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
whisper bugfix
游雁7f6099140a78
diff --git a/funasr/models/whisper/model.py b/funasr/models/whisper/model.pyindex a3321000e..791fddd47 100644--- a/funasr/models/whisper/model.py+++ b/funasr/models/whisper/model.py@@ -9,6 +9,7 @@ from torch import Tensorfrom torch import nnimport whisper+# import whisper_timestamped as whisperfrom funasr.utils.load_utils import load_audio_text_image_video, extract_fbank@@ -111,10 +112,10 @@ class WhisperWarp(nn.Module):# decode the audiooptions = whisper.DecodingOptions(**kwargs.get("DecodingOptions", {}))-- result = whisper.decode(self.model, speech, language='english')++ result = whisper.decode(self.model, speech, options=options)# result = whisper.transcribe(self.model, speech)-+results = []result_i = {"key": key[0], "text": result.text}