All pull requests

Pull requests

• Resample ASR array inputs from the given sampling_rate

  #460·Kayvan Zahiri (Kayvan-Zahiri) opened Oct 14, 2026·

  ·

• Add a vLLM image for streaming ASR

  #456·mallocnew opened Sep 28, 2026·

  ·
  1

  1

• Fix session-scoped stopping in the ASR Gradio demo

  #454·Rajat Jaiswal (onlinerj) opened Sep 20, 2026·

  ·
  1

  1

• fix: use torch_dtype instead of dtype in ASR from_pretrained calls

  #452·Ammar Hashmi (ammar-iitm) opened Sep 16, 2026·2 of 3 tasks·

  ·
  1

• feat(demo): add copy-friendly Plain Text tab to ASR gradio demo

  #451·Ammar Hashmi (ammar-iitm) opened Sep 16, 2026·1 of 2 tasks·

  ·
  1

• docs: document VRAM-vs-audio-duration ceiling for VibeVoice-ASR

  #450·Ammar Hashmi (ammar-iitm) opened Sep 15, 2026·

  ·
  1

• docs: note venv workaround for PEP 668 install friction outside NVIDIA containers

  #449·Ammar Hashmi (ammar-iitm) opened Sep 15, 2026·

  ·

• Fix pip install -e . editable-install failures

  #448·Ammar Hashmi (ammar-iitm) opened Sep 11, 2026·2 tasks done·

  ·
  1

• vllm_plugin: fix multimodal-processing API drift for vLLM >=0.15 (incl. 0.21)

  #447·Ammar Hashmi (ammar-iitm) opened Sep 11, 2026·2 of 3 tasks·

  ·
  1

• fix(asr): add exempt_numeral_tokens option to prevent digit corruption under repetition penalty

  #445·Syl (syl-ky) opened Sep 8, 2026·

  ·
  1

• fix: collapse decode-loop repeats when assembling streaming ASR segments

  #444·Ben Younes (ousamabenyounes) opened Sep 6, 2026·

  ·
  1

• fix(tts): normalize Chinese punctuation before synthesis

  #443·Ben Younes (ousamabenyounes) opened Sep 5, 2026·

  ·
  1

• Fix vllm_plugin import crash when vLLM's AudioMediaIO location changes

  #442·Ammar Hashmi (ammar-iitm) opened Sep 5, 2026·2 tasks done·

  ·
  1

• Fix device mismatch in VibeVoiceASR.encode_speech under device_map=auto

  #441·Ammar Hashmi (ammar-iitm) opened Sep 4, 2026·2 of 3 tasks·

  ·
  1

• Fix weights_only=True load failure for bundled voice presets

  #440·Ammar Hashmi (ammar-iitm) opened Sep 4, 2026·

  ·
  1

• Fix GPU memory accumulation across files in ASR batch inference demo

  #439·Ammar Hashmi (ammar-iitm) opened Sep 4, 2026·1 of 2 tasks·

  ·
  1

  2

• Add VibeVoice 1.5B Transformers release tooling

  #438·Justin Chu (justinchuby) opened Sep 4, 2026·

  ·

• build(deps): bump transformers from 4.51.3 to 5.10.1

  dependenciespython
  #434·dependabot[bot] opened Sep 2, 2026Bot·

  ·

• Fix broken star history chart

  #430·Dessalines39394 opened Aug 17, 2026·

  ·

• fix: declare supported vLLM installation extra

  #429·guyua9 opened Aug 13, 2026·

  ·

• fix(audio): soft-knee limiter in AudioNormalizer.avoid_clipping

  #427·Aaron (aarochu) opened Jul 30, 2026·3 tasks done·

  ·
  1

  1

• Add demo link for VibeVoice-ASR-BitNet in README.md

  #426·Yan Xia (sd983527) opened Jul 30, 2026Contributor·

  ·

• Fix three critical bugs: distributed all_reduce on scalars, undefined speech_len, and async loop crash

  #424·Adwait M. (adwaitm1301) opened Jul 29, 2026·

  ·
  1

• fix(asr): derive speech token compression ratio from encoder config

  #420·Urvi Tyagi (Urvity03) opened Jul 15, 2026·

  ·
  1

• deps: bump accelerate 1.13.0 → 1.14.0

  #413·bytheby72 opened Jun 22, 2026·

  ·
  1