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
Actually there is. Using the sklearn function involves the necessity to convert from torch.Tensor to numpy.ndarray before converting. While this is totally fine on cpu (no memory copy made here), this really causes a slow-down for gpu tensors, since it involves a GPU-sync.
For now #1327 already implements a sklearn version, but it shall be replaced with a native torch one.
馃殌 Feature
Implement general AUC (to be combined with other metrics like ROC)
The text was updated successfully, but these errors were encountered: