stt-realtime/requirements.txt

# Low-latency STT dependencies
websockets>=12.0
numpy>=1.24.0

# Faster-whisper backend (GPU accelerated)
faster-whisper>=1.0.0
ctranslate2>=4.4.0

# Audio processing
soundfile>=0.12.0

# VAD - Silero (loaded via torch.hub, runs on CPU via ONNX)
# Requires torch (CPU-only) - see requirements-gpu-torch.txt

# Utilities
aiohttp>=3.9.0
refactor: Implement low-latency STT pipeline with speculative transcription Major architectural overhaul of the speech-to-text pipeline for real-time voice chat: STT Server Rewrite: - Replaced RealtimeSTT dependency with direct Silero VAD + Faster-Whisper integration - Achieved sub-second latency by eliminating unnecessary abstractions - Uses small.en Whisper model for fast transcription (~850ms) Speculative Transcription (NEW): - Start transcribing at 150ms silence (speculative) while still listening - If speech continues, discard speculative result and keep buffering - If 400ms silence confirmed, use pre-computed speculative result immediately - Reduces latency by ~250-850ms for typical utterances with clear pauses VAD Implementation: - Silero VAD with ONNX (CPU-efficient) for 32ms chunk processing - Direct speech boundary detection without RealtimeSTT overhead - Configurable thresholds for silence detection (400ms final, 150ms speculative) Architecture: - Single Whisper model loaded once, shared across sessions - VAD runs on every 512-sample chunk for immediate speech detection - Background transcription worker thread for non-blocking processing - Greedy decoding (beam_size=1) for maximum speed Performance: - Previous: 400ms silence wait + ~850ms transcription = ~1.25s total latency - Current: 400ms silence wait + 0ms (speculative ready) = ~400ms (best case) - Single model reduces VRAM usage, prevents OOM on GTX 1660 Container Manager Updates: - Updated health check logic to work with new response format - Changed from checking 'warmed_up' flag to just 'status: ready' - Improved terminology from 'warmup' to 'models loading' Files Changed: - stt-realtime/stt_server.py: Complete rewrite with Silero VAD + speculative transcription - stt-realtime/requirements.txt: Removed RealtimeSTT, using torch.hub for Silero VAD - bot/utils/container_manager.py: Updated health check for new STT response format - bot/api.py: Updated docstring to reflect new architecture - backups/: Archived old RealtimeSTT-based implementation This addresses low latency requirements while maintaining accuracy with configurable speech detection thresholds. 2026-01-22 22:08:07 +02:00			`# Low-latency STT dependencies`
Implemented experimental real production ready voice chat, relegated old flow to voice debug mode. New Web UI panel for Voice Chat. 2026-01-20 23:06:17 +02:00			`websockets>=12.0`
			`numpy>=1.24.0`

refactor: Implement low-latency STT pipeline with speculative transcription Major architectural overhaul of the speech-to-text pipeline for real-time voice chat: STT Server Rewrite: - Replaced RealtimeSTT dependency with direct Silero VAD + Faster-Whisper integration - Achieved sub-second latency by eliminating unnecessary abstractions - Uses small.en Whisper model for fast transcription (~850ms) Speculative Transcription (NEW): - Start transcribing at 150ms silence (speculative) while still listening - If speech continues, discard speculative result and keep buffering - If 400ms silence confirmed, use pre-computed speculative result immediately - Reduces latency by ~250-850ms for typical utterances with clear pauses VAD Implementation: - Silero VAD with ONNX (CPU-efficient) for 32ms chunk processing - Direct speech boundary detection without RealtimeSTT overhead - Configurable thresholds for silence detection (400ms final, 150ms speculative) Architecture: - Single Whisper model loaded once, shared across sessions - VAD runs on every 512-sample chunk for immediate speech detection - Background transcription worker thread for non-blocking processing - Greedy decoding (beam_size=1) for maximum speed Performance: - Previous: 400ms silence wait + ~850ms transcription = ~1.25s total latency - Current: 400ms silence wait + 0ms (speculative ready) = ~400ms (best case) - Single model reduces VRAM usage, prevents OOM on GTX 1660 Container Manager Updates: - Updated health check logic to work with new response format - Changed from checking 'warmed_up' flag to just 'status: ready' - Improved terminology from 'warmup' to 'models loading' Files Changed: - stt-realtime/stt_server.py: Complete rewrite with Silero VAD + speculative transcription - stt-realtime/requirements.txt: Removed RealtimeSTT, using torch.hub for Silero VAD - bot/utils/container_manager.py: Updated health check for new STT response format - bot/api.py: Updated docstring to reflect new architecture - backups/: Archived old RealtimeSTT-based implementation This addresses low latency requirements while maintaining accuracy with configurable speech detection thresholds. 2026-01-22 22:08:07 +02:00			`# Faster-whisper backend (GPU accelerated)`
Implemented experimental real production ready voice chat, relegated old flow to voice debug mode. New Web UI panel for Voice Chat. 2026-01-20 23:06:17 +02:00			`faster-whisper>=1.0.0`
			`ctranslate2>=4.4.0`

			`# Audio processing`
			`soundfile>=0.12.0`

perf: reduce container sizes and build times - miku-stt: switch PyTorch CUDA -> CPU-only (~2.5 GB savings) - Silero VAD already runs on CPU via ONNX (onnx=True), CUDA PyTorch was waste - faster-whisper/CTranslate2 uses CUDA directly, no PyTorch GPU needed - torch+torchaudio layer: 3.3 GB -> 796 MB; total image 9+ GB -> 6.83 GB - Tested: Silero VAD loads (ONNX), Whisper loads on cuda, server ready - llama-swap-rocm: add root .dockerignore to fix 31 GB build context - Dockerfile clones all sources from git, never COPYs from context - 19 GB of GGUF model files were being transferred on every build - Now excludes everything (*), near-zero context transfer - anime-face-detector: add .dockerignore to exclude accumulated outputs - api/outputs/ (56 accumulated detection files) no longer baked into image - api/__pycache__/ and images/ also excluded - .gitignore: remove .dockerignore exclusion so these files are tracked 2026-02-25 14:41:04 +02:00			`# VAD - Silero (loaded via torch.hub, runs on CPU via ONNX)`
			`# Requires torch (CPU-only) - see requirements-gpu-torch.txt`
Implemented experimental real production ready voice chat, relegated old flow to voice debug mode. New Web UI panel for Voice Chat. 2026-01-20 23:06:17 +02:00
			`# Utilities`
			`aiohttp>=3.9.0`