Arda Göktoğan

KnetArray doesn't support broadcasting for Float16

2

When I tried the following code: ``` a = [1.0;2.0] a = convert( KnetArray{Float16,1} , a) a .* a ``` I get this error: ``` ERROR: MethodError: no method matching...

InexactError when broadcasting two KnetArrays

3

When I write this code: ``` a=rand(512,4,1024,1,1) b=rand(1,1,1024,16,64) a .* b ``` it computes the result but when I converted these arrays to `KnetArray{Flaot32,5}`: ``` a = convert( KnetArray{Float32,5}, a...

https://github.com/zihangdai/xlnet/blob/0d0dd5088d5ce2d72fa7890a633f990efbe3920f/modeling.py#L100 Layer normalization is done over last dimension (which is 768 in XLNet-Base), however, trained weights gamma and beta have a shape (768,). I thought that gamma and beta should...

Hello @manuelbrack, I was reviewing the source code and came across a line that seems to have a hardcoded value in a conditional statement. The line in question is: https://github.com/huggingface/diffusers/blob/0bae6e447cba0459456c4f7e7e87d7db141d3235/src/diffusers/pipelines/ledits_pp/pipeline_leditspp_stable_diffusion.py#L1053...