Issue investigation

This commit is contained in:
msqr1
2024-02-05 15:31:21 -08:00
parent 09a4a2844d
commit 6d097a9853
8 changed files with 164 additions and 197 deletions

2
.gitignore vendored
View File

@@ -1,5 +1,5 @@
zstd
vosk-api
vosk
kaldi
minitest
libarchive

View File

@@ -13,15 +13,15 @@ COMPILE_JOBS?:=$(nproc)
SRC:=$(realpath src)
KALDI:=$(realpath kaldi)
VOSK:=$(realpath vosk-api)
VOSK:=$(realpath vosk)
OPENFST:=$(realpath openfst)
LIBARCHIVE:=$(realpath libarchive)
ZSTD:=$(realpath zstd)
CLAPACK_WASM:=$(realpath clapack-wasm)
browser-recognizer: vosk libarchive
BrowserRecognizer.js: | vosk libarchive
cd $(SRC) && \
em++ -O3 global.cc genericModel.cc model.cc spkModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sTRUSTED_TYPES -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY=$(MAX_MEMORY) -sPTHREAD_POOL_SIZE=$(MAX_THREADS) -sPOLYFILL=0 -sSUPPORT_LONGJMP=0 -sEXPORT_NAME=loadBR -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sMALLOC=emmaloc -sEXPORTED_FUNCTIONS=_malloc,_main -sENVIRONMENT=web -I. -I$(LIBARCHIVE)/include -I$(VOSK)/src -L$(LIBARCHIVE)/lib -larchive -L$(ZSTD)/lib -lzstd -L$(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$(VOSK)/src -l:vosk.a -lopfs.js -lembind -pthread -flto --pre-js pre.js ../BrowserRecognizer.js && \
em++ -O3 global.cc genericModel.cc model.cc spkModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sTRUSTED_TYPES -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY=$(MAX_MEMORY) -sPTHREAD_POOL_SIZE=$(MAX_THREADS) -sPOLYFILL=0 -sSUPPORT_LONGJMP=0 -sEXPORT_NAME=loadBR -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sMALLOC=emmalloc -sEXPORTED_FUNCTIONS=_malloc,_main -sENVIRONMENT=web,worker -I. -I$(LIBARCHIVE)/include -I$(VOSK)/src -L$(LIBARCHIVE)/lib -larchive -L$(ZSTD)/lib -lzstd -L$(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 -lopfs.js -lembind -pthread -flto --pre-js pre.js -o ../BrowserRecognizer.js && \
cd .. && \
rm -f BrowserRecognizer.worker.js && \
sed -i 's/locateFile("BrowserRecognizer.worker.js")/pthreadUrl/g' BrowserRecognizer.js && \
@@ -32,7 +32,7 @@ prepare:
[ $(EMSDK) != ../emsdk ! -d $(EMSDK) ] && \
echo "Invalid emsdk path"; \
exit 1; \
[ $(MAX_THREAD) -lt 2 ] && \
[ $(MAX_THREADS) -lt 2 ] && \
echo "MAX_THREADS must be greater than or equal to 2"; \
exit 1; \
[ $(COMPILE_JOBS) -lt 1 ] && \
@@ -57,7 +57,7 @@ zstd: prepare
HAVE_THREAD=0 ZSTD_LEGACY_SUPPORT=0 HAVE_ZLIB=0 HAVE_LZMA=0 HAVE_LZ4=0 ZSTD_NOBENCH=1 ZSTD_NODICT=1 ZSTD_NOCOMPRESS=1 BACKTRACE=0 PREFIX=$(ZSTD) CPPFLAGS="-O3 -flto" LDFLAGS="-O3 -flto" emmake make -j$(COMPILE_JOBS) install && \
rm -rf /tmp/zstd
libarchive: zstd
libarchive: | zstd
rm -rf /tmp/libarchive && \
git clone -b v3.7.2 --depth=1 https://github.com/libarchive/libarchive /tmp/libarchive && \
cd /tmp/libarchive && \
@@ -82,14 +82,14 @@ openfst: prepare
echo "PACKAGE_VERSION = 1.8.0" >> $(OPENFST)/Makefile && \
rm -rf /tmp/openfst
kaldi: openfst clapack-wasm
kaldi: | openfst clapack-wasm
git clone -b vosk --depth=1 https://github.com/alphacep/kaldi $(KALDI) && \
cd $(KALDI)/src && \
git apply $(SRC)/kaldi.patch && \
CXXFLAGS="-O3 -msimd128 -UHAVE_EXECINFO_H -pthread -flto" LDFLAGS="-O3 -sERROR_ON_UNDEFINED_SYMBOLS=0 -lembind -pthread -flto" emconfigure ./configure --use-cuda=no --with-cudadecoder=no --static --static-math=yes --static-fst=yes --debug-level=0 --double-precision=yes --fst-root=$(OPENFST) --clapack-root=$(CLAPACK_WASM) --host=WASM && \
emmake make -j$(COMPILE_JOBS) online2 lm rnnlm
vosk: kaldi
vosk: | kaldi
git clone -b go/v0.3.46 --depth=1 https://github.com/alphacep/vosk-api $(VOSK) && \
cd $(VOSK)/src && \
git apply $(SRC)/vosk.patch && \

View File

@@ -59,3 +59,19 @@ index 6221401..d93b87f 100644
mv $*.xxx $*.o
## Under Solaris (and other systems that do not understand ld -x),
## omit -x in the ld line above.
diff --git a/libf2c/main.c b/libf2c/main.c
index d95fdc9..ac82f68 100644
--- a/libf2c/main.c
+++ b/libf2c/main.c
@@ -105,9 +105,9 @@ char **xargv;
int
#ifdef KR_headers
-main(argc, argv) int argc; char **argv;
+m(argc, argv) int argc; char **argv;
#else
-main(int argc, char **argv)
+m(int argc, char **argv)
#endif
{
xargc = argc;

View File

@@ -3,16 +3,16 @@
genericModel::genericModel(const std::string& storepath, const std::string &id, int index) : storepath(storepath), id(id), index(index) {
fs::current_path("/opfs", tank);
if(tank.value() != 0) {
throwJS("Unable to change to OPFS directory");
throwJS("Unable to cd OPFS root");
return;
}
if(!fs::create_directories(storepath, tank)) {
throwJS("Unable to make model directory");
return;
};
fs::create_directories(storepath, tank);
if(tank.value() != 0) {
throwJS("Unable to create storepath");
}
fs::current_path(storepath, tank);
if(tank.value() != 0) {
throwJS("Unable to change to model directory");
throwJS("Unable to cd storepath");
}
}
bool genericModel::checkModel() {
@@ -29,7 +29,7 @@ bool genericModel::checkModel() {
void genericModel::afterFetch() {
thrd.setTask1([this](){
if(!extractModel()) {
fireEv("_continue", "Unable to extract model", index);
fireEv("_continue", "Unable to extract model and delete .tzst file", index);
return;
}
std::ofstream idFile("id");
@@ -41,7 +41,18 @@ void genericModel::afterFetch() {
}
idFile << id;
idFile.close();
load(false);
char a[3] {};
int fd = open("./conf/model.conf", O_RDONLY);
read(fd, a, 3);
emscripten_console_log(a);
close(fd);
std::ifstream is("./conf/model.conf");
emscripten_console_logf("%d", is.good());
emscripten_console_logf("%d", is.bad());
emscripten_console_logf("%d", is.eof());
emscripten_console_logf("%d", is.fail());
//load(false);
});
}
bool genericModel::extractModel() {
@@ -50,7 +61,7 @@ bool genericModel::extractModel() {
archive_entry* entry {};
archive_read_support_filter_all(src);
archive_read_support_format_all(src);
archive_read_open_filename(src, "m0dEl.tzst", 10240);
archive_read_open_filename(src, "/opfs/m0dEl.tzst", 10240);
if(archive_errno(src) != 0) return false;
while (archive_read_next_header(src, &entry) == ARCHIVE_OK) {
path = archive_entry_pathname(entry);
@@ -61,5 +72,5 @@ bool genericModel::extractModel() {
archive_read_extract(src, entry, ARCHIVE_EXTRACT_UNLINK);
}
archive_read_free(src);
return true;
return fs::remove("/opfs/m0dEl.tzst",tank);
}

View File

@@ -75,12 +75,7 @@ class SpkModel extends EventTarget {
}
}
Module.makeModel = async (url, storepath, id) => {
try {
let mdl = new Model(storepath, id)
}
catch(e) {
return Promise.reject(e)
}
let mdl = new Model(storepath, id)
return new Promise((resolve, reject) => {
mdl.addEventListener("_continue", (ev) => {
if(ev.detail === ".") {
@@ -98,18 +93,15 @@ Module.makeModel = async (url, storepath, id) => {
if(!res.ok) {
return reject("Unable to download model")
}
await (await (await root.getFileHandle("m0dEl.tzst", {create : true})).createWritable()).write(await res.arrayBuffer())
let wStream = await (await (await navigator.storage.getDirectory()).getFileHandle("m0dEl.tzst", {create : true})).createWritable()
await wStream.write(await res.arrayBuffer())
await wStream.close()
mdl.obj.afterFetch()
})()
})
}
Module.makeSpkModel = async (url, storepath, id) => {
try {
let mdl = new SpkModel(storepath, id)
}
catch(e) {
return Promise.reject(e)
}
let mdl = new SpkModel(storepath, id)
return new Promise((resolve, reject) => {
mdl.addEventListener("_continue", (ev) => {
if(ev.detail === ".") {
@@ -149,7 +141,7 @@ Module.makeRecognizer = (model, sampleRate) => {
rec._init(model.obj, sampleRate)
return retval
}
let processorUrl = URL.createObjectURL(new Blob([
let processorUrl = URL.createObjectURL(new Blob(['(',
(() => {
registerProcessor("BRProcessor", class extends AudioWorkletProcessor {
constructor(options) {
@@ -169,9 +161,9 @@ let processorUrl = URL.createObjectURL(new Blob([
}
})
}).toString()
], {type : "text/javascript"}))
, ')()'], {type : "text/javascript"}))
// Taken from the worker.js file
let pthreadUrl = URL.createObjectURL(new Blob([
let pthreadUrl = URL.createObjectURL(new Blob(['(',
(() => {
/**
* @license
@@ -342,4 +334,4 @@ let pthreadUrl = URL.createObjectURL(new Blob([
self.onmessage = handleMessage;
}).toString()
], {type : "text/javascript"}))
, ')()'], {type : "text/javascript"}))

View File

@@ -1,5 +1,5 @@
diff --git a/src/model.cc b/src/model.cc
index 035ffee..9cb7e30 100644
index 035ffee..7f5148a 100644
--- a/src/model.cc
+++ b/src/model.cc
@@ -74,38 +74,48 @@ static void KaldiLogHandler(const LogMessageEnvelope &env, const char *message)
@@ -78,6 +78,18 @@ index 035ffee..9cb7e30 100644
}
#endif
diff --git a/src/model.h b/src/model.h
index 7fc09df..608d28a 100644
--- a/src/model.h
+++ b/src/model.h
@@ -32,6 +32,7 @@
#include "rnnlm/rnnlm-utils.h"
#include "rnnlm/rnnlm-lattice-rescoring.h"
#include <atomic>
+#include <emscripten/console.h>
using namespace kaldi;
using namespace std;
diff --git a/src/recognizer.cc b/src/recognizer.cc
index 1da6e6b..5e0652a 100644
--- a/src/recognizer.cc

File diff suppressed because one or more lines are too long

View File

@@ -1,8 +1,8 @@
cd .. &&
SRC=$(realpath src) &&
KALDI=$(realpath kaldi) &&
VOSK=$(realpath vosk-api) &&
OPENFST=$KALDI/tools/openfst &&
VOSK=$(realpath vosk) &&
OPENFST=$(realpath openfst) &&
LIBARCHIVE=$(realpath libarchive) &&
ZSTD=$(realpath zstd) &&
CLAPACK_WASM=$(realpath clapack-wasm) &&
@@ -15,7 +15,7 @@ if [ ! -d $EMSDK_PATH ]; then
echo "Invalid EMSDK path"
exit 1
fi
if [ $MAX_THREAD -lt 2 ]; then
if [ $MAX_THREADS -lt 2 ]; then
echo "MAX_THREAD must be greater or equal to 2" &&
exit 1
fi
@@ -23,7 +23,7 @@ if ! [[ $MAX_MEMORY =~ ^[0-9]+([kmgt]b)?$ ]]; then
echo "MAX_MEMORY valid suffixes are kb, mb, gb, tb, none (bytes)" &&
exit 1
fi
if [ $(realpath $EMSDK) == $(realpath emsdk) ]; then
if [ $EMSDK = ../emsdk ]; then
echo "EMSDK is current directory, installing emsdk and Emscripten..." &&
git clone --depth=1 https://github.com/emscripten-core/emsdk.git &&
cd emsdk &&
@@ -35,7 +35,7 @@ fi
export PATH=:$PATH:$EMSDK/upstream/bin &&
cd $SRC &&
em++ -O0 global.cc genericModel.cc model.cc spkModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sEMBIND_STD_STRING_IS_UTF8 -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sMODULARIZE -sEXPORTED_FUNCTIONS=_malloc,_free,_main -sSUPPORT_LONGJMP=0 -sTRUSTED_TYPES -sEXPORT_NAME=loadBR -sENVIRONMENT=web,worker -sINITIAL_MEMORY=$MAX_MEMORY -sPTHREAD_POOL_SIZE=$MAX_THREADS -sPTHREAD_POOL_SIZE_STRICT=0 -sEXIT_RUNTIME -sALLOW_BLOCKING_ON_MAIN_THREAD=1 -sPOLYFILL=0 -sPTHREAD_POOL_DELAY_LOAD --pre-js pre.js -I. -I$LIBARCHIVE/include -I$VOSK/src -L$LIBARCHIVE/lib -larchive -L$ZSTD/lib -lzstd -L$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$VOSK/src -l:vosk.a -lopfs.js -lembind -pthread -flto -o ../test/BrowserRecognizer.js &&
em++ -O0 global.cc genericModel.cc model.cc spkModel.cc recognizer.cc bindings.cc -sWASMFS -sWASM_BIGINT -sSINGLE_FILE -sMODULARIZE -sTRUSTED_TYPES -sEMBIND_STD_STRING_IS_UTF8 -sPTHREAD_POOL_DELAY_LOAD -sTEXTDECODER=2 -sPTHREAD_POOL_SIZE_STRICT=2 -sINITIAL_MEMORY=$MAX_MEMORY -sPTHREAD_POOL_SIZE=$MAX_THREADS -sPOLYFILL=0 -sSUPPORT_LONGJMP=0 -sEXPORTED_FUNCTIONS=_malloc,_main -sEXPORT_NAME=loadBR -sEXPORTED_RUNTIME_METHODS=UTF8ToString,stringToUTF8OnStack -sMALLOC=emmalloc -sENVIRONMENT=web,worker -I. -I$LIBARCHIVE/include -I$VOSK/src -L$LIBARCHIVE/lib -larchive -L$ZSTD/lib -lzstd -L$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 -lopfs.js -lembind -pthread -flto --pre-js pre.js -o ../test/BrowserRecognizer.js &&
cd ../test &&
rm -f BrowserRecognizer.worker.js &&
sed -i "s/locateFile('BrowserRecognizer.worker.js')/pthreadUrl/g" BrowserRecognizer.js &&