extern/whisper.cpp
1
0
Fork 1
mirror of https://github.com/ggerganov/whisper.cpp.git synced 2024-11-07 08:34:37 +01:00
Code Issues Packages Projects Releases Wiki Activity

examples : add tinydiarization support for streaming (#1137)

Browse Source
This commit is contained in:
Duncan McConnell 2023-08-03 03:24:07 -05:00 committed by GitHub
parent a792c4079c
commit b948361956
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 17 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
examples/stream/stream.cpp
View File

@ -47,6 +47,7 @@ struct whisper_params {
bool print_special = false; bool print_special = false;
bool no_context = true; bool no_context = true;
bool no_timestamps = false; bool no_timestamps = false;
bool tinydiarize = false;
std::string language = "en"; std::string language = "en";
std::string model = "models/ggml-base.en.bin"; std::string model = "models/ggml-base.en.bin";
@ -80,6 +81,8 @@ bool whisper_params_parse(int argc, char ** argv, whisper_params & params) {
else if (arg == "-l" || arg == "--language") { params.language = argv[++i]; } else if (arg == "-l" || arg == "--language") { params.language = argv[++i]; }
else if (arg == "-m" || arg == "--model") { params.model = argv[++i]; } else if (arg == "-m" || arg == "--model") { params.model = argv[++i]; }
else if (arg == "-f" || arg == "--file") { params.fname_out = argv[++i]; } else if (arg == "-f" || arg == "--file") { params.fname_out = argv[++i]; }
else if (arg == "-tdrz" || arg == "--tinydiarize") { params.tinydiarize = true; }
else { else {
fprintf(stderr, "error: unknown argument: %s\n", arg.c_str()); fprintf(stderr, "error: unknown argument: %s\n", arg.c_str());
whisper_print_usage(argc, argv, params); whisper_print_usage(argc, argv, params);
@ -113,6 +116,7 @@ void whisper_print_usage(int /*argc*/, char ** argv, const whisper_params & para
fprintf(stderr, " -l LANG, --language LANG [%-7s] spoken language\n", params.language.c_str()); fprintf(stderr, " -l LANG, --language LANG [%-7s] spoken language\n", params.language.c_str());
fprintf(stderr, " -m FNAME, --model FNAME [%-7s] model path\n", params.model.c_str()); fprintf(stderr, " -m FNAME, --model FNAME [%-7s] model path\n", params.model.c_str());
fprintf(stderr, " -f FNAME, --file FNAME [%-7s] text output file name\n", params.fname_out.c_str()); fprintf(stderr, " -f FNAME, --file FNAME [%-7s] text output file name\n", params.fname_out.c_str());
fprintf(stderr, " -tdrz, --tinydiarize [%-7s] enable tinydiarize (requires a tdrz model)\n", params.tinydiarize ? "true" : "false");
fprintf(stderr, "\n"); fprintf(stderr, "\n");
} }
@ -299,6 +303,8 @@ int main(int argc, char ** argv) {
wparams.audio_ctx = params.audio_ctx; wparams.audio_ctx = params.audio_ctx;
wparams.speed_up = params.speed_up; wparams.speed_up = params.speed_up;
wparams.tdrz_enable = params.tinydiarize; // [TDRZ]
// disable temperature fallback // disable temperature fallback
//wparams.temperature_inc = -1.0f; //wparams.temperature_inc = -1.0f;
wparams.temperature_inc = params.no_fallback ? 0.0f : wparams.temperature_inc; wparams.temperature_inc = params.no_fallback ? 0.0f : wparams.temperature_inc;
@ -344,10 +350,19 @@ int main(int argc, char ** argv) {
const int64_t t0 = whisper_full_get_segment_t0(ctx, i); const int64_t t0 = whisper_full_get_segment_t0(ctx, i);
const int64_t t1 = whisper_full_get_segment_t1(ctx, i); const int64_t t1 = whisper_full_get_segment_t1(ctx, i);
printf ("[%s --> %s] %s\n", to_timestamp(t0).c_str(), to_timestamp(t1).c_str(), text); std::string output = "[" + to_timestamp(t0) + " --> " + to_timestamp(t1) + "] " + text;
if (whisper_full_get_segment_speaker_turn_next(ctx, i)) {
output += " [SPEAKER_TURN]";
}
output += "\n";
printf("%s", output.c_str());
fflush(stdout);
if (params.fname_out.length() > 0) { if (params.fname_out.length() > 0) {
fout << "[" << to_timestamp(t0) << " --> " << to_timestamp(t1) << "] " << text << std::endl; fout << output;
} }
} }
} }