Slow inference

[Echolink50]

I couldn't find a solid answer on the speed of inference so I just gave it a try. Running RTX 2060 12GB. My examples didn't run with some tensors not found or some kinda error. I ran demo 1 and it took 10 minutes for the 10 second clip. Is this normal? I have to say for a 10GB worth of models and pickle files that seems disappointing. Thanks for the program though

[petermg]

think that's normal. I'm on a 4090 and it takes a while. But I think the results are totally worth it.

[Echolink50]

Thanks. The results are not good enough to justify how long it takes in my case. Maybe if you have better hardware like a RTX XX90 series its faster and worth the time.