Commit activity
Last 12 weeks · 0 commits
Community health
2 of 6 standards met
42
✓README✓License○Contributing○Code of Conduct○Issue Template○PR Template
Recent PRs & issues
Quiet · 2 in progress · Last activity 9mo ago
See all on GitHub →
AI-Generated Unit Tests This PR adds AI-generated unit tests to improve coverage and ensure key functionality is tested. Test coverage summary Test files and commits How to run tests is included as a dev dependency. ### Run tests directly ### Optional: add npm scripts You can add convenient scripts to : ### Usage examples Run tests once: Run tests in watch mode during development: Run tests with coverage summary: This contribution was created with assistance from ToTheos (https://totheos.com) to support test generation, code refactoring, and pull request preparation for an open-source codebases. The tool was used to automate routine software development tasks.
Great product. But when I'm using the application only the Tiny Models work and transcribing my audio. Other models just keep loading.
Fixes #5 issue. stop recording on when modal is closed. fix audio context closing error, as react-audio-visualize is already handling the closing of the audio context.
Recent fixes
This PR moves the heavy text summarization work into a dedicated Web Worker so that summarization no longer blocks the main thread. The change aims to improve UI responsiveness, reduce jank during summarization operations, and provide a clean fallback for environments where Workers are unavailable. Running summarization on the main thread can cause UI freezes for larger inputs and on lower-end devices. Offloading work to a Worker keeps the main thread responsive and improves perceived performance.
Fix incorrect model load progress reporting. This change adjusts progress calculation and emission so consumers receive accurate progress updates while models are loading.
@addyosmani This change makes the summarization model usage configurable instead of hard-coded. The goal is to allow selecting different summarization models without changing code, simplifying experimentation and future model swaps. !video
Structured data for AI agents
Repository: addyosmani/say. Description: Say - A Whisper AI Notes app Stars: 395, Forks: 42. Primary language: TypeScript. Languages: TypeScript (90.6%), JavaScript (7%), HTML (2.1%), CSS (0.2%). License: MIT. Homepage: https://say.addy.ie Topics: dictation, reactjs, transcription, voice-dictation, whisper-ai. Open PRs: 2, open issues: 2. Last activity: 9mo ago. Community health: 42%. Top contributors: addyosmani, anton-karlovskiy, Ashish-simpleCoder.