fixed target device being fixated on CUDA on whisper model onnx export#244

Open

taken-yjyoon wants to merge 1 commit intoairockchip:mainfrom taken-yjyoon:main

+9-9

Commits on Dec 18, 2024

fixed a line that is causing whisper to load the model to cuda on supported devices, despite it is not needed.
taken-yjyoon
committed