facebookresearch / robust_cvd

Robust Consistent Video Depth Estimation
https://robust-cvd.github.io/
Other
300 stars 24 forks source link

Torch and cuda version #38

Open XihangYU630 opened 1 year ago

XihangYU630 commented 1 year ago

What torch and cuda version this codebase use? Since one error arises: RuntimeError: CUDA error: no kernel image is available for execution on the device. It arises when running "robust_cvd/cvd2/monodepth/midas_v2_model.py", line 50, in estimatedepth. input = (input_ - self.norm_mean.to(self.device)) / self.norm_stdev.to(self.device)