Meanwhile, a new version of nn.ar is out: Release v0.0.8-alpha · elgiano/nn.ar · GitHub
Enjoy, and let me know how it goes if you use this! <3
4 Likes
Question for everybody using RAVE:
what is the largest number of models you are able to run at the same time on your cpus?
we are setting up a larger-scale multichannel installation and I wonder if anyone has run around 16-20 models, and on which processor…
thanks!
Hi,
6 in a AMD Ryzen™ 7 PRO 8840HS w/ Radeon™ 780M Graphics × 16. I think I can get 7 or even 8 working. I wonder if the total amount of latent values can be a good measure instead of the amount of models. In my case they were 64 values considering the 6 models. Mind that I do not understand deeply the internal functioning of the models, and I did not test whether the same amount of latent values, despite the amount of models, uses the same resources.