新着
lnenad/lipwise — 音声エンジンのビルドを AVX2 基準で固定
lnenad/lipwise が内蔵の音声認識エンジンを、ビルドサーバーの CPU ではなく AVX2・FMA・F16C・BMI2 基準で固定し、AVX-512 対応ランナーが生成した Windows ビルドの illegal instruction クラッシュを回避した。
lnenad/lipwise が内蔵の音声認識エンジンを、ビルドサーバーの CPU ではなく AVX2・FMA・F16C・BMI2 基準で固定し、AVX-512 対応ランナーが生成した Windows ビルドの illegal instruction クラッシュを回避した。
GitHub の説明: Private voice dictation for Windows, macOS and Linux. Hold a shortcut, talk, and Lipwise types what you meant into any app: speech recognition runs on your computer, and an AI editor understands "scratch that", lists and rewrites. Works with local AI, Claude or OpenAI.
GitHub で見る ↗lnenad/lipwise が内蔵の音声認識エンジンを、ビルドサーバーの CPU ではなく AVX2・FMA・F16C・BMI2 基準で固定し、AVX-512 対応ランナーが生成した Windows ビルドの illegal instruction クラッシュを回避した。
該当するニュースはありません。