Simple Whisper

Implements the Whisper model via the Burn WGPU backend or whisper-rs.

Burn implmentation is based over sudomonikers/whisper-burn, provides abstraction over different models and languages.

Weights are automatically downloaded from Hugging Face.

Feature flags

• burn_vulkan = default enables the Burn WGPU backend
• whisper_cpp_vulkan = enables the alternative whisper.cpp backed using Vulkan shaders

Other resources

See newfla/simple-whisper for prebuilt cli & server binaries