mozilla
/
DeepSpeech
зеркало из https://github.com/mozilla/DeepSpeech.git
1
0
Форкнуть 0
Код Задачи Пакеты Проекты Релизы Вики Активность

Address review comments

This commit is contained in:
Reuben Morais 2018-11-09 08:49:08 -02:00
Родитель 56dc024d29
Коммит 4d05c9af46
3 изменённых файлов: 22 добавлений и 13 удалений
Показать статистику Скачать Patch файл Скачать Diff файл Показать все файлы Свернуть все файлы

2
native_client/ctcdecode/Makefile
Просмотреть файл

@ -6,7 +6,7 @@ NUM_PROCESSES ?= 1
# ARM64 can't find the proper libm.so without this # ARM64 can't find the proper libm.so without this
ifeq ($(TARGET),rpi3-armv8) ifeq ($(TARGET),rpi3-armv8)
LDFLAGS_NEEDED += $(RASPBIAN)/lib/aarch64-linux-gnu/libm-2.24.so LDFLAGS_NEEDED += $(RASPBIAN)/lib/aarch64-linux-gnu/libm.so.6
endif endif
all: bindings all: bindings

31
native_client/ctcdecode/scorer.cpp
Просмотреть файл

@ -50,13 +50,14 @@ Scorer::Scorer(double alpha,
const char* alphabet_config_path) const char* alphabet_config_path)
: Scorer(alpha, beta, lm_path, trie_path, Alphabet(alphabet_config_path)) : Scorer(alpha, beta, lm_path, trie_path, Alphabet(alphabet_config_path))
{ {
} }
Scorer::~Scorer() { Scorer::~Scorer()
{
} }
void Scorer::setup(const std::string& lm_path, const std::string& trie_path) { void Scorer::setup(const std::string& lm_path, const std::string& trie_path)
{
// load language model // load language model
const char* filename = lm_path.c_str(); const char* filename = lm_path.c_str();
VALID_CHECK_EQ(access(filename, R_OK), 0, "Invalid language model path"); VALID_CHECK_EQ(access(filename, R_OK), 0, "Invalid language model path");
@ -114,7 +115,8 @@ void Scorer::setup(const std::string& lm_path, const std::string& trie_path) {
max_order_ = language_model_->Order(); max_order_ = language_model_->Order();
} }
void Scorer::save_dictionary(const std::string& path) { void Scorer::save_dictionary(const std::string& path)
{
std::ofstream fout(path, std::ios::binary); std::ofstream fout(path, std::ios::binary);
fout.write(reinterpret_cast<const char*>(&MAGIC), sizeof(MAGIC)); fout.write(reinterpret_cast<const char*>(&MAGIC), sizeof(MAGIC));
fout.write(reinterpret_cast<const char*>(&FILE_VERSION), sizeof(FILE_VERSION)); fout.write(reinterpret_cast<const char*>(&FILE_VERSION), sizeof(FILE_VERSION));
@ -123,7 +125,8 @@ void Scorer::save_dictionary(const std::string& path) {
dictionary->Write(fout, opt); dictionary->Write(fout, opt);
} }
double Scorer::get_log_cond_prob(const std::vector<std::string>& words) { double Scorer::get_log_cond_prob(const std::vector<std::string>& words)
{
double cond_prob = OOV_SCORE; double cond_prob = OOV_SCORE;
lm::ngram::State state, tmp_state, out_state; lm::ngram::State state, tmp_state, out_state;
// avoid to inserting <s> in begin // avoid to inserting <s> in begin
@ -143,7 +146,8 @@ double Scorer::get_log_cond_prob(const std::vector<std::string>& words) {
return cond_prob/NUM_FLT_LOGE; return cond_prob/NUM_FLT_LOGE;
} }
double Scorer::get_sent_log_prob(const std::vector<std::string>& words) { double Scorer::get_sent_log_prob(const std::vector<std::string>& words)
{
std::vector<std::string> sentence; std::vector<std::string> sentence;
if (words.size() == 0) { if (words.size() == 0) {
for (size_t i = 0; i < max_order_; ++i) { for (size_t i = 0; i < max_order_; ++i) {
@ -159,7 +163,8 @@ double Scorer::get_sent_log_prob(const std::vector<std::string>& words) {
return get_log_prob(sentence); return get_log_prob(sentence);
} }
double Scorer::get_log_prob(const std::vector<std::string>& words) { double Scorer::get_log_prob(const std::vector<std::string>& words)
{
assert(words.size() > max_order_); assert(words.size() > max_order_);
double score = 0.0; double score = 0.0;
for (size_t i = 0; i < words.size() - max_order_ + 1; ++i) { for (size_t i = 0; i < words.size() - max_order_ + 1; ++i) {
@ -170,12 +175,14 @@ double Scorer::get_log_prob(const std::vector<std::string>& words) {
return score; return score;
} }
void Scorer::reset_params(float alpha, float beta) { void Scorer::reset_params(float alpha, float beta)
{
this->alpha = alpha; this->alpha = alpha;
this->beta = beta; this->beta = beta;
} }
std::vector<std::string> Scorer::split_labels(const std::vector<int>& labels) { std::vector<std::string> Scorer::split_labels(const std::vector<int>& labels)
{
if (labels.empty()) return {}; if (labels.empty()) return {};
std::string s = alphabet_.LabelsToString(labels); std::string s = alphabet_.LabelsToString(labels);
@ -188,7 +195,8 @@ std::vector<std::string> Scorer::split_labels(const std::vector<int>& labels) {
return words; return words;
} }
std::vector<std::string> Scorer::make_ngram(PathTrie* prefix) { std::vector<std::string> Scorer::make_ngram(PathTrie* prefix)
{
std::vector<std::string> ngram; std::vector<std::string> ngram;
PathTrie* current_node = prefix; PathTrie* current_node = prefix;
PathTrie* new_node = nullptr; PathTrie* new_node = nullptr;
@ -221,7 +229,8 @@ std::vector<std::string> Scorer::make_ngram(PathTrie* prefix) {
return ngram; return ngram;
} }
void Scorer::fill_dictionary(const std::vector<std::string>& vocabulary, bool add_space) { void Scorer::fill_dictionary(const std::vector<std::string>& vocabulary, bool add_space)
{
fst::StdVectorFst dictionary; fst::StdVectorFst dictionary;
// For each unigram convert to ints and put in trie // For each unigram convert to ints and put in trie
for (const auto& word : vocabulary) { for (const auto& word : vocabulary) {

2
tc-single-shot-inference.sh
Просмотреть файл

@ -48,7 +48,7 @@ platform=$(python -c 'import sys; import platform; plat = platform.system().lowe
whl_ds_version="$(python -c 'from pkg_resources import parse_version; print(parse_version("'${DS_VERSION}'"))')" whl_ds_version="$(python -c 'from pkg_resources import parse_version; print(parse_version("'${DS_VERSION}'"))')"
decoder_pkg="ds_ctcdecoder-${whl_ds_version}-cp${pyver_pkg}-cp${pyver_pkg}${py_unicode_type}-${platform}.whl" decoder_pkg="ds_ctcdecoder-${whl_ds_version}-cp${pyver_pkg}-cp${pyver_pkg}${py_unicode_type}-${platform}.whl"
decoder_pkg_url=${DEEPSPEECH_ARTIFACTS_ROOT}/${decoder_pkg} decoder_pkg_url=${DECODER_ARTIFACTS_ROOT}/${decoder_pkg}
LD_LIBRARY_PATH=${PY37_LDPATH}:$LD_LIBRARY_PATH pip install --verbose --only-binary :all: ${PY37_SOURCE_PACKAGE} --upgrade ${decoder_pkg_url} | cat LD_LIBRARY_PATH=${PY37_LDPATH}:$LD_LIBRARY_PATH pip install --verbose --only-binary :all: ${PY37_SOURCE_PACKAGE} --upgrade ${decoder_pkg_url} | cat