-
-
Notifications
You must be signed in to change notification settings - Fork 255
Description
I am having so much trouble with visomaster on new pc. My prior pc was a surface studio 2 with nvidia 1060 imbedded on intel processor chip. Visomaster ran on this fine with the auto window install file on github but the buffer was virtually unplayable and to slow. So I just bought this Asus pc and the video card is dedicated.
ROG GM700 Gaming Desktop - AMD Ryzen 7 8700F - 32GB Memory - NVIDIA GeForce RTX 5060Ti 16GB - 1TB SSD
I tried running the windows auto intall file and spent the whole night trying to get it to work but it doesn't. Not sure if it is the rtx 5060ti or the fact that it has an amd processor or the fact that the video card is dedicated.
After working on this all night: Copilot said the following:
⚡ Bottom Line
Your GPU is too new for current PyTorch binaries. Until CUDA 13.1 wheels with sm_120 support are published, the only stable way to run VisoMaster is CPU mode using CUDA_VISIBLE_DEVICES="".
Would you like me to show you exactly how the edited batch file should look line‑by‑line, so you can just copy‑paste it into Start_Portable.bat and lock VisoMaster into CPU mode automatically?
I can't believe that the program wont work with the current nvidia cards and that it worked with a 10 year old pc that had an nvidia 1060 card. Is there any advice you have since i spent all this money on this new pc just to use this program? Thanks.
copilot summary below: (this can't be the case because many people are using visomaster with rtx 50xx series video cards.
⚡ Bottom Line
Your GPU is too new for current PyTorch binaries. Until CUDA 13.1 wheels with sm_120 support are published, the only stable way to run VisoMaster is CPU mode using CUDA_VISIBLE_DEVICES="".
Would you like me to show you exactly how the edited batch file should look line‑by‑line, so you can just copy‑paste it into Start_Portable.bat and lock VisoMaster into CPU mode automatically?