You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I conducted the Gaussian Mixture experiments using SCVM-NODE, following the instructions provided in the Readme file. However, I encountered an estimated training time of ten hours, which is significantly longer than the few minutes mentioned in your paper. I used a single V100 GPU for this experiment. Could you please clarify what I might be missing?
The text was updated successfully, but these errors were encountered:
Hi, could you double-check if JAX was installed correctly and that you were using GPU when running the code (if JAX failed to initialize CUDA, all tensors will be on the CPU)? I just tried on my lab machine (which has A6000 GPU) and the training for the mixture of Gaussians example should take under 10min.
I conducted the Gaussian Mixture experiments using SCVM-NODE, following the instructions provided in the Readme file. However, I encountered an estimated training time of ten hours, which is significantly longer than the few minutes mentioned in your paper. I used a single V100 GPU for this experiment. Could you please clarify what I might be missing?
The text was updated successfully, but these errors were encountered: