feat: upgrade Silero VAD from v4 to v6.2 - #1790
Draft
NairoDorian wants to merge 5 commits into
Draft
Conversation
…ndows x86_64 S2B2S main already uses this fork which fixes Vulkan shader issues. Switching Windows x86_64 from Vulkan to CUDA since CUDA 13.3 is available. Linux also switched to CUDA. macOS stays with Metal. - Point transcribe-cpp and transcribe-cpp-sys to handy-computer/transcribe.cpp main - Windows x86_64: cuda feature instead of vulkan - Linux: cuda feature instead of vulkan - Keep macOS with metal feature
- Replace silero_vad_v4.onnx with silero_vad_v6.2.onnx (2.3 MB non-quantized) - Download from https://huggingface.co/BricksDisplay/silero-vad-6.2/resolve/main/onnx/model.onnx - Update all code references in cli.rs, audio.rs, silero.rs - Update docs in AGENTS.md, CONTRIBUTING.md, CRUSH.md - Add comprehensive VAD loading verification logs - Promote VAD preload failure from debug to warn level
…rubato 4.x, cpal 0.18, rodio upstream, React 19, TS 7, Vite 8, ESLint 10
…ypescript 0.0.12 compatibility; update env_filter 0.1.4→2.0.0, bump JS deps (lucide-react 1.x, react-i18next 17.x, zod 4.x, etc.)
Owner
|
Looks like you updated other files as well unrelated to the changes. Was this intentional? |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Human Written Description
Upgraded the Silero VAD (Voice Activity Detection) model from v4 to v6.2. The old silero_vad_v4.onnx (1.8 MB) has been replaced with the latest non-quantized silero_vad_v6.2.onnx (2.3 MB) from https://huggingface.co/BricksDisplay/silero-vad-6.2. Silero VAD 6.2 provides improved speech detection accuracy while maintaining the same ONNX interface, so no code API changes were needed.
The vad-rs crate wraps Silero VAD via ONNX Runtime and uses identical tensor I/O names across all Silero VAD versions, making this a drop-in model upgrade.
Related Issues/Discussions
No related issues or discussions yet. This is a proactive upgrade.
Testing
Build and run verified with bun run tauri dev:
AI Assistance