V10#13

Merged

cs-util merged 30 commits intomainfrom

v10

Oct 31, 2025

Commits on Oct 28, 2025

Implement client-side chunked transcription pipeline
cs-util
committed

Commits on Oct 29, 2025

Disable Multi-Threading: Set ort.env.wasm.numThreads = 1 to avoid the cross-origin isolation requirement
cs-util
committed
Refactor chunking tests for clarity and accuracy; update VAD to use CDN for ONNX Runtime
cs-util
committed
Update ONNX Runtime import to use CDN and disable multi-threading for cross-origin isolation
cs-util
committed
Update ONNX Runtime import to use ESM and adjust multi-threading settings for cross-origin isolation
cs-util
committed
Refactor ONNX Runtime import to simplify promise handling and ensure single-threaded execution for cross-origin isolation
cs-util
committed
Enhance error handling in ONNX Runtime import to check for module availability and InferenceSession
cs-util
committed
Enhance error handling in ensureSession to check for InferenceSession.create availability
cs-util
committed
Enhance ONNX Runtime import to support multiple CDN sources and improve error handling for module availability
cs-util
committed
Add diagnostic logging for ONNX Runtime Web URL loading
cs-util
committed
Add fetch script for ONNX Runtime Web distribution assets
cs-util
committed
Added local ort vendor directory
cs-util
committed
Added explicit ignore patterns for ort/** in eslint.config.js so ESLint's flat config will exclude the local vendor files.
cs-util
committed
Add local ORT and Silero VAD model paths for development
cs-util
committed
Update src/stt/vad.js

cs-util
and
gemini-code-assist[bot]
authored
Update scripts/fetch-onnx-dist.sh

cs-util
and
gemini-code-assist[bot]
authored
Update scripts/fetch-onnx-dist.sh

cs-util
and
gemini-code-assist[bot]
authored
Merge commit 'b2de2cd577b316d1780fd015f9952886de730b50' into v8
cs-util
committed
updated the default model URL to the v4 version, commented out the local model override in index.html (since the file doesn't exist), and disabled the custom WASM paths to ensure compatibility with…
cs-util
committed
Add Silero VAD model path and update README instructions for local setup
cs-util
committed
Add logging for ONNX Runtime object keys and InferenceSession details
cs-util
committed
Update default ONNX Runtime version in fetch script and enhance VAD session logging
cs-util
committed
Updated ONNX Runtime Web to version 1.19.0 and added logging to show the model URL being used and the exact error during session creation. This should help identify if the issue is with the model f…
cs-util
committed
Update Silero VAD model path to '/model/silero_vad.onnx' and reflect changes in README
cs-util
committed
Fix tensor shape in createStateTensor function for correct dimensionality
cs-util
committed
Fix tensor shape in createStateTensor function for correct dimensionality
cs-util
committed
Enhance logging in detectSpeechSegments function to reduce console output frequency and update state tensor assignment
cs-util
committed
Refactor logging in VAD functions for improved readability and consistency
cs-util
committed
unused createHiddenTensor function
cs-util
committed
Add testing helpers for injecting fake ort/session and enhance tests for detectSpeechSegments
cs-util
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

V10#13

V10#13
cs-util merged 30 commits intomainfrom
v10

Commits on Oct 28, 2025

Implement client-side chunked transcription pipeline

Commits on Oct 29, 2025

Disable Multi-Threading: Set ort.env.wasm.numThreads = 1 to avoid the cross-origin isolation requirement

Refactor chunking tests for clarity and accuracy; update VAD to use CDN for ONNX Runtime

Update ONNX Runtime import to use CDN and disable multi-threading for cross-origin isolation

Update ONNX Runtime import to use ESM and adjust multi-threading settings for cross-origin isolation

Refactor ONNX Runtime import to simplify promise handling and ensure single-threaded execution for cross-origin isolation

Enhance error handling in ONNX Runtime import to check for module availability and InferenceSession

Enhance error handling in ensureSession to check for InferenceSession.create availability

Enhance ONNX Runtime import to support multiple CDN sources and improve error handling for module availability

Add diagnostic logging for ONNX Runtime Web URL loading

Add fetch script for ONNX Runtime Web distribution assets

Added local ort vendor directory

Added explicit ignore patterns for ort/** in eslint.config.js so ESLint's flat config will exclude the local vendor files.

Add local ORT and Silero VAD model paths for development

Update src/stt/vad.js

Update scripts/fetch-onnx-dist.sh

Update scripts/fetch-onnx-dist.sh

Merge commit 'b2de2cd577b316d1780fd015f9952886de730b50' into v8

updated the default model URL to the v4 version, commented out the local model override in index.html (since the file doesn't exist), and disabled the custom WASM paths to ensure compatibility with…

Add Silero VAD model path and update README instructions for local setup

Add logging for ONNX Runtime object keys and InferenceSession details

Update default ONNX Runtime version in fetch script and enhance VAD session logging

Updated ONNX Runtime Web to version 1.19.0 and added logging to show the model URL being used and the exact error during session creation. This should help identify if the issue is with the model f…

Update Silero VAD model path to '/model/silero_vad.onnx' and reflect changes in README

Fix tensor shape in createStateTensor function for correct dimensionality

Fix tensor shape in createStateTensor function for correct dimensionality

Enhance logging in detectSpeechSegments function to reduce console output frequency and update state tensor assignment

Refactor logging in VAD functions for improved readability and consistency

unused createHiddenTensor function

Add testing helpers for injecting fake ort/session and enhance tests for detectSpeechSegments