extern/whisper.cpp
1
0
Fork 1
mirror of https://github.com/ggerganov/whisper.cpp.git synced 2025-05-30 22:47:22 +02:00
Code Issues Packages Projects Releases Wiki Activity

vad : store VAD context in whisper_state (#3156)

Browse Source 
* vad : store VAD context in whisper_state

This commit stores the VAD context in the whisper_state structure,
allowing for better management and reuse of the VAD context across
multiple calls to the whisper_vad function.

The motivation for this change is that when updating the stream example
I noticed that the VAD context was being re-initialized every time the
whisper_vad function was called. This involved loading the VAD model
which is expensive and unnecessary if the context can be reused.

Storing this in the whisper_state seems follow the pattern simliar to
how whisper_coreml_context and whisper_openvion_context are stored.

* vad : free vad_context in whisper_free_state
This commit is contained in:
Daniel Bevenius 2025-05-16 07:53:26 +02:00 committed by GitHub
parent 20a20decd9
commit bae5d074c7
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 16 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
src/whisper.cpp
View File

@ -954,6 +954,8 @@ struct whisper_state {
// [EXPERIMENTAL] speed-up techniques
int32_t exp_n_audio_ctx = 0; // 0 - use default
whisper_vad_context * vad_context = nullptr;
struct vad_segment_info {
float orig_start;
float orig_end;
@ -3853,6 +3855,11 @@ void whisper_free_state(struct whisper_state * state) {
// [EXPERIMENTAL] Token-level timestamps with DTW
aheads_masks_free(state->aheads_masks);
if (state->vad_context != nullptr) {
whisper_vad_free(state->vad_context);
state->vad_context = nullptr;
}
delete state;
}
}
@ -6613,12 +6620,16 @@ static bool whisper_vad(
WHISPER_LOG_INFO("%s: VAD is enabled, processing speach segments only\n", __func__);
filtered_n_samples = 0;
struct whisper_vad_context_params vad_ctx_params = whisper_vad_default_context_params();
struct whisper_vad_context * vctx = whisper_vad_init_from_file_with_params(params.vad_model_path, vad_ctx_params);
if (vctx == nullptr) {
WHISPER_LOG_ERROR("%s: failed to initialize VAD context\n", __func__);
return false;
if (state->vad_context == nullptr) {
struct whisper_vad_context_params vad_ctx_params = whisper_vad_default_context_params();
struct whisper_vad_context * vctx = whisper_vad_init_from_file_with_params(params.vad_model_path, vad_ctx_params);
if (vctx == nullptr) {
WHISPER_LOG_ERROR("%s: failed to initialize VAD context\n", __func__);
return false;
}
state->vad_context = vctx;
}
auto vctx = state->vad_context;
const whisper_vad_params & vad_params = params.vad_params;