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
vad conf
游雁60a7d39f3939
diff --git a/funasr/auto/auto_model.py b/funasr/auto/auto_model.pyindex d0444557d..bba44e792 100644--- a/funasr/auto/auto_model.py+++ b/funasr/auto/auto_model.py@@ -187,7 +187,8 @@ class AutoModel:kwargs["frontend"] = frontend# build modelmodel_class = tables.model_classes.get(kwargs["model"])- model_conf = kwargs.get("model_conf", {})+ model_conf = {}+ deep_update(model_conf, kwargs.get("model_conf", {}))deep_update(model_conf, kwargs)model = model_class(**model_conf, vocab_size=vocab_size)model.to(device)