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
Love the library! I've been pseudo-polyfilling some of the other pytorch methods, but there was an issue I ran into that I can't find a good workaround for. Because of that I'm not sure if this should be classified as a bug, since its more of a design issue.
In pytorch, doing an operation such as .sum() doesn't produce a vector (shape = [1]), it produces a scalar (shape = []). This solves some edgecases and makes shape validation practical.
Is there any chance js-pytorch will be getting scalar support?
The text was updated successfully, but these errors were encountered:
Love the library! I've been pseudo-polyfilling some of the other pytorch methods, but there was an issue I ran into that I can't find a good workaround for. Because of that I'm not sure if this should be classified as a bug, since its more of a design issue.
In pytorch, doing an operation such as .sum() doesn't produce a vector (shape = [1]), it produces a scalar (shape = []). This solves some edgecases and makes shape validation practical.
Is there any chance js-pytorch will be getting scalar support?
The text was updated successfully, but these errors were encountered: