ExternalVendorCode/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/whisper.cpp.git synced 2025-06-13 04:28:07 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

examples : refactor in order to reuse code and reduce duplication (#482)

Browse Source 
* examples : refactor common code into a library

* examples : refactor common SDL code into a library

* make : update Makefile to use common libs

* common : fix MSVC M_PI ..

* addon.node : link common lib
This commit is contained in:
Georgi Gerganov
2023-02-15 19:28:10 +02:00
committed by GitHub
parent 0336161b7d
commit 09d7d2b68e
19 changed files with 580 additions and 1254 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
examples/command.wasm/CMakeLists.txt
View File

@ -11,6 +11,7 @@ add_executable(${TARGET}
include(DefaultTargetOptions)
target_link_libraries(${TARGET} PRIVATE
common
whisper
)

62
examples/command.wasm/emscripten.cpp
View File

@ -1,4 +1,5 @@
#include "ggml.h"
#include "common.h"
#include "whisper.h"
#include <emscripten.h>
@ -27,24 +28,6 @@ std::string g_transcribed = "";
std::vector<float> g_pcmf32;
static std::string trim(const std::string & s) {
std::regex e("^\\s+|\\s+$");
return std::regex_replace(s, e, "");
}
static void high_pass_filter(std::vector<float> & data, float cutoff, float sample_rate) {
const float rc = 1.0f / (2.0f * M_PI * cutoff);
const float dt = 1.0f / sample_rate;
const float alpha = dt / (rc + dt);
float y = data[0];
for (size_t i = 1; i < data.size(); i++) {
y = alpha * (y + data[i] - data[i - 1]);
data[i] = y;
}
}
// compute similarity between two strings using Levenshtein distance
static float similarity(const std::string & s0, const std::string & s1) {
const size_t len0 = s0.size() + 1;
@ -75,44 +58,6 @@ void command_set_status(const std::string & status) {
g_status = status;
}
bool command_vad_simple(std::vector<float> & pcmf32, int sample_rate, int last_ms, float vad_thold, float freq_thold, bool verbose) {
const int n_samples = pcmf32.size();
const int n_samples_last = (sample_rate * last_ms) / 1000;
if (n_samples_last >= n_samples) {
// not enough samples - assume no speech
return false;
}
if (freq_thold > 0.0f) {
high_pass_filter(pcmf32, freq_thold, sample_rate);
}
float energy_all = 0.0f;
float energy_last = 0.0f;
for (size_t i = 0; i < n_samples; i++) {
energy_all += fabsf(pcmf32[i]);
if (i >= n_samples - n_samples_last) {
energy_last += fabsf(pcmf32[i]);
}
}
energy_all /= n_samples;
energy_last /= n_samples_last;
if (verbose) {
fprintf(stderr, "%s: energy_all: %f, energy_last: %f, vad_thold: %f, freq_thold: %f\n", __func__, energy_all, energy_last, vad_thold, freq_thold);
}
if (energy_last > vad_thold*energy_all) {
return false;
}
return true;
}
std::string command_transcribe(whisper_context * ctx, const whisper_full_params & wparams, const std::vector<float> & pcmf32, float & prob, int64_t & t_ms) {
const auto t_start = std::chrono::high_resolution_clock::now();
@ -155,7 +100,7 @@ void command_get_audio(int ms, int sample_rate, std::vector<float> & audio) {
const int64_t n_samples = (ms * sample_rate) / 1000;
int64_t n_take = 0;
if (g_pcmf32.size() < n_samples) {
if (n_samples > (int) g_pcmf32.size()) {
n_take = g_pcmf32.size();
} else {
n_take = n_samples;
@ -187,7 +132,6 @@ void command_main(size_t index) {
printf("command: using %d threads\n", wparams.n_threads);
bool is_running = true;
bool have_prompt = false;
bool ask_prompt = true;
bool print_energy = false;
@ -233,7 +177,7 @@ void command_main(size_t index) {
{
command_get_audio(vad_ms, WHISPER_SAMPLE_RATE, pcmf32_cur);
if (command_vad_simple(pcmf32_cur, WHISPER_SAMPLE_RATE, 1000, vad_thold, freq_thold, print_energy)) {
if (::vad_simple(pcmf32_cur, WHISPER_SAMPLE_RATE, 1000, vad_thold, freq_thold, print_energy)) {
fprintf(stdout, "%s: Speech detected! Processing ...\n", __func__);
command_set_status("Speech detected! Processing ...");