whisper: expose detected lang_id and lang_prob from internal state

This commit is contained in:
Seven Du 2026-01-04 22:58:01 +08:00
parent 7aa8818647
commit 6d507e377e
No known key found for this signature in database
2 changed files with 14 additions and 1 deletions

View File

@ -415,6 +415,9 @@ extern "C" {
WHISPER_API float * whisper_get_logits (struct whisper_context * ctx);
WHISPER_API float * whisper_get_logits_from_state(struct whisper_state * state);
WHISPER_API int whisper_get_lang_id_from_state(struct whisper_state * state);
WHISPER_API float whisper_get_lang_prob_from_state(struct whisper_state * state);
// Token Id -> String. Uses the vocabulary in the provided context
WHISPER_API const char * whisper_token_to_str(struct whisper_context * ctx, whisper_token token);
WHISPER_API const char * whisper_model_type_readable(struct whisper_context * ctx);

View File

@ -892,6 +892,7 @@ struct whisper_state {
std::vector<whisper_token> prompt_past1; // dynamic context from decoded output
int lang_id = 0; // english by default
float lang_prob = 0.0f; // probability of the detected language
std::string path_model; // populated by whisper_init_from_file_with_params()
@ -4198,6 +4199,14 @@ float * whisper_get_logits_from_state(struct whisper_state * state) {
return state->logits.data();
}
int whisper_get_lang_id_from_state(struct whisper_state * state) {
return state->lang_id;
}
float whisper_get_lang_prob_from_state(struct whisper_state * state) {
return state->lang_prob;
}
const char * whisper_token_to_str(struct whisper_context * ctx, whisper_token token) {
return ctx->vocab.id_to_token.at(token).c_str();
}
@ -6831,9 +6840,10 @@ int whisper_full_with_state(
return -3;
}
state->lang_id = lang_id;
state->lang_prob = probs[lang_id];
params.language = whisper_lang_str(lang_id);
WHISPER_LOG_INFO("%s: auto-detected language: %s (p = %f)\n", __func__, params.language, probs[whisper_lang_id(params.language)]);
WHISPER_LOG_INFO("%s: auto-detected language: %s (p = %f)\n", __func__, params.language, probs[lang_id]);
if (params.detect_language) {
return 0;
}