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
bugfix seg_dict_file
游雁e02f361de651
diff --git a/funasr/tokenizer/char_tokenizer.py b/funasr/tokenizer/char_tokenizer.pyindex 28b4f407c..2efc0b056 100644--- a/funasr/tokenizer/char_tokenizer.py+++ b/funasr/tokenizer/char_tokenizer.py@@ -36,6 +36,7 @@ class CharTokenizer(BaseTokenizer):self.remove_non_linguistic_symbols = remove_non_linguistic_symbolsself.split_with_space = split_with_spaceself.seg_dict = None+ seg_dict = seg_dict if seg_dict is not None else kwargs.get("seg_dict_file", None)if seg_dict is not None:self.seg_dict = load_seg_dict(seg_dict)