Underrstanding matrix multiplication for 3D tensors

[sumitfromculinda]

Hi As suggested that for matrix multiplication the input dimension should match.

But when I try the below, I am getting an error:

new_tensor=torch.matmul(torch.rand(224,224,3),torch.rand(3,224,224))

Below is the error:
RuntimeError: The size of tensor a (224) must match the size of tensor b (3) at non-singleton dimension 0

Please suggest why we are getting this error.