Testing Audio Transcription with GPU

[paulobreim]

I'm testing audio transcription. So I submitted a single video file using the option implementationClass = iped.engine.task.transcript.WhisperTranscriptTask.
I'm attaching the configuration file and the 2 logs. The only difference in the test was the device parameter, which I set to device = gpu for one test and device = cpu for the other test.
The configuration for device = gpu didn't transcribe anything, and the one with device = cpu performed the transcription but took about 1 hour.
I believe the installation of what's needed for the GPU is correct.
Could someone take a quick look at the files and see if they can find the problem?
AudioTranscriptConfig.txt
IPED-2025-11-24-21-04-51.log
IPED-2025-11-24-21-07-46.log

Thank you
paulo