1bit.MONSTERDocs GitHub ↗

Whisper — Speech-to-Text

OpenAI Whisper V3 Turbo. Provides the speech-to-text (STT) stage of the JARVIS pipeline — full FFT/STFT + encoder-decoder path through GPU HIP kernels, with an NPU xclbin build ready.

Models

Model Backend(s) Status
Whisper-V3-Turbo NPU (5 xclbins) / GPU HIP 🔄

Notes

See also: JARVIS pipeline · full model support detail · all families