Search notes:
torch: multiplication
>>> a = torch.tensor([ 2, 4, 3 ])
>>> a * 3
tensor([ 6, 12, 9])
>>> b = torch.tensor([ 5, 7, 2 ])
>>> torch.mul(a, b)
tensor([10, 28, 6])
>>> a * b
tensor([10, 28, 6])
>>> a = torch.tensor([[ 1, 2, 3, 4 ]])
>>> a
tensor([[1, 2, 3, 4]])
>>> b = torch.tensor([ [40], [30], [20], [10] ])
>>> b
tensor([[40],
[30],
[20],
[10]])
>>> a*b
tensor([[ 40, 80, 120, 160],
[ 30, 60, 90, 120],
[ 20, 40, 60, 80],
[ 10, 20, 30, 40]])
mm
import torch
m1 = torch.tensor([
[ 0.9 , 1.3 ],
[ 4.2 , 5.2 ],
[ 2.1 , 3.0 ]
])
m2 = torch.tensor([
[ 0.1 , 2.8 , 3.1 , 1.8 , 2.6 ],
[ 4.8 , 3.0 , 0.2 , 2.5 , 3.8 ]
])
r = m1.mm(m2) ; print(r)
r = torch.mm(m1, m2); print(r) # same thing
print(r)
Multiplying a 3x2 matrix with a 2x5 matrix produces a 3x5 matrix, the ouptut of print(r) is:
tensor([[ 6.3300, 6.4200, 3.0500, 4.8700, 7.2800],
[25.3800, 27.3600, 14.0600, 20.5600, 30.6800],
[14.6100, 14.8800, 7.1100, 11.2800, 16.8600]])