Search results

reproduce retrain result

Status: Open.

#18 In megvii-model/SinglePathOneShot;

· zhanglei1172 opened on Sep 12, 2022

Evaluate pretrained model

Status: Open.

#17 In megvii-model/SinglePathOneShot;

· yaolu-zjut opened on Apr 9, 2022

Missing arguments in validate.

Status: Open.

#14 In megvii-model/SinglePathOneShot;

· ghost opened on Mar 7, 2021

how can I constraint FLOPs?

Status: Open.

#13 In megvii-model/SinglePathOneShot;

· cingyoung opened on Jan 28, 2021

What is the purpose of setting flops bins when training SuperNet?

Status: Open.

#11 In megvii-model/SinglePathOneShot;

· jianghaojun opened on Oct 15, 2020

What's the minimum flops-limit in the search?

Status: Open.

#10 In megvii-model/SinglePathOneShot;

· betterhalfwzm opened on Mar 18, 2020

Supernet accuracy

Status: Open.

#8 In megvii-model/SinglePathOneShot;

· xiaoshingshing opened on Feb 17, 2020

Evolution search results

Status: Open.

#6 In megvii-model/SinglePathOneShot;

· ultmaster opened on Dec 6, 2019

Why set affine=False in all nn.BatchNorm2d

Status: Open.

#3 In megvii-model/SinglePathOneShot;

· kaleidoscopical opened on Nov 21, 2019