Repository Issue Activity (beta)

#416 Audio.cpp: Optimized VibeVoice impl in C++/ggml. Generate 90-min podcast in 22.95 min, 4.08x real-time, 2.86x Python on CUDA

#373 VibeVoice-ASR (4-bit MLX) decoder degenerates into infinite "Yes. Yes." loop on a 36-min recording

#415 Encountered a bug where the content keeps repeating during transcription

#361 Seems that the windows and macOS clients are no longer available

#414 Text-Only Inference Report for Upstream

#407 Speaker diarization collapses on tiny waveform changes (e.g. +0.7 dB linear gain)

#402 pip install -e . failed

#405 Aligning existing transcript to audio

#403 Performance Optimization: Fusing Pre-Attention Normalization to Eliminate HBM Materialization Bottlenecks

#199 Object of type dtype is not JSON serializable

#401 Alternative decoder

#400 Feature: Add SenseVoice/FunASR as alternative STT backend — non-autoregressive, 5x faster

#399 Down sampling of voices is producing noise in the speech

#376 为什么 demo 中的普通话带着英语口音 ?

#381 Is there a way to add a model I’ve downloaded from HuggingFace to Vibevoice?

#390 [demo] VibeVoice-ASR demo failed!

#391 How do I generate my own voice .pt file? What is the training process?

#392 App/Demo exception since fix: use weights_only=True (CWE-502)

#395 العلاج الطبيعي

#396 vllm_plugin: `_merge_mm_kwargs` IndexError at predict time with vllm 0.21.0 (cannot serve VibeVoice-ASR via vllm)

#398 Any plans to add hotwords weights later?

#322 Inquiry on DER Evaluation Method for AliMeeting

#394 Inference script and decoding settings for reproducing MLC results

#225 pip install vllm

#151 Polish language thread