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
Merge pull request #3751: explicit PCM16 input and VAD source rates
bb303f8 zhifu gao committed 1 month ago
feat(audio): support explicit PCM16 input without container probing
9d61129 LauraGPT committed 1 month ago
ci: run rich transcription regressions in ONNX package checks (#3749)
4a84879 zhifu gao committed 1 month ago
ci: run rich transcription regressions in ONNX package checks
afd5c88 LauraGPT committed 1 month ago
fix(runtime): skip empty deduplicated event segments (#3748)
800f756 zhifu gao committed 1 month ago
fix(sensevoice): stop rendering the Cough event as the Sneeze emoji (#3747)
217d10a Mohammad Hijjawi committed 1 month ago
fix: keep token separators in timestamp_sentence raw_text (#3746)
b0fa25b Mohammad Hijjawi committed 1 month ago
fix(hotwords): stop fuzzy postprocess matching from rewriting nearby text (#3744)
43329e8 Chi-Wei Lee committed 1 month ago
docs: surface merged speech-to-speech SenseVoice integration (#3743)
460f006 zhifu gao committed 1 month ago
fix: stop timestamp_sentence_en reusing a stale sentence start (#3741)
be7557f Chi-Wei Lee committed 1 month ago
fix(auto): warn when unavailable accelerators fall back to CPU (#3742)
7b098ac zhifu gao committed 1 month ago
fix(auto): warn when unavailable accelerators fall back to CPU
9c9067f LauraGPT committed 1 month ago
fix(amp): resolve autocast/GradScaler device type for non-CUDA accelerators (#3735)
7f85ddd li-lizhe committed 1 month ago
Merge pull request #3736: correct LiteLLM website transcription route
12e417f zhifu gao committed 1 month ago
Merge pull request #3737: preserve SenseVoice word boundaries
15c1959 zhifu gao committed 1 month ago
fix(sensevoice): keep words separate at standalone boundaries
a57cc3f LauraGPT committed 1 month ago
fix(sensevoice): normalize first word before joining subpieces
2ddabb2 LauraGPT committed 1 month ago
feat(benchmark): report first nonempty transcript latency (#3734)
55203a9 zhifu gao committed 1 month ago
Merge pull request #3733 from modelscope/codex/speaker-overlap-accounting-20260928
e1ceba0 zhifu gao committed 1 month ago