baddbmm

Performs a batch matrix-matrix product of matrices in batch1 and batch2. x is added to the final result.

tch_baddbmm(x, batch1, batch2, beta = 1, alpha = 1)

Arguments

x	tensor object
batch1	the first batch of matrices to be multiplied
batch2	the second batch of matrices to be multiplied
beta	multiplier for x
alpha	multiplier for batch1 * batch2

Examples

x <- tensor(array(runif(45), dim = c(3, 3, 5)))
batch1 <- tensor(array(runif(36), dim = c(3, 3, 4)))
batch2 <- tensor(array(runif(60), dim = c(3, 4, 5)))
tch_baddbmm(x, batch1, batch2)
#> tensor 
#> (1,.,.) = 
#>   1.8235  1.8044  1.1738  1.3777  1.0120
#>   1.9102  3.7442  1.8650  1.5697  2.0777
#>   1.1358  2.5151  2.1979  1.1476  1.4231
#> 
#> (2,.,.) = 
#>   0.9822  1.4914  1.4809  1.8327  1.6085
#>   1.6679  2.4760  1.7026  2.9018  2.5509
#>   1.4881  1.7932  0.9740  2.1827  2.3150
#> 
#> (3,.,.) = 
#>   1.7097  1.6868  1.9578  0.7815  1.2295
#>   1.1103  1.2823  1.1187  0.5574  1.0791
#>   1.3530  1.4687  1.8853  1.5337  1.3748
#> [ Variable[CPUFloatType]{3,3,5} ]