haraldwolff
/
whisper.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

examples : fix + refactor Levenshtein distance

pull/842/head
Georgi Gerganov 2023-04-30 19:12:49 +03:00
parent 794b162a46
commit 4a7d49af95
No known key found for this signature in database
GPG Key ID: 449E073F9DC10735
4 changed files with 26 additions and 50 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

25
examples/command.wasm/emscripten.cpp
View File

@ -28,31 +28,6 @@ std::string g_transcribed = "";
std::vector<float> g_pcmf32;
// 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;
const size_t len1 = s1.size() + 1;
std::vector<int> col(len1, 0);
std::vector<int> prevCol(len1, 0);
for (size_t i = 0; i < len1; i++) {
prevCol[i] = i;
}
for (size_t i = 0; i < len0; i++) {
col[0] = i;
for (size_t j = 1; j < len1; j++) {
col[j] = std::min(std::min(1 + col[j - 1], 1 + prevCol[j]), prevCol[j - 1] + (s0[i - 1] == s1[j - 1] ? 0 : 1));
}
col.swap(prevCol);
}
const float dist = prevCol[len1 - 1];
return 1.0f - (dist / std::max(s0.size(), s1.size()));
}
void command_set_status(const std::string & status) {
std::lock_guard<std::mutex> lock(g_mutex);
g_status = status;

25
examples/command/command.cpp
View File

@ -163,31 +163,6 @@ std::string transcribe(whisper_context * ctx, const whisper_params & params, con
return result;
}
// compute similarity between two strings using Levenshtein distance
float similarity(const std::string & s0, const std::string & s1) {
const size_t len0 = s0.size() + 1;
const size_t len1 = s1.size() + 1;
std::vector<int> col(len1, 0);
std::vector<int> prevCol(len1, 0);
for (size_t i = 0; i < len1; i++) {
prevCol[i] = i;
}
for (size_t i = 0; i < len0; i++) {
col[0] = i;
for (size_t j = 1; j < len1; j++) {
col[j] = std::min(std::min(1 + col[j - 1], 1 + prevCol[j]), prevCol[j - 1] + (s0[i - 1] == s1[j - 1] ? 0 : 1));
}
col.swap(prevCol);
}
const float dist = prevCol[len1 - 1];
return 1.0f - (dist / std::max(s0.size(), s1.size()));
}
std::vector<std::string> read_allowed_commands(const std::string & fname) {
std::vector<std::string> allowed_commands;

24
examples/common.cpp
View File

@ -479,3 +479,27 @@ bool vad_simple(std::vector<float> & pcmf32, int sample_rate, int last_ms, float
return true;
}
float similarity(const std::string & s0, const std::string & s1) {
const size_t len0 = s0.size() + 1;
const size_t len1 = s1.size() + 1;
std::vector<int> col(len1, 0);
std::vector<int> prevCol(len1, 0);
for (size_t i = 0; i < len1; i++) {
prevCol[i] = i;
}
for (size_t i = 0; i < len0; i++) {
col[0] = i;
for (size_t j = 1; j < len1; j++) {
col[j] = std::min(std::min(1 + col[j - 1], 1 + prevCol[j]), prevCol[j - 1] + (i > 0 && s0[i - 1] == s1[j - 1] ? 0 : 1));
}
col.swap(prevCol);
}
const float dist = prevCol[len1 - 1];
return 1.0f - (dist / std::max(s0.size(), s1.size()));
}

2
examples/common.h
View File

@ -118,3 +118,5 @@ bool vad_simple(
float freq_thold,
bool verbose);
// compute similarity between two strings using Levenshtein distance
float similarity(const std::string & s0, const std::string & s1);