Open ElmiraNa opened 3 years ago
I think it should use system RAM for the avatars array, and only send the current avatar to the GPU
When the images are loaded they are resized to 256*256, and I think with 32 bits/pixel. So unless I'm mistaken, having 1000 images would use 250MB RAM. I don't think there's any relevant constraint other than system RAM.
Thank you @JohanAR for your quick answer and that is right according to my today test after installing. please let me ask you another question if you would not mind: if I run the avatarify-python as a server side program, and I just get images from other clients, does it possible to run in a parallel form? or for each client I need to donate a separate GPU? (consider that the camera is just on server side and photos are send by multi clients at same time and in a parallel mode)
Sorry, I don't know that much about the server-client implementation.. But I would guess that you could run multiple server instances on the same computer, and they share the GPU
ElmiraNa pls I have a question for you, if you don’t mind
Hello, As I see in your code, the avatar images are taken in an array, (we should put them in avatar folder). is there any constraint about the number of photos? I mean does it use the memory of gpu or any image after being animated will be removed from the memory of gpu?