Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.
Python asraudiochineseemotion-recognitionfunasrmcp-servermultilingual-asropenai-compatible-apiparaformerpunctuationpytorchreal-time-asrspeaker-diarizationspeech-recognitionspeech-to-textstreaming-asrtranscriptionvllmvoice-activity-detectionwhisper-alternative
update funasrruntime
雾聪687277ae9d86
diff --git a/runtime/onnxruntime/src/funasrruntime.cpp b/runtime/onnxruntime/src/funasrruntime.cppindex e9a3f59f1..e1b147764 100644--- a/runtime/onnxruntime/src/funasrruntime.cpp+++ b/runtime/onnxruntime/src/funasrruntime.cpp@@ -252,7 +252,7 @@if (wfst_decoder){wfst_decoder->StartUtterance();}- vector<string> msgs = (offline_stream->asr_handle)->Forward(buff, len, true, hw_emb, dec_handle);+ vector<string> msgs = (offline_stream->asr_handle)->Forward(buff, len, true, hw_emb, dec_handle, batch_in);for(int idx=0; idx<batch_in; idx++){string msg = msgs[idx];std::vector<std::string> msg_vec = funasr::split(msg, '|');@@ -359,7 +359,7 @@if (wfst_decoder){wfst_decoder->StartUtterance();}- vector<string> msgs = (offline_stream->asr_handle)->Forward(buff, len, true, hw_emb, dec_handle);+ vector<string> msgs = (offline_stream->asr_handle)->Forward(buff, len, true, hw_emb, dec_handle, batch_in);for(int idx=0; idx<batch_in; idx++){string msg = msgs[idx];std::vector<std::string> msg_vec = funasr::split(msg, '|');