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
Commits on main
fix(runtime): release multi-input ONNX VAD fix as 0.4.6 (#3834)
4b19f43 zhifu gao committed 16 days ago
fix(vad): apply the offline dynamic silence schedule per frame
c9a4feb lingfengyao committed 16 days ago
fix(punctuation): preserve streaming cache on empty input (#3824)
2c54fd2 zhifu gao committed 16 days ago
fix(runtime): release streaming ONNX wrapper fixes as 0.4.5 (#3828)
6c20419 zhifu gao committed 17 days ago
fix(runtime): handle streaming Paraformer chunks without features (#3827)
7eae5ba zhifu gao committed 17 days ago
fix(runtime): handle streaming Paraformer chunks without features
cf22a34 zhifu gao committed 17 days ago
fix(runtime): release streaming ONNX wrapper fixes as 0.4.4 (#3826)
a71ad60 zhifu gao committed 17 days ago
fix(onnx): return [] and reset state at the end of short VAD streams
f903065 Tanishq committed 18 days ago
fix(punc): return an empty result for empty text in CTTransformerStreaming
8dfa8f5 Tanishq committed 18 days ago
fix(onnx): avoid UnboundLocalError in Fsmn_vad_online for chunks without feature frames
a45faf9 Tanishq committed 18 days ago
docs: scope Ascend compatibility and latency evidence (#3820)
8710e1c zhifu gao committed 18 days ago
fix(auto): preserve optional speaker dependency errors (#3819)
2b8b15a zhifu gao committed 18 days ago
fix(runtime): release the repaired Torch package as 0.1.4 (#3817)
c8d3252 zhifu gao committed 18 days ago