Search results

Type instabilities lead to insane number of CPU allocations on grouped convolutions

performance

Status: Open.

#520 In FluxML/NNlib.jl;

· mashu opened on Jul 5, 2023

warn on <code>batched_mul_generic!</code>?

performance

Status: Open.

#478 In FluxML/NNlib.jl;

· chengchingwen opened on Feb 25, 2023

Replace the fallback <code>mul!</code> in <code>batchedmul</code> with <code>Octavian.matmul!</code>?

performance

Status: Open.

#446 In FluxML/NNlib.jl;

· chengchingwen opened on Dec 23, 2022

Investigate why upsampling tests are so slow

ciCI related issues or enhancementsCI related issues or enhancements

help wanted

performance

Status: Open.

#442 In FluxML/NNlib.jl;

· ToucheSir opened on Nov 2, 2022

<code>Conv</code> is 2x slow than pytorch <code>Conv</code> on cpu

performance

Status: Open.

#234 In FluxML/NNlib.jl;

· yiyuezhuo opened on Sep 6, 2020

use vmap for all activations

performance

Status: Open.

#220 In FluxML/NNlib.jl;

· CarloLucibello opened on Jul 11, 2020

Use oneDNN

help wanted

performance

Status: Open.

#74 In FluxML/NNlib.jl;

· vchuravy opened on Oct 22, 2018