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
test
语帆5ecd13bd0460
diff --git a/funasr/utils/load_utils.py b/funasr/utils/load_utils.pyindex b7d0200cc..20fa0fd2e 100644--- a/funasr/utils/load_utils.py+++ b/funasr/utils/load_utils.py@@ -56,8 +56,7 @@ def load_audio_text_image_video(data_or_path_or_list, fs: int = 16000, audio_fs:elif isinstance(data_or_path_or_list, str) and data_type == "kaldi_ark":data_mat = kaldiio.load_mat(data_or_path_or_list)if isinstance(data_mat, tuple):- sampling_rate, mat = data_mat- assert sampling_rate == audio_fs+ audio_fs, mat = data_matelse:mat = data_matif mat.dtype == 'int16' or mat.dtype == 'int32':