-
-
Notifications
You must be signed in to change notification settings - Fork 2.1k
Open
Description
Hello, I am using WhisperX for a project in Blender, but the timings of each word aren't always accurate. What can contribute to accurate alignments?
Does the compute type/device matter? I am stuck using INT8/CPU. I wonder if it is possible to use FP16/CUDA on WSL2 with my RX 6800.
Does the model used matter?
Metadata
Metadata
Assignees
Labels
No labels