Remove single file deployment for streaming instantiation and smaller size.
This commit is contained in:
@@ -6,7 +6,7 @@ using namespace emscripten;
|
||||
|
||||
EMSCRIPTEN_BINDINGS() {
|
||||
function("setLogLevel", &vosk_set_log_level, allow_raw_pointers());
|
||||
|
||||
|
||||
enum_<VoskEndpointerMode>("EpMode")
|
||||
.value("ANSWER_DEFAULT", VOSK_EP_ANSWER_DEFAULT)
|
||||
.value("ANSWER_SHORT", VOSK_EP_ANSWER_SHORT)
|
||||
@@ -16,8 +16,8 @@ EMSCRIPTEN_BINDINGS() {
|
||||
class_<CommonModel>("CommonModel")
|
||||
.constructor<int, bool, std::string, std::string, int, int>(allow_raw_pointers())
|
||||
.function("findWord", &CommonModel::findWord, allow_raw_pointers());
|
||||
|
||||
class_<Recognizer>("Recognizer")
|
||||
|
||||
class_<Recognizer>("Recognizer")
|
||||
.constructor<int, float, CommonModel*>(allow_raw_pointers())
|
||||
.constructor<int, float, CommonModel*, CommonModel*>(allow_raw_pointers())
|
||||
.constructor<int, float, CommonModel*, std::string, int>(allow_raw_pointers())
|
||||
|
||||
@@ -1,9 +1,9 @@
|
||||
#include "CommonModel.h"
|
||||
|
||||
CommonModel::CommonModel(int index, bool normalMdl, std::string storepath, std::string id, int tarStart, int tarSize) :
|
||||
normalMdl{normalMdl}, index{index},
|
||||
storepath{std::move(storepath)},
|
||||
id{std::move(id)}
|
||||
CommonModel::CommonModel(int index, bool normalMdl, std::string storepath, std::string id, int tarStart, int tarSize) :
|
||||
normalMdl{normalMdl}, index{index},
|
||||
storepath{std::move(storepath)},
|
||||
id{std::move(id)}
|
||||
{
|
||||
globalPool.exec([this, tarStart, tarSize]{
|
||||
extractAndLoad(reinterpret_cast<unsigned char*>(tarStart), tarSize);
|
||||
@@ -22,12 +22,12 @@ void CommonModel::extractAndLoad(unsigned char* tar, int tarSize) {
|
||||
case FailedOpen:
|
||||
fireEv(index, "Untar: Unable to open file for write");
|
||||
return;
|
||||
case FailedWrite:
|
||||
case FailedWrite:
|
||||
fireEv(index, "Untar: Unable to write file");
|
||||
return;
|
||||
case FailedClose:
|
||||
fireEv(index, "Untar: Unable to close file after write");
|
||||
return;
|
||||
return;
|
||||
};
|
||||
if(normalMdl) mdl = vosk_model_new(storepath.c_str());
|
||||
else mdl = vosk_spk_model_new(storepath.c_str());
|
||||
|
||||
@@ -2,6 +2,7 @@
|
||||
#include "Util.h"
|
||||
|
||||
#include <vosk_api.h>
|
||||
|
||||
struct CommonModel {
|
||||
bool normalMdl;
|
||||
int index;
|
||||
@@ -12,5 +13,4 @@ struct CommonModel {
|
||||
int findWord(std::string word);
|
||||
CommonModel(int index, bool normalMdl, std::string storepath, std::string id, int tarStart, int tarSize);
|
||||
~CommonModel();
|
||||
};
|
||||
|
||||
};
|
||||
@@ -1,18 +1,18 @@
|
||||
#include "Recognizer.h"
|
||||
#include "Recognizer.h"
|
||||
#include "emscripten/atomic.h"
|
||||
Recognizer::Recognizer(int index, float sampleRate, CommonModel* model) :
|
||||
index{index},
|
||||
Recognizer::Recognizer(int index, float sampleRate, CommonModel* model) :
|
||||
index{index},
|
||||
rec{vosk_recognizer_new(std::get<VoskModel*>(model->mdl), sampleRate)}
|
||||
{
|
||||
finishConstruction(model);
|
||||
}
|
||||
Recognizer::Recognizer(int index, float sampleRate, CommonModel* model, CommonModel* spkModel) :
|
||||
Recognizer::Recognizer(int index, float sampleRate, CommonModel* model, CommonModel* spkModel) :
|
||||
index{index},
|
||||
rec{vosk_recognizer_new_spk(std::get<VoskModel*>(model->mdl), sampleRate, std::get<VoskSpkModel*>(spkModel->mdl))} {
|
||||
finishConstruction(model, spkModel);
|
||||
}
|
||||
Recognizer::Recognizer(int index, float sampleRate, CommonModel* model, const std::string& grm, int) :
|
||||
index{index},
|
||||
Recognizer::Recognizer(int index, float sampleRate, CommonModel* model, const std::string& grm, int) :
|
||||
index{index},
|
||||
rec{vosk_recognizer_new_grm(std::get<VoskModel*>(model->mdl), sampleRate, grm.c_str())} {
|
||||
finishConstruction(model);
|
||||
}
|
||||
|
||||
@@ -13,7 +13,7 @@ int untar(unsigned char* tar, int tarSize, const std::string& storepath) {
|
||||
path.reserve(100); // Max length
|
||||
unsigned char* end = tar + tarSize;
|
||||
while(tar <= end) {
|
||||
if(tar[156] != '5' && tar[156] != 0 &&
|
||||
if(tar[156] != '5' && tar[156] != 0 &&
|
||||
tar[156] != '0') {
|
||||
return IncorrectFiletype;
|
||||
}
|
||||
@@ -64,7 +64,7 @@ void Thread::startup(ThreadPool* pool) {
|
||||
emscripten_atomic_store_u32(&pool->qLock, false);
|
||||
emscripten_atomic_notify(&pool->qLock, 1);
|
||||
fn();
|
||||
}
|
||||
}
|
||||
}
|
||||
ThreadPool::ThreadPool() {
|
||||
for(Thread& thrd : threads) {
|
||||
|
||||
@@ -8,7 +8,7 @@
|
||||
#include <emscripten/atomic.h>
|
||||
#include <emscripten/console.h>
|
||||
namespace fs = std::filesystem;
|
||||
|
||||
|
||||
struct AudioData {
|
||||
float* data;
|
||||
int len;
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
let objs = []
|
||||
let objs = []
|
||||
let processorURL = URL.createObjectURL(new Blob(['(', (() => {
|
||||
registerProcessor("VoskletTransferer", class extends AudioWorkletProcessor {
|
||||
constructor(opts) {
|
||||
@@ -28,8 +28,8 @@ Module.cleanUp = () => {
|
||||
|
||||
Module.createTransferer = async (ctx, bufferSize) => {
|
||||
await ctx.audioWorklet.addModule(processorURL)
|
||||
return new AudioWorkletNode(ctx, "VoskletTransferer", {
|
||||
channelCountMode : "explicit",
|
||||
return new AudioWorkletNode(ctx, "VoskletTransferer", {
|
||||
channelCountMode : "explicit",
|
||||
numberOfInputs : 1,
|
||||
numberOfOutputs : 0,
|
||||
channelCount : 1,
|
||||
@@ -37,7 +37,7 @@ Module.createTransferer = async (ctx, bufferSize) => {
|
||||
})
|
||||
}
|
||||
|
||||
async function getFileHandle(path, create = false) {
|
||||
getFileHandle = async (path, create = false) => {
|
||||
let components = path.split("/")
|
||||
let prevDir = await navigator.storage.getDirectory()
|
||||
for(let component of components.slice(0, -1)) prevDir = await prevDir.getDirectoryHandle(component, { create : create })
|
||||
@@ -65,7 +65,7 @@ class CommonModel extends EventTarget {
|
||||
let dataFile = await (await getFileHandle(storepath + "/model.tgz")).getFile()
|
||||
let idFile = await (await getFileHandle(storepath + "/id")).getFile()
|
||||
if(await idFile.text() != id) throw ""
|
||||
tar = await new Response(dataFile.stream().pipeThrough(new DecompressionStream("gzip"))).arrayBuffer()
|
||||
tar = await new Response(dataFile.stream().pipeThrough(new DecompressionStream("gzip"))).arrayBuffer()
|
||||
}
|
||||
catch {
|
||||
try {
|
||||
@@ -105,7 +105,7 @@ class Recognizer extends EventTarget {
|
||||
super()
|
||||
objs.push(this)
|
||||
return new Proxy(this, {
|
||||
get(self, prop, receiver) {
|
||||
get(self, prop, _) {
|
||||
return self.obj && Object.keys(Object.getPrototypeOf(self.obj)).includes(prop) ? self.obj[prop].bind(self.obj) : self[prop] ? self[prop].bind ? self[prop].bind(self) : self[prop] : undefined
|
||||
}
|
||||
})
|
||||
@@ -123,13 +123,13 @@ class Recognizer extends EventTarget {
|
||||
rec.obj = new Module.Recognizer(objs.length - 1, sampleRate, model)
|
||||
break
|
||||
case 2:
|
||||
rec.obj = new Module.Recognizer(objs.length -1, sampleRate, model, spkModel)
|
||||
rec.obj = new Module.Recognizer(objs.length -1, sampleRate, model, spkModel)
|
||||
break
|
||||
default:
|
||||
rec.obj = new Module.Recognizer(objs.length - 1, sampleRate, model, grammar, 0)
|
||||
rec.obj = new Module.Recognizer(objs.length - 1, sampleRate, model, grammar, 0)
|
||||
}
|
||||
return result
|
||||
}
|
||||
}
|
||||
acceptWaveform(audioData) {
|
||||
let start = Module._malloc(audioData.length * 4)
|
||||
Module.HEAPF32.set(audioData, start / 4)
|
||||
|
||||
61
src/make
61
src/make
@@ -4,15 +4,15 @@ MAX_THREADS=${MAX_THREADS:-1}
|
||||
EMSDK=${EMSDK:-../emsdk}
|
||||
JOBS=${JOBS:-$(nproc)}
|
||||
|
||||
if [ $EMSDK != ../emsdk ] && [ ! -d $EMSDK ]; then
|
||||
if [ "$EMSDK" != ../emsdk ] && [ ! -d "$EMSDK" ]; then
|
||||
echo "Invalid emsdk path"
|
||||
exit 1
|
||||
fi
|
||||
if [ $MAX_THREADS -lt 1 ]; then
|
||||
fi
|
||||
if [ "$MAX_THREADS" -lt 1 ]; then
|
||||
echo "MAX_THREADS must be greater than or equal to 1"
|
||||
exit 1
|
||||
fi
|
||||
if [ $JOBS -lt 1 ]; then
|
||||
if [ "$JOBS" -lt 1 ]; then
|
||||
echo "JOBS must be greater than or equal to 1"
|
||||
exit 1
|
||||
fi
|
||||
@@ -20,15 +20,15 @@ if ! [[ $INITIAL_MEMORY =~ ^[0-9]+([kmgt]b)?$ ]]; then
|
||||
echo "INITIAL_MEMORY valid suffixes are kb, mb, gb, tb, none (bytes)"
|
||||
exit 1
|
||||
fi
|
||||
if [ $EMSDK = ../emsdk ] && [ ! -d $EMSDK ]; then
|
||||
if [ "$EMSDK" = ../emsdk ] && [ ! -d "$EMSDK" ]; then
|
||||
echo "Installing emsdk + Emscripten..."
|
||||
git clone --depth=1 https://github.com/emscripten-core/emsdk.git ../emsdk &&
|
||||
cd ../emsdk &&
|
||||
./emsdk install 3.1.65 &&
|
||||
./emsdk activate 3.1.65
|
||||
fi
|
||||
. $(realpath $EMSDK)/emsdk_env.sh &&
|
||||
export PATH=:$PATH:$(realpath $EMSDK)/upstream/bin &&
|
||||
. $(realpath "$EMSDK")/emsdk_env.sh &&
|
||||
export PATH=:$PATH:$(realpath "$EMSDK")/upstream/bin &&
|
||||
|
||||
cd .. &&
|
||||
SRC=$(realpath src)
|
||||
@@ -37,47 +37,48 @@ VOSK=$(realpath vosk)
|
||||
OPENFST=$(realpath openfst)
|
||||
CLAPACK_WASM=$(realpath clapack-wasm)
|
||||
|
||||
if [ ! -d $OPENFST ]; then
|
||||
if [ ! -d "$OPENFST" ]; then
|
||||
rm -rf /tmp/openfst &&
|
||||
git clone --depth=1 https://github.com/alphacep/openfst /tmp/openfst &&
|
||||
cd /tmp/openfst &&
|
||||
git apply $SRC/Openfst.patch
|
||||
git apply "$SRC"/Openfst.patch
|
||||
autoreconf -is &&
|
||||
CXXFLAGS="-r -O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals" LDFLAGS="-O3 -flto" emconfigure ./configure --prefix=$OPENFST --enable-static --disable-shared --enable-lookahead-fsts --enable-ngram-fsts --disable-bin &&
|
||||
emmake make -j$JOBS install &&
|
||||
echo "PACKAGE_VERSION = 1.8.0" >> $OPENFST/Makefile
|
||||
CXXFLAGS="-r -O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals" LDFLAGS="-O3 -flto" emconfigure ./configure --prefix="$OPENFST" --enable-static --disable-shared --enable-lookahead-fsts --enable-ngram-fsts --disable-bin &&
|
||||
emmake make -j"$JOBS" install &&
|
||||
echo "PACKAGE_VERSION = 1.8.0" >> "$OPENFST"/Makefile
|
||||
fi
|
||||
|
||||
if [ ! -d $CLAPACK_WASM ]; then
|
||||
git clone --depth=1 https://gitlab.inria.fr/multispeech/kaldi.web/clapack-wasm.git $CLAPACK_WASM &&
|
||||
cd $CLAPACK_WASM &&
|
||||
git apply $SRC/Clapack-wasm.patch &&
|
||||
if [ ! -d "$CLAPACK_WASM" ]; then
|
||||
git clone --depth=1 https://gitlab.inria.fr/multispeech/kaldi.web/clapack-wasm.git "$CLAPACK_WASM" &&
|
||||
cd "$CLAPACK_WASM" &&
|
||||
git apply "$SRC"/Clapack-wasm.patch &&
|
||||
bash install_repo.sh emcc
|
||||
fi
|
||||
|
||||
if [ ! -d $KALDI ]; then
|
||||
git clone -b vosk --depth=1 https://github.com/alphacep/kaldi $KALDI &&
|
||||
cd $KALDI/src &&
|
||||
git apply $SRC/Kaldi.patch &&
|
||||
CXXFLAGS="-O3 -UHAVE_EXECINFO_H -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals -Wno-unused-variable -Wno-unused-but-set-variable -g0" LDFLAGS="-O3 -lembind -flto -g0" emconfigure ./configure --use-cuda=no --with-cudadecoder=no --static --static-math=yes --static-fst=yes --debug-level=0 --fst-root=$OPENFST --clapack-root=$CLAPACK_WASM --host=WASM &&
|
||||
emmake make -j$JOBS online2 rnnlm
|
||||
if [ ! -d "$KALDI" ]; then
|
||||
git clone -b vosk --depth=1 https://github.com/alphacep/kaldi "$KALDI" &&
|
||||
cd "$KALDI"/src &&
|
||||
git apply "$SRC"/Kaldi.patch &&
|
||||
CXXFLAGS="-O3 -UHAVE_EXECINFO_H -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals -Wno-unused-variable -Wno-unused-but-set-variable -g0" LDFLAGS="-O3 -lembind -flto -g0" emconfigure ./configure --use-cuda=no --with-cudadecoder=no --static --static-math=yes --static-fst=yes --debug-level=0 --fst-root="$OPENFST" --clapack-root="$CLAPACK_WASM" --host=WASM &&
|
||||
emmake make -j"$JOBS" online2 rnnlm
|
||||
fi
|
||||
|
||||
if [ ! -d $VOSK ]; then
|
||||
git clone -b v0.3.50 --depth=1 https://github.com/alphacep/vosk-api $VOSK &&
|
||||
cd $VOSK/src &&
|
||||
git apply $SRC/Vosk.patch &&
|
||||
if [ ! -d "$VOSK" ]; then
|
||||
git clone -b v0.3.50 --depth=1 https://github.com/alphacep/vosk-api "$VOSK" &&
|
||||
cd "$VOSK"/src &&
|
||||
git apply "$SRC"/Vosk.patch &&
|
||||
VOSK_FILES="Recognizer.cc language_model.cc model.cc spk_model.cc vosk_api.cc" &&
|
||||
em++ -O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals -Wno-deprecated -I. -I$KALDI/src -I$OPENFST/include $VOSK_FILES -c &&
|
||||
emar -rcs vosk.a ${VOSK_FILES//.cc/.o}
|
||||
em++ -O3 -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals -Wno-deprecated -I. -I"$KALDI"/src -I"$OPENFST"/include "$VOSK_FILES" -c &&
|
||||
emar -rcs vosk.a "${VOSK_FILES//.cc/.o}"
|
||||
fi
|
||||
|
||||
cd $SRC &&
|
||||
em++ Util.cc CommonModel.cc Recognizer.cc Bindings.cc -O3 -Wno-pthreads-mem-growth -DEMSCRIPTEN_HAS_UNBOUND_TYPE_NAMES=0 -fno-rtti -DMAX_THREADS=$MAX_THREADS -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY=$INITIAL_MEMORY -sALLOW_MEMORY_GROWTH -sPTHREAD_POOL_SIZE=$MAX_THREADS -sPOLYFILL=0 -sEXIT_RUNTIME=0 -sINVOKE_RUN=0 -sSUPPORT_LONGJMP=0 -sALLOW_BLOCKING_ON_MAIN_THREAD=0 -sEXPORTED_FUNCTIONS=_malloc -sEXPORT_NAME=loadVosklet -sMALLOC=emmalloc -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sENVIRONMENT=web,worker -I. -I$VOSK/src -L$KALDI/src -l:online2/kaldi-online2.a -l:decoder/kaldi-decoder.a -l:ivector/kaldi-ivector.a -l:gmm/kaldi-gmm.a -l:tree/kaldi-tree.a -l:feat/kaldi-feat.a -l:cudamatrix/kaldi-cudamatrix.a -l:lat/kaldi-lat.a -l:lm/kaldi-lm.a -l:rnnlm/kaldi-rnnlm.a -l:hmm/kaldi-hmm.a -l:nnet3/kaldi-nnet3.a -l:transform/kaldi-transform.a -l:matrix/kaldi-matrix.a -l:fstext/kaldi-fstext.a -l:util/kaldi-util.a -l:base/kaldi-base.a -L$OPENFST/lib -l:libfst.a -l:libfstngram.a -L$CLAPACK_WASM -l:CBLAS/lib/cblas.a -l:CLAPACK-3.2.1/lapack.a -l:CLAPACK-3.2.1/libcblaswr.a -l:f2c_BLAS-3.8.0/blas.a -l:libf2c/libf2c.a -L$VOSK/src -l:vosk.a -lembind -pthread -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals --pre-js Wrapper.js -o ../Vosklet.js &&
|
||||
cd "$SRC" &&
|
||||
em++ Util.cc CommonModel.cc Recognizer.cc Bindings.cc -O3 -Wno-pthreads-mem-growth -DEMSCRIPTEN_HAS_UNBOUND_TYPE_NAMES=0 -fno-rtti -DMAX_THREADS="$MAX_THREADS" -sWASMFS -sWASM_BIGINT -sMODULARIZE -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY="$INITIAL_MEMORY" -sALLOW_MEMORY_GROWTH -sPTHREAD_POOL_SIZE="$MAX_THREADS" -sPOLYFILL=0 -sEXIT_RUNTIME=0 -sINVOKE_RUN=0 -sSUPPORT_LONGJMP=0 -sALLOW_BLOCKING_ON_MAIN_THREAD=0 -sEXPORTED_FUNCTIONS=_malloc -sEXPORT_NAME=loadVosklet -sMALLOC=emmalloc -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sENVIRONMENT=web,worker -I. -I"$VOSK"/src -L"$KALDI"/src -l:online2/kaldi-online2.a -l:decoder/kaldi-decoder.a -l:ivector/kaldi-ivector.a -l:gmm/kaldi-gmm.a -l:tree/kaldi-tree.a -l:feat/kaldi-feat.a -l:cudamatrix/kaldi-cudamatrix.a -l:lat/kaldi-lat.a -l:lm/kaldi-lm.a -l:rnnlm/kaldi-rnnlm.a -l:hmm/kaldi-hmm.a -l:nnet3/kaldi-nnet3.a -l:transform/kaldi-transform.a -l:matrix/kaldi-matrix.a -l:fstext/kaldi-fstext.a -l:util/kaldi-util.a -l:base/kaldi-base.a -L"$OPENFST"/lib -l:libfst.a -l:libfstngram.a -L"$CLAPACK_WASM" -l:CBLAS/lib/cblas.a -l:CLAPACK-3.2.1/lapack.a -l:CLAPACK-3.2.1/libcblaswr.a -l:f2c_BLAS-3.8.0/blas.a -l:libf2c/libf2c.a -L"$VOSK"/src -l:vosk.a -lembind -pthread -flto -msimd128 -mreference-types -mnontrapping-fptoint -mextended-const -msign-ext -mmutable-globals --pre-js Wrapper.js -o ../Vosklet.js &&
|
||||
cd .. &&
|
||||
rm -f Vosklet.worker.js
|
||||
|
||||
cp Vosklet.js Examples/Vosklet.js &&
|
||||
cp Vosklet.wasm Example/Vosklet.wasm &&
|
||||
|
||||
# Can't serve files from raw.githubusercontent with Content-Encoding: gzip header so the browser won't decompress automatically. Manually decompressing instead.
|
||||
sed -i 's/res.body/new Response(res.body.pipeThrough(new DecompressionStream("gzip"))).body/' Examples/Vosklet.js &&
|
||||
|
||||
Reference in New Issue
Block a user