Open-source models transcribe 8.7 hours of clinic audio in 18 minutes on one GPU
Developer Maziyar Panahi demonstrated a fully open-source speech-to-text pipeline that transcribed 57 simulated GP consultations totaling 8 hours 38 minutes in just 18 minutes on a single GPU, achieving 27.8x real-time speed.
2026-11-19 ~ 2026-11-19 · 2 related posts
- One GPU transcribes 8.7 hours of mock GP calls in 18 minutes, 27.8x realtime — MaziyarPanahi · 2026-11-19
1 near-duplicate retellings: MaziyarPanahi