Lines Matching +full:backwards +full:- +full:incompatible
1 .. _broadcasting-semantics:
13 -----------------
16 - Each tensor has at least one dimension.
17 - When iterating over the dimension sizes, starting at the trailing dimension,
48 - If the number of dimensions of :attr:`x` and :attr:`y` are not equal, prepend 1
50 - Then, for each dimension size, the resulting dimension size is the max of the sizes of
70 …RuntimeError: The size of tensor a (2) must match the size of tensor b (3) at non-singleton dimens…
72 In-place semantics
73 ------------------
74 One complication is that in-place operations do not allow the in-place tensor to change shape
88 …RuntimeError: The expanded size of the tensor (1) must match the existing size (7) at non-singleto…
90 Backwards compatibility
91 -----------------------
94 out by viewing each tensor as 1-dimensional. PyTorch now supports broadcasting and the "1-dimensio…
98 Note that the introduction of broadcasting can cause backwards incompatible changes in the case whe…
105 In order to help identify cases in your code where backwards incompatibilities introduced by broadc…
114 …Changing behavior in a backwards incompatible manner to broadcasting rather than viewing as 1-dime…