vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-14 02:49:41 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

phonemizer fixes, now kokoro works very well

Browse source
This commit is contained in:
Concedo 2025-08-18 16:13:16 +08:00
parent 3f621be7dd
commit 67ef5e6c02
13 changed files with 58645 additions and 58612 deletions
Download patch file Download diff file Expand all files Collapse all files

15
otherarch/tts_adapter.cpp
View file

@ -504,10 +504,12 @@ static generation_configuration * ttscpp_config = nullptr;
static struct tts_runner * ttscpp_runner = nullptr;
int total_tts_gens = 0;
static std::string tts_executable_path = "";
bool ttstype_load_model(const tts_load_model_inputs inputs)
{
tts_is_quiet = inputs.quiet;
tts_executable_path = inputs.executable_path;
//duplicated from expose.cpp
int cl_parseinfo = inputs.clblast_info; //first digit is whether configured, second is platform, third is devices
@ -544,6 +546,11 @@ bool ttstype_load_model(const tts_load_model_inputs inputs)
if (detectedarch!="" && SUPPORTED_ARCHITECTURES.find(detectedarch) != SUPPORTED_ARCHITECTURES.end()) {
is_ttscpp_file = true;
printf("\nLoading TTS.CPP Model Arch: %s \n", detectedarch.c_str());
if(detectedarch=="kokoro")
{
//setup kokoro IPA
populate_kokoro_ipa_map(tts_executable_path);
}
}else{
printf("\nLoading OuteTTS Model, OuteTTS: %s \nWavTokenizer: %s \n",modelfile_ttc.c_str(),modelfile_cts.c_str());
if(modelfile_ttc=="" || modelfile_cts=="")
@ -557,7 +564,7 @@ bool ttstype_load_model(const tts_load_model_inputs inputs)
// tts init
if (is_ttscpp_file) {
ttscpp_config = new generation_configuration("am_adam", 25, 1.0, 1.0, true, "", 2048, 1.0);
ttscpp_config = new generation_configuration("am_echo", 25, 1.0, 1.0, true, "", 2048, 1.0);
ttscpp_runner = runner_from_file(modelfile_ttc, inputs.threads, ttscpp_config, true);
if (ttscpp_runner == nullptr) {
printf("\nTTS Load Error: Failed to initialize TTSCPP!\n");
@ -652,14 +659,14 @@ static tts_generation_outputs ttstype_generate_ttscpp(const tts_generation_input
return output;
}
int speaker_seed = inputs.speaker_seed;
std::string voiceused = "am_adam";
std::string voiceused = "am_echo";
std::string prompt = inputs.prompt;
double ttstime = 0;
timer_start();
switch(speaker_seed)
{
case 1:
voiceused = "am_adam";
voiceused = "am_echo";
break;
case 2:
voiceused = "af_alloy";
@ -668,7 +675,7 @@ static tts_generation_outputs ttstype_generate_ttscpp(const tts_generation_input
voiceused = "af_jessica";
break;
case 4:
voiceused = "bm_george";
voiceused = "bm_daniel";
break;
case 5:
voiceused = "bf_isabella";