaOnnxOnlineRecognizer recognizer, SherpaOnnxOnlineStream stream);
/// Destroy the pointer returned by GetOnlineStreamResult().
///
/// @param r A pointer returned by GetOnlineStreamResult()
[DllImport(dllName, EntryPoint = "DestroyOnlineRecognizerResult", CallingConvention = CallingConvention.Cdecl)]
internal static extern void DestroyOnlineRecognizerResult(IntPtr result);
/// Reset an OnlineStream , which clears the neural network model state
/// and the state for decoding.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer().
/// @param stream A pointer returned by CreateOnlineStream
[DllImport(dllName, EntryPoint = "Reset", CallingConvention = CallingConvention.Cdecl)]
internal static extern void Reset(SherpaOnnxOnlineRecognizer recognizer,
SherpaOnnxOnlineStream stream);
/// Signal that no more audio samples would be available.
/// After this call, you cannot call AcceptWaveform() any more.
///
/// @param stream A pointer returned by CreateOnlineStream()
[DllImport(dllName, EntryPoint = "InputFinished", CallingConvention = CallingConvention.Cdecl)]
internal static extern void InputFinished(SherpaOnnxOnlineStream stream);
/// Return 1 if an endpoint has been detected.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer()
/// @param stream A pointer returned by CreateOnlineStream()
/// @return Return 1 if an endpoint is detected. Return 0 otherwise.
[DllImport(dllName, EntryPoint = "IsEndpoint", CallingConvention = CallingConvention.Cdecl)]
internal static extern int IsEndpoint(SherpaOnnxOnlineRecognizer recognizer,
SherpaOnnxOnlineStream stream);
}
internal struct SherpaOnnxOfflineTransducer
{
public string encoder_filename;
public string decoder_filename;
public string joiner_filename;
public SherpaOnnxOfflineTransducer()
{
encoder_filename = "";
decoder_filename = "";
joiner_filename = "";
}
};
internal struct SherpaOnnxOfflineParaformer
{
public string model;
public SherpaOnnxOfflineParaformer()
{
model = "";
}
};
internal struct SherpaOnnxOfflineNemoEncDecCtc
{
public string model;
public SherpaOnnxOfflineNemoEncDecCtc()
{
model = "";
}
};
internal struct SherpaOnnxOfflineModelConfig
{
public SherpaOnnxOfflineTransducer transducer;
public SherpaOnnxOfflineParaformer paraformer;
public SherpaOnnxOfflineNemoEncDecCtc nemo_ctc;
public string tokens;
public int num_threads;
public bool debug;
};
/// It expects 16 kHz 16-bit single channel wave format.
internal struct SherpaOnnxFeatureConfig
{
/// Sample rate of the input data. MUST match the one expected
/// by the model. For instance, it should be 16000 for models provided
/// by us.
public int sample_rate;
/// Feature dimension of the model.
/// For instance, it should be 80 for models provided by us.
public int feature_dim;
};
internal struct SherpaOnnxOfflineRecognizerConfig
{
public SherpaOnnxFeatureConfig feat_config;
public SherpaOnnxOfflineModelConfig model_config;
/// Possible values are: greedy_search, modified_beam_search
public string decoding_method;
};
internal struct SherpaOnnxOfflineRecognizer
{
public IntPtr impl;
};
[StructLayout(LayoutKind.Sequential, CharSet = CharSet.Ansi, Pack = 1)]
internal struct SherpaOnnxOfflineStream
{
public IntPtr impl;
};
internal struct SherpaOnnxOfflineRecognizerResult
{
public IntPtr text;
public int text_len;
}
internal struct SherpaOnnxOnlineTransducer
{
public string encoder_filename;
public string decoder_filename;
public string joiner_filename;
public SherpaOnnxOnlineTransducer()
{
encoder_filename = string.Empty;
decoder_filename = string.Empty;
joiner_filename = string.Empty;
}
};
internal struct SherpaOnnxOnlineModelConfig
{
public SherpaOnnxOnlineTransducer transducer;
public string tokens;
public int num_threads;
public bool debug; // true to print debug information of the model
};
internal struct SherpaOnnxOnlineRecognizerConfig
{
public SherpaOnnxFeatureConfig feat_config;
public SherpaOnnxOnlineModelConfig model_config;
/// Possible values are: greedy_search, modified_beam_search
public string decoding_method;
/// Used only when decoding_method is modified_beam_search
/// Example value: 4
public int max_active_paths;
/// 0 to disable endpoint detection.
/// A non-zero value to enable endpoint detection.
public int enable_endpoint;
/// An endpoint is detected if trailing silence in seconds is larger than
/// this value even if nothing has been decoded.
/// Used only when enable_endpoint is not 0.
public float rule1_min_trailing_silence;
/// An endpoint is detected if trailing silence in seconds is larger than
/// this value after something that is not blank has been decoded.
/// Used only when enable_endpoint is not 0.
public float rule2_min_trailing_silence;
/// An endpoint is detected if the utterance in seconds is larger than
/// this value.
/// Used only when enable_endpoint is not 0.
public float rule3_min_utterance_length;
};
internal struct SherpaOnnxOnlineRecognizerResult
{
public IntPtr text;
public int text_len;
// TODO: Add more fields
}
internal struct SherpaOnnxOnlineRecognizer
{
public IntPtr impl;
};
[StructLayout(LayoutKind.Sequential, CharSet = CharSet.Ansi, Pack = 1)]
internal struct SherpaOnnxOnlineStream
{
public IntPtr impl;
};
public class OfflineNemoEncDecCtc
{
private string model = string.Empty;
public string Model { get => model; set => model = value; }
}
public class OfflineParaformer
{
private string model = string.Empty;
public string Model { get => model; set => model = value; }
}
public class OfflineRecognizerResultEntity
{
/// <summary>
/// recognizer result
/// </summary>
public string? text { get; set; }
/// <summary>
/// recognizer result length
/// </summary>
public int text_len { get; set; }
/// <summary>
/// decode tokens
/// </summary>
public List<string>? tokens { get; set; }
/// <summary>
/// timestamps
/// </summary>
public List<float>? timestamps { get; set; }
}
public class OfflineTransducer
{
private string encoderFilename = string.Empty;
private string decoderFilename = string.Empty;
private string joinerFilename = string.Empty;
public string EncoderFilename { get => encoderFilename; set => encoderFilename = value; }
public string DecoderFilename { get => decoderFilename; set => decoderFilename = value; }
public string JoinerFilename { get => joinerFilename; set => joinerFilename = value; }
}
public class OnlineEndpoint
{
/// 0 to disable endpoint detection.
/// A non-zero value to enable endpoint detection.
private int enableEndpoint;
/// An endpoint is detected if trailing silence in seconds is larger than
/// this value even if nothing has been decoded.
/// Used only when enable_endpoint is not 0.
private float rule1MinTrailingSilence;
/// An endpoint is detected if trailing silence in seconds is larger than
/// this value after something that is not blank has been decoded.
/// Used only when enable_endpoint is not 0.
private float rule2MinTrailingSilence;
/// An endpoint is detected if the utterance in seconds is larger than
/// this value.
/// Used only when enable_endpoint is not 0.
private float rule3MinUtteranceLength;
public int EnableEndpoint { get => enableEndpoint; set => enableEndpoint = value; }
public float Rule1MinTrailingSilence { get => rule1MinTrailingSilence; set => rule1MinTrailingSilence = value; }
public float Rule2MinTrailingSilence { get => rule2MinTrailingSilence; set => rule2MinTrailingSilence = value; }
public float Rule3MinUtteranceLength { get => rule3MinUtteranceLength; set => rule3MinUtteranceLength = value; }
}
public class OnlineRecognizerResultEntity
{
/// <summary>
/// recognizer result
/// </summary>
public string? text { get; set; }
/// <summary>
/// recognizer result length
/// </summary>
public int text_len { get; set; }
/// <summary>
/// decode tokens
/// </summary>
public List<string>? tokens { get; set; }
/// <summary>
/// timestamps
/// </summary>
public List<float>? timestamps { get; set; }
}
public class OnlineTransducer
{
private string encoderFilename = string.Empty;
private string decoderFilename = string.Empty;
private string joinerFilename = string.Empty;
public string EncoderFilename { get => encoderFilename; set => encoderFilename = value; }
public string DecoderFilename { get => decoderFilename; set => decoderFilename = value; }
public string JoinerFilename { get => joinerFilename; set => joinerFilename = value; }
}
}
\ No newline at end of file
// sherpa-onnx/sharp-api/offline-api.cpp
//
// Copyright (c) 2023 Manyeyes Corporation
#include "offline-api.h"
#include "sherpa-onnx/csrc/display.h"
#include "sherpa-onnx/csrc/offline-recognizer.h"
namespace sherpa_onnx
{
struct SherpaOnnxOfflineRecognizer {
sherpa_onnx::OfflineRecognizer* impl;
};
struct SherpaOnnxOfflineStream {
std::unique_ptr<sherpa_onnx::OfflineStream> impl;
explicit SherpaOnnxOfflineStream(std::unique_ptr<sherpa_onnx::OfflineStream> p)
: impl(std::move(p)) {}
};
struct SherpaOnnxDisplay {
std::unique_ptr<sherpa_onnx::Display> impl;
};
SherpaOnnxOfflineRecognizer* __stdcall CreateOfflineRecognizer(
const SherpaOnnxOfflineRecognizerConfig* config) {
sherpa_onnx::OfflineRecognizerConfig recognizer_config;
recognizer_config.feat_config.sampling_rate = config->feat_config.sample_rate;
recognizer_config.feat_config.feature_dim = config->feat_config.feature_dim;
if (strlen(config->model_config.transducer.encoder_filename) > 0) {
recognizer_config.model_config.transducer.encoder_filename =
config->model_config.transducer.encoder_filename;
recognizer_config.model_config.transducer.decoder_filename =
config->model_config.transducer.decoder_filename;
recognizer_config.model_config.transducer.joiner_filename =
config->model_config.transducer.joiner_filename;
}
else if (strlen(config->model_config.paraformer.model) > 0) {
recognizer_config.model_config.paraformer.model =
config->model_config.paraformer.model;
}
else if (strlen(config->model_config.nemo_ctc.model) > 0) {
recognizer_config.model_config.nemo_ctc.model =
config->model_config.nemo_ctc.model;
}
recognizer_config.model_config.tokens =
config->model_config.tokens;
recognizer_config.model_config.num_threads =
config->model_config.num_threads;
recognizer_config.model_config.debug =
config->model_config.debug;
recognizer_config.decoding_method = config->decoding_method;
SherpaOnnxOfflineRecognizer* recognizer =
new SherpaOnnxOfflineRecognizer;
recognizer->impl =
new sherpa_onnx::OfflineRecognizer(recognizer_config);
return recognizer;
}
SherpaOnnxOfflineStream* __stdcall CreateOfflineStream(
SherpaOnnxOfflineRecognizer* recognizer) {
SherpaOnnxOfflineStream* stream =
new SherpaOnnxOfflineStream(recognizer->impl->CreateStream());
return stream;
}
void __stdcall AcceptWaveform(
SherpaOnnxOfflineStream* stream,
int32_t sample_rate,
const float* samples, int32_t samples_size) {
std::vector<float> waveform{ samples, samples + samples_size };
stream->impl->AcceptWaveform(sample_rate, waveform.data(), waveform.size());
}
void __stdcall DecodeOfflineStream(
SherpaOnnxOfflineRecognizer* recognizer,
SherpaOnnxOfflineStream* stream) {
recognizer->impl->DecodeStream(stream->impl.get());
}
void __stdcall DecodeMultipleOfflineStreams(
SherpaOnnxOfflineRecognizer* recognizer,
SherpaOnnxOfflineStream** streams, int32_t n) {
std::vector<sherpa_onnx::OfflineStream*> ss(n);
for (int32_t i = 0; i != n; ++i) {
ss[i] = streams[i]->impl.get();
}
recognizer->impl->DecodeStreams(ss.data(), n);
}
SherpaOnnxOfflineRecognizerResult* __stdcall GetOfflineStreamResult(
SherpaOnnxOfflineStream* stream) {
sherpa_onnx::OfflineRecognitionResult result =
stream->impl->GetResult();
const auto& text = result.text;
auto r = new SherpaOnnxOfflineRecognizerResult;
r->text = new char[text.size() + 1];
std::copy(text.begin(), text.end(), const_cast<char*>(r->text));
const_cast<char*>(r->text)[text.size()] = 0;
r->text_len = text.size();
return r;
}
/// Free a pointer returned by CreateOfflineRecognizer()
///
/// @param p A pointer returned by CreateOfflineRecognizer()
void __stdcall DestroyOfflineRecognizer(
SherpaOnnxOfflineRecognizer* recognizer) {
delete recognizer->impl;
delete recognizer;
}
/// Destory an offline stream.
///
/// @param stream A pointer returned by CreateOfflineStream()
void __stdcall DestroyOfflineStream(SherpaOnnxOfflineStream* stream) {
delete stream;
}
/// Destroy the pointer returned by GetOfflineStreamResult().
///
/// @param r A pointer returned by GetOfflineStreamResult()
void __stdcall DestroyOfflineRecognizerResult(
SherpaOnnxOfflineRecognizerResult* r) {
delete r->text;
delete r;
}
}// namespace sherpa_onnx
\ No newline at end of file
// sherpa-onnx/sharp-api/offline-api.h
//
// Copyright (c) 2023 Manyeyes Corporation
#pragma once
#include <list>
namespace sherpa_onnx
{
/// Please refer to
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
/// to download pre-trained models. That is, you can find encoder-xxx.onnx
/// decoder-xxx.onnx, joiner-xxx.onnx, and tokens.txt for this struct
/// from there.
typedef struct SherpaOnnxOfflineTransducer {
const char* encoder_filename;
const char* decoder_filename;
const char* joiner_filename;
} SherpaOnnxOfflineTransducer;
typedef struct SherpaOnnxOfflineParaformer {
const char* model;
}SherpaOnnxOfflineParaformer;
typedef struct SherpaOnnxOfflineNemoEncDecCtc {
const char* model;
}SherpaOnnxOfflineNemoEncDecCtc;
typedef struct SherpaOnnxOfflineModelConfig {
SherpaOnnxOfflineTransducer transducer;
SherpaOnnxOfflineParaformer paraformer;
SherpaOnnxOfflineNemoEncDecCtc nemo_ctc;
const char* tokens;
const int32_t num_threads;
const bool debug;
} SherpaOnnxOfflineModelConfig;
/// It expects 16 kHz 16-bit single channel wave format.
typedef struct SherpaOnnxFeatureConfig {
/// Sample rate of the input data. MUST match the one expected
/// by the model. For instance, it should be 16000 for models provided
/// by us.
int32_t sample_rate;
/// Feature dimension of the model.
/// For instance, it should be 80 for models provided by us.
int32_t feature_dim;
} SherpaOnnxFeatureConfig;
typedef struct SherpaOnnxOfflineRecognizerConfig {
SherpaOnnxFeatureConfig feat_config;
SherpaOnnxOfflineModelConfig model_config;
/// Possible values are: greedy_search, modified_beam_search
const char* decoding_method;
} SherpaOnnxOfflineRecognizerConfig;
typedef struct SherpaOnnxOfflineRecognizerResult {
// Recognition results.
// For English, it consists of space separated words.
// For Chinese, it consists of Chinese words without spaces.
char* text;
int text_len;
// Decoded results at the token level.
// For instance, for BPE-based models it consists of a list of BPE tokens.
// std::vector<std::string> tokens;
// timestamps.size() == tokens.size()
// timestamps[i] records the time in seconds when tokens[i] is decoded.
// std::vector<float> timestamps;
} SherpaOnnxOfflineRecognizerResult;
/// Note: OfflineRecognizer here means StreamingRecognizer.
/// It does not need to access the Internet during recognition.
/// Everything is run locally.
typedef struct SherpaOnnxOfflineRecognizer SherpaOnnxOfflineRecognizer;
typedef struct SherpaOnnxOfflineStream SherpaOnnxOfflineStream;
extern "C" __declspec(dllexport)
SherpaOnnxOfflineRecognizer * __stdcall CreateOfflineRecognizer(
const SherpaOnnxOfflineRecognizerConfig * config);
extern "C" __declspec(dllexport)
SherpaOnnxOfflineStream * __stdcall CreateOfflineStream(
SherpaOnnxOfflineRecognizer * sherpaOnnxOfflineRecognizer);
extern "C" __declspec(dllexport)
void __stdcall AcceptWaveform(
SherpaOnnxOfflineStream * stream, int32_t sample_rate,
const float* samples, int32_t samples_size);
extern "C" __declspec(dllexport)
void __stdcall DecodeOfflineStream(
SherpaOnnxOfflineRecognizer * recognizer,
SherpaOnnxOfflineStream * stream);
extern "C" __declspec(dllexport)
void __stdcall DecodeMultipleOfflineStreams(
SherpaOnnxOfflineRecognizer * recognizer,
SherpaOnnxOfflineStream * *streams, int32_t n);
extern "C" __declspec(dllexport)
SherpaOnnxOfflineRecognizerResult * __stdcall GetOfflineStreamResult(
SherpaOnnxOfflineStream * stream);
extern "C" __declspec(dllexport)
void __stdcall DestroyOfflineRecognizer(
SherpaOnnxOfflineRecognizer * recognizer);
extern "C" __declspec(dllexport)
void __stdcall DestroyOfflineStream(
SherpaOnnxOfflineStream * stream);
extern "C" __declspec(dllexport)
void __stdcall DestroyOfflineRecognizerResult(
SherpaOnnxOfflineRecognizerResult * r);
}// namespace sherpa_onnx
\ No newline at end of file
// sherpa-onnx/cpp-api/c-api.cc
//
// Copyright (c) 2023 Xiaomi Corporation
#include "online-api.h"
#include <algorithm>
#include <memory>
#include <utility>
#include <vector>
#include "../../sherpa-onnx/csrc/display.h"
#include "../../sherpa-onnx/csrc/online-recognizer.h"
namespace sherpa_onnx
{
struct SherpaOnnxOnlineRecognizer {
sherpa_onnx::OnlineRecognizer* impl;
};
struct SherpaOnnxOnlineStream {
std::unique_ptr<sherpa_onnx::OnlineStream> impl;
explicit SherpaOnnxOnlineStream(std::unique_ptr<sherpa_onnx::OnlineStream> p)
: impl(std::move(p)) {}
};
struct SherpaOnnxDisplay {
std::unique_ptr<sherpa_onnx::Display> impl;
};
SherpaOnnxOnlineRecognizer* __stdcall CreateOnlineRecognizer(
const SherpaOnnxOnlineRecognizerConfig* config) {
sherpa_onnx::OnlineRecognizerConfig recognizer_config;
recognizer_config.feat_config.sampling_rate = config->feat_config.sample_rate;
recognizer_config.feat_config.feature_dim = config->feat_config.feature_dim;
recognizer_config.model_config.encoder_filename =
config->model_config.transducer.encoder;
recognizer_config.model_config.decoder_filename =
config->model_config.transducer.decoder;
recognizer_config.model_config.joiner_filename = config->model_config.transducer.joiner;
recognizer_config.model_config.tokens = config->model_config.tokens;
recognizer_config.model_config.num_threads = config->model_config.num_threads;
recognizer_config.model_config.debug = config->model_config.debug;
recognizer_config.decoding_method = config->decoding_method;
recognizer_config.max_active_paths = config->max_active_paths;
recognizer_config.enable_endpoint = config->enable_endpoint;
recognizer_config.endpoint_config.rule1.min_trailing_silence =
config->rule1_min_trailing_silence;
recognizer_config.endpoint_config.rule2.min_trailing_silence =
config->rule2_min_trailing_silence;
recognizer_config.endpoint_config.rule3.min_utterance_length =
config->rule3_min_utterance_length;
SherpaOnnxOnlineRecognizer* recognizer = new SherpaOnnxOnlineRecognizer;
recognizer->impl = new sherpa_onnx::OnlineRecognizer(recognizer_config);
return recognizer;
}
void __stdcall DestroyOnlineRecognizer(SherpaOnnxOnlineRecognizer* recognizer) {
delete recognizer->impl;
delete recognizer;
}
SherpaOnnxOnlineStream* __stdcall CreateOnlineStream(
const SherpaOnnxOnlineRecognizer* recognizer) {
SherpaOnnxOnlineStream* stream =
new SherpaOnnxOnlineStream(recognizer->impl->CreateStream());
return stream;
}
void __stdcall DestroyOnlineStream(SherpaOnnxOnlineStream* stream) { delete stream; }
void __stdcall AcceptOnlineWaveform(SherpaOnnxOnlineStream* stream, int32_t sample_rate,
const float* samples, int32_t n) {
stream->impl->AcceptWaveform(sample_rate, samples, n);
}
int32_t __stdcall IsOnlineStreamReady(SherpaOnnxOnlineRecognizer* recognizer,
SherpaOnnxOnlineStream* stream) {
return recognizer->impl->IsReady(stream->impl.get());
}
void __stdcall DecodeOnlineStream(SherpaOnnxOnlineRecognizer* recognizer,
SherpaOnnxOnlineStream* stream) {
recognizer->impl->DecodeStream(stream->impl.get());
}
void __stdcall DecodeMultipleOnlineStreams(SherpaOnnxOnlineRecognizer* recognizer,
SherpaOnnxOnlineStream** streams, int32_t n) {
std::vector<sherpa_onnx::OnlineStream*> ss(n);
for (int32_t i = 0; i != n; ++i) {
ss[i] = streams[i]->impl.get();
}
recognizer->impl->DecodeStreams(ss.data(), n);
}
SherpaOnnxOnlineRecognizerResult* __stdcall GetOnlineStreamResult(
SherpaOnnxOnlineRecognizer* recognizer, SherpaOnnxOnlineStream* stream) {
sherpa_onnx::OnlineRecognizerResult result =
recognizer->impl->GetResult(stream->impl.get());
const auto& text = result.text;
auto r = new SherpaOnnxOnlineRecognizerResult;
r->text = new char[text.size() + 1];
std::copy(text.begin(), text.end(), const_cast<char*>(r->text));
const_cast<char*>(r->text)[text.size()] = 0;
r->text_len = text.size();
return r;
}
void __stdcall DestroyOnlineRecognizerResult(const SherpaOnnxOnlineRecognizerResult* r) {
delete[] r->text;
delete r;
}
void __stdcall Reset(SherpaOnnxOnlineRecognizer* recognizer,
SherpaOnnxOnlineStream* stream) {
recognizer->impl->Reset(stream->impl.get());
}
void __stdcall InputFinished(SherpaOnnxOnlineStream* stream) {
stream->impl->InputFinished();
}
int32_t __stdcall IsEndpoint(SherpaOnnxOnlineRecognizer* recognizer,
SherpaOnnxOnlineStream* stream) {
return recognizer->impl->IsEndpoint(stream->impl.get());
}
SherpaOnnxDisplay* __stdcall CreateDisplay(int32_t max_word_per_line) {
SherpaOnnxDisplay* ans = new SherpaOnnxDisplay;
ans->impl = std::make_unique<sherpa_onnx::Display>(max_word_per_line);
return ans;
}
void __stdcall DestroyDisplay(SherpaOnnxDisplay* display) { delete display; }
void __stdcall SherpaOnnxPrint(SherpaOnnxDisplay* display, int32_t idx, const char* s) {
display->impl->Print(idx, s);
}
}
\ No newline at end of file
// sherpa-onnx/cpp-api/c-api.h
//
// Copyright (c) 2023 Xiaomi Corporation
// C API for sherpa-onnx
//
// Please refer to
// https://github.com/k2-fsa/sherpa-onnx/blob/master/c-api-examples/decode-file-c-api.c
// for usages.
//
#ifndef SHERPA_ONNX_CPP_API_C_API_H_
#define SHERPA_ONNX_CPP_API_C_API_H_
#include <stdint.h>
#ifdef __cplusplus
extern "C" {
#endif
namespace sherpa_onnx
{
/// Please refer to
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
/// to download pre-trained models. That is, you can find encoder-xxx.onnx
/// decoder-xxx.onnx, joiner-xxx.onnx, and tokens.txt for this struct
/// from there.
typedef struct SherpaOnnxOnlineTransducer {
const char* encoder;
const char* decoder;
const char* joiner;
} SherpaOnnxOnlineTransducer;
typedef struct SherpaOnnxOnlineModelConfig
{
const SherpaOnnxOnlineTransducer transducer;
const char* tokens;
const int32_t num_threads;
const bool debug; // true to print debug information of the model
}SherpaOnnxOnlineModelConfig;
/// It expects 16 kHz 16-bit single channel wave format.
typedef struct SherpaOnnxFeatureConfig {
/// Sample rate of the input data. MUST match the one expected
/// by the model. For instance, it should be 16000 for models provided
/// by us.
int32_t sample_rate;
/// Feature dimension of the model.
/// For instance, it should be 80 for models provided by us.
int32_t feature_dim;
} SherpaOnnxFeatureConfig;
typedef struct SherpaOnnxOnlineRecognizerConfig {
SherpaOnnxFeatureConfig feat_config;
SherpaOnnxOnlineModelConfig model_config;
/// Possible values are: greedy_search, modified_beam_search
const char* decoding_method;
/// Used only when decoding_method is modified_beam_search
/// Example value: 4
int32_t max_active_paths;
/// 0 to disable endpoint detection.
/// A non-zero value to enable endpoint detection.
int enable_endpoint;
/// An endpoint is detected if trailing silence in seconds is larger than
/// this value even if nothing has been decoded.
/// Used only when enable_endpoint is not 0.
float rule1_min_trailing_silence;
/// An endpoint is detected if trailing silence in seconds is larger than
/// this value after something that is not blank has been decoded.
/// Used only when enable_endpoint is not 0.
float rule2_min_trailing_silence;
/// An endpoint is detected if the utterance in seconds is larger than
/// this value.
/// Used only when enable_endpoint is not 0.
float rule3_min_utterance_length;
} SherpaOnnxOnlineRecognizerConfig;
typedef struct SherpaOnnxOnlineRecognizerResult {
const char* text;
int text_len;
// TODO(fangjun): Add more fields
} SherpaOnnxOnlineRecognizerResult;
/// Note: OnlineRecognizer here means StreamingRecognizer.
/// It does not need to access the Internet during recognition.
/// Everything is run locally.
typedef struct SherpaOnnxOnlineRecognizer SherpaOnnxOnlineRecognizer;
typedef struct SherpaOnnxOnlineStream SherpaOnnxOnlineStream;
/// @param config Config for the recongizer.
/// @return Return a pointer to the recognizer. The user has to invoke
// DestroyOnlineRecognizer() to free it to avoid memory leak.
extern "C" __declspec(dllexport)
SherpaOnnxOnlineRecognizer* __stdcall CreateOnlineRecognizer(
const SherpaOnnxOnlineRecognizerConfig * config);
/// Free a pointer returned by CreateOnlineRecognizer()
///
/// @param p A pointer returned by CreateOnlineRecognizer()
extern "C" __declspec(dllexport)
void __stdcall DestroyOnlineRecognizer(SherpaOnnxOnlineRecognizer* recognizer);
/// Create an online stream for accepting wave samples.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer()
/// @return Return a pointer to an OnlineStream. The user has to invoke
/// DestroyOnlineStream() to free it to avoid memory leak.
extern "C" __declspec(dllexport)
SherpaOnnxOnlineStream* __stdcall CreateOnlineStream(
const SherpaOnnxOnlineRecognizer* recognizer);
/// Destroy an online stream.
///
/// @param stream A pointer returned by CreateOnlineStream()
extern "C" __declspec(dllexport)
void __stdcall DestroyOnlineStream(SherpaOnnxOnlineStream* stream);
/// Accept input audio samples and compute the features.
/// The user has to invoke DecodeOnlineStream() to run the neural network and
/// decoding.
///
/// @param stream A pointer returned by CreateOnlineStream().
/// @param sample_rate Sample rate of the input samples. If it is different
/// from config.feat_config.sample_rate, we will do
/// resampling inside sherpa-onnx.
/// @param samples A pointer to a 1-D array containing audio samples.
/// The range of samples has to be normalized to [-1, 1].
/// @param n Number of elements in the samples array.
extern "C" __declspec(dllexport)
void __stdcall AcceptOnlineWaveform(SherpaOnnxOnlineStream* stream, int32_t sample_rate,
const float* samples, int32_t n);
/// Return 1 if there are enough number of feature frames for decoding.
/// Return 0 otherwise.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer
/// @param stream A pointer returned by CreateOnlineStream
extern "C" __declspec(dllexport)
int32_t __stdcall IsOnlineStreamReady(SherpaOnnxOnlineRecognizer* recognizer,
SherpaOnnxOnlineStream* stream);
/// Call this function to run the neural network model and decoding.
//
/// Precondition for this function: IsOnlineStreamReady() MUST return 1.
///
/// Usage example:
///
/// while (IsOnlineStreamReady(recognizer, stream)) {
/// DecodeOnlineStream(recognizer, stream);
/// }
///
extern "C" __declspec(dllexport)
void __stdcall DecodeOnlineStream(SherpaOnnxOnlineRecognizer* recognizer,
SherpaOnnxOnlineStream* stream);
/// This function is similar to DecodeOnlineStream(). It decodes multiple
/// OnlineStream in parallel.
///
/// Caution: The caller has to ensure each OnlineStream is ready, i.e.,
/// IsOnlineStreamReady() for that stream should return 1.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer()
/// @param streams A pointer array containing pointers returned by
/// CreateOnlineRecognizer()
/// @param n Number of elements in the given streams array.
extern "C" __declspec(dllexport)
void __stdcall DecodeMultipleOnlineStreams(SherpaOnnxOnlineRecognizer* recognizer,
SherpaOnnxOnlineStream** streams, int32_t n);
/// Get the decoding results so far for an OnlineStream.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer().
/// @param stream A pointer returned by CreateOnlineStream().
/// @return A pointer containing the result. The user has to invoke
/// DestroyOnlineRecognizerResult() to free the returned pointer to
/// avoid memory leak.
extern "C" __declspec(dllexport)
SherpaOnnxOnlineRecognizerResult* __stdcall GetOnlineStreamResult(
SherpaOnnxOnlineRecognizer* recognizer, SherpaOnnxOnlineStream* stream);
/// Destroy the pointer returned by GetOnlineStreamResult().
///
/// @param r A pointer returned by GetOnlineStreamResult()
extern "C" __declspec(dllexport)
void __stdcall DestroyOnlineRecognizerResult(const SherpaOnnxOnlineRecognizerResult* r);
/// Reset an OnlineStream , which clears the neural network model state
/// and the state for decoding.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer().
/// @param stream A pointer returned by CreateOnlineStream
extern "C" __declspec(dllexport)
void __stdcall Reset(SherpaOnnxOnlineRecognizer* recognizer,
SherpaOnnxOnlineStream* stream);
/// Signal that no more audio samples would be available.
/// After this call, you cannot call AcceptWaveform() any more.
///
/// @param stream A pointer returned by CreateOnlineStream()
extern "C" __declspec(dllexport)
void __stdcall InputFinished(SherpaOnnxOnlineStream* stream);
/// Return 1 if an endpoint has been detected.
///
/// @param recognizer A pointer returned by CreateOnlineRecognizer()
/// @param stream A pointer returned by CreateOnlineStream()
/// @return Return 1 if an endpoint is detected. Return 0 otherwise.
extern "C" __declspec(dllexport)
int32_t __stdcall IsEndpoint(SherpaOnnxOnlineRecognizer* recognizer,
SherpaOnnxOnlineStream* stream);
// for displaying results on Linux/macOS.
typedef struct SherpaOnnxDisplay SherpaOnnxDisplay;
/// Create a display object. Must be freed using DestroyDisplay to avoid
/// memory leak.
extern "C" __declspec(dllexport)
SherpaOnnxDisplay* __stdcall CreateDisplay(int32_t max_word_per_line);
extern "C" __declspec(dllexport)
void __stdcall DestroyDisplay(SherpaOnnxDisplay* display);
/// Print the result.
extern "C" __declspec(dllexport)
void __stdcall SherpaOnnxPrint(SherpaOnnxDisplay* display, int32_t idx, const char* s);
}
#ifdef __cplusplus
} /* extern "C" */
#endif
#endif // SHERPA_ONNX_C_API_C_API_H_