Alex Rogozhnikov

https://github.com/takluyver/bash_kernel/issues/13#issuecomment-527510754 solved the issue for me

@bprabhakar thanks for noting. Indeed, I have 1.8 and it diverges for me whatever I do.

it also works with tensorflow 1.9, but not tensorflow 1.8. I was not able to find something relevant in changelog

@Jinex2012 no, I would be stuck for days if I haven't seen this thread

ok, seems the problem in conv2d_transposed according to this thread https://github.com/tensorflow/tensorflow/issues/19200

Unfortunately, neither combination of numpy and k-means-constrained works for me. Tried 0.6.0 and numpy 1.20.* and numpy 1.20.3 Tried 0.7.0 and numpy 1.21.* and numpy 1.21.5 Tried building with --no-binary....

@shoyer, I've been thinking about this quite extensively in the past, and probably I'm settled about how that would like. Not expecting everyone to like it though. I've put some...

Julia's line-level macros really shine for this kind of stuff ``` @tullio M[x,y,c] := N[x+i, y+j,c] * K[i,j] # sum over i,j, and create M ``` Here is the problem...

Hi @lvyufeng thanks for a well-organized PR. Github CI does not have cuda, so effectively tests were turned off in your branch. I've modified it to include tests (but skip...

Thanks! I am not mentioning this in the docs for purpose as I expect this feature to be misused. So, for now I'd better keep it as is.