콘텐츠로 이동

safestacksilero-vad

SherpaONNXVAD

Uses sherpa-onnx streaming Silero state with an explicit threshold and segment padding.

Voice activity detectionVoiceHub-nativenative-vad-dispatchParameters: Not reportedNot text-language conditionedTraining: nativeLicense: LicenseRef-TEN-VAD-Open-Source-License

Parameter metadata: Not reported: the audited metadata available for the registered default does not provide an exact parameter total.

Use this model Checkpoint
Resources
Upstream GitHub VoiceHub source Open in Colab

Usage

Complete the VoiceHub installation once, then run this repository-authored example. Model pages intentionally contain no package-install command.

This example is maintained against VoiceHub's public API; it is not copied from an upstream demo or package README.

Model-specific path: Uses sherpa-onnx streaming Silero state with an explicit threshold and segment padding.

Inputs and controls: Keep streaming state per audio stream; do not share one detector instance across unrelated calls.

from pathlib import Path

from voicehub import AutoModelForVoiceActivityDetection

AUDIO_FILE = Path("speech.wav")
if not AUDIO_FILE.is_file():
    raise FileNotFoundError(AUDIO_FILE)

model = AutoModelForVoiceActivityDetection.from_pretrained(
    'safestack/silero-vad',
    model_type='vad_sherpa_onnx',
    device="cpu",
    lazy_load=True,
)
output = model.detect(
    AUDIO_FILE,
    threshold=0.5,
    speech_pad_ms=100,
    max_speech_duration_s=30.0,
)
for segment in output.segments:
    print(segment.start, segment.end, segment.score)

Use authorized recordings. Verify hardware needs and pin a revision in production.

Overview

vad_sherpa_onnx is a VoiceHub voice activity detection integration. This page is generated from its registry contract. Open the vad_sherpa_onnx Colab notebook.

Property Value
Task Voice activity detection
Architecture native-vad-dispatch
Runtime VoiceHub-native
Languages Not text-language conditioned
Capabilities voice-activity-detection, voicehub-native, safetensors, explicit-onnx-weight-conversion, fine-tuning, streaming, sherpa-compatible-segmentation, silero, ten-vad
Reusable components —
Normalized output VADOutput

Language support

The public VAD contract does not select a spoken language; validate checkpoint acoustic coverage on the target languages and recording conditions.

Paper and GitHub

Configuration

Load configuration without constructing the model:

from voicehub import AutoConfig

config = AutoConfig.for_model('vad_sherpa_onnx')
print(config.model_type)
Property Value
Canonical model type vad_sherpa_onnx
Configuration class SherpaONNXVADConfig
Architecture class SherpaONNXVADForVoiceActivityDetection

Processing

Create the registered processor without allocating model weights:

from voicehub import AutoProcessor

processor = AutoProcessor.from_pretrained(
    'safestack/silero-vad',
    model_type='vad_sherpa_onnx',
)
print(type(processor).__name__)

Inference

The Usage example returns VADOutput through AutoModelForVoiceActivityDetection.

Input and output contract

Property Value
Label boundary Clip-, frame-, or segment-level labels
Required training inputs labels

Use authorized audio and preserve annotation provenance. See the ASR and VAD data workflow.

Training and optimization

Use available_optimization_passes() to discover reversible public passes. Unsupported runtime or hardware fails closed before mutation.

Training contract

Property Value
Support native
Family frame-classification
Recipe single-phase
Default phase voice_activity_detection
Training checkpoint safestack/silero-vad
Native training graph yes
Phase Kind Components Required inputs Loss keys
voice_activity_detection objective model labels loss

The integration accepts its declared source or prepared contract directly. Call model.validate_training_support() first, then follow the training workflow.

Checkpoints, provenance, license, and limitations

Property Value
Default checkpoint safestack/silero-vad
Hugging Face ID safestack/silero-vad
Repository availability verified through the Hugging Face model API on 2026-08-11; pin a revision before production use.
Checkpoint status Registry default; pin an immutable revision for production and reproducible evidence
Optional dependency extra Core package
Hardware and runtime Usage selects cpu; verify checkpoint-specific requirements
Real-checkpoint evidence Release evidence; a registry default alone is not execution evidence
Implementation voicehub.models.vad_sherpa_onnx.modeling_vad_sherpa_onnx.SherpaONNXVADForVoiceActivityDetection
Configuration voicehub.models.vad_sherpa_onnx.configuration_vad_sherpa_onnx.SherpaONNXVADConfig
Source provenance voicehub/architectures/ten_vad/SOURCE.json
License LicenseRef-TEN-VAD-Open-Source-License

The provider's optional TEN family is governed by a non-standard license with additional deployment restrictions, including limits on competing with Agora. Review the bundled THIRD_PARTY_LICENSE before conversion, fine-tuning, distribution, or deployment. The default Silero family retains its own checkpoint terms. Commercial use: review required.

Confirm the checkpoint revision, access terms, provenance, and license.

Limitations

  • No integration-specific checkpoint limitation is registered. Verify the selected checkpoint revision and its documented runtime requirements.
  • Validate memory, precision, and optional dependencies on the target system.
  • Public optimizations fail closed when the runtime or hardware cannot satisfy their validation contract; an unavailable pass is not reported as applied.
  • Contract tests do not replace the linked released-checkpoint evidence.

Public API

Use the stable configuration, processor, and task-model facades below.

Configuration

SherpaONNXVADConfig

View source

SherpaONNXVADConfig(**config_kwargs)

Parameters

  • **config_kwargs — Configuration fields validated by SherpaONNXVADConfig.

Model

SherpaONNXVADForVoiceActivityDetection

View source

AutoModelForVoiceActivityDetection.from_pretrained(
    pretrained_model_name_or_path,
    *,
    model_type='vad_sherpa_onnx',
    config=None,
    **model_kwargs,
)

Parameters

  • pretrained_model_name_or_path — Hub ID or compatible local directory.
  • model_type — Canonical model type; use 'vad_sherpa_onnx'.
  • config — Optional preloaded SherpaONNXVADConfig instance.
  • **model_kwargs — Model-specific loading arguments.
from voicehub import get_model_spec

spec = get_model_spec('vad_sherpa_onnx')
print(spec.display_name, spec.task.value)
Purpose Public object
Discover get_model_spec('vad_sherpa_onnx')
Load and run AutoModelForVoiceActivityDetection
Configure SherpaONNXVADConfig
Process AutoProcessor
Model implementation SherpaONNXVADForVoiceActivityDetection
Normalized output VADOutput
Training contract get_training_spec('vad_sherpa_onnx')
Optimization lifecycle available_optimization_passes, apply_optimization_plan, optimization_manifest, restore_optimization_plan

See all model guides, inference, and the training matrix.