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
docs(punctuation): clarify SenseVoice postprocessing
zhifu gao176e35ec7c9d
diff --git a/funasr/models/ct_transformer/model.py b/funasr/models/ct_transformer/model.pyindex c7beb1ba0..b82cafdae 100644--- a/funasr/models/ct_transformer/model.py+++ b/funasr/models/ct_transformer/model.py@@ -46,8 +46,9 @@ class CTTransformer(torch.nn.Module):Output: {"key": "...", "text": "punctuated text", "punc_array": Tensor}punc_array encoding: 1=none, 2=comma(,), 3=period(。), 4=question(?)- Note: Not needed for Fun-ASR-Nano/SenseVoice/Qwen3-ASR (they output punctuation natively).- Only required for Paraformer models.+ For unpunctuated SenseVoiceSmall or Paraformer output, configure punc_model+ when punctuation is needed. Models that already emit punctuation may not+ need this postprocessor.Author: Speech Lab of DAMO Academy, Alibaba GroupCT-Transformer: Controllable time-delay transformer for real-time punctuation prediction and disfluency detection