Hungry, hungry, Hedy?
complete
Julian
complete
Thanks for all the feedback! There was indeed an issue with the GPU optimization, and the CPU was carrying too much load. This is resolved in version 2.0.2.
A
AJ
I have the exact same issue with an M2 Pro. The fans spin up as if the laptop is going to take off.
Frankly, I didn't think that the large model's size was too big, considering that I have other apps using far larger models for transcription, and the laptop barely breaks a sweat doing it. Here are some screenshots from other apps in case there are approaches in there that Hedy could use (WhisperKit is what works best but neither use much resources).
Julian
Thanks for sharing this! Did you enable any options, such as realtime transcripts or the large model? The CPU load is significantly higher than what we've seen for the basic settings in our testing using M2 and M4 macbooks.
S
Stephen
Julian I enabled the large model, but left transcript speed set to the standard. No session context and set to business meetings for session type.
Julian
Stephen thanks! From the initial beta feedback we're getting, we might have pushed it to hard with the large model. It works, but everyone's fans are running like crazy 🤣 We'll probably scale it back to a smaller model in the next update.