A basic introduction to NumPy's einsum

Re:A basic introduction to NumPy's einsum by teddykoker on Saturday April 09, 2022 @11:01PM (#30974127)

Also see Einops: https://github.com/arogozhnikov/einops, which uses a einsum-like notation for various tensor operations used in deep learning.

https://einops.rocks/pytorch-examples.html shows how it can be used to implement various neural network architectures in a more simplified manor.

[ Parent | on HN ]

Re:A basic introduction to NumPy's einsum by johnnymellor on Sunday April 10, 2022 @05:58AM (#30975710)

Einops looks nice! It reminds me of https://github.com/deepmind/einshape which is another attempt at unifying reshape, squeeze, expand_dims, transpose, tile, flatten, etc under an einsum-inspired DSL.

[ Parent | on HN ]

Re:A basic introduction to NumPy's einsum by pizza on Sunday April 10, 2022 @03:13AM (#30975053)

Somebody also realized that much of the time you can use one single function to describe all 3 of the einops operations. I present to you, einop: https://github.com/cgarciae/einop

[ Parent | on HN ]

Re:A basic introduction to NumPy's einsum by PartiallyTyped on Sunday April 10, 2022 @01:21AM (#30974665)

In addition, einsum in jax uses einops.

[ Parent | on HN ]