Patches Are All You Need? ?

This repository contains an implementation of ConvMixer for the ICLR 2022 submission “Patches Are All You Need?” by Asher Trockman and Zico Kolter.

? New: Check out this repository for training ConvMixers on CIFAR-10.

Code overview

The most important code is in We trained ConvMixers using the timm framework, which we copied from here.

Update: ConvMixer is now integrated into the timm framework itself. You can see the PR here.

Inside pytorch-image-models, we have made the following modifications. (Though one could look at the diff, we think it is convenient to summarize them here.)

  • Added ConvMixers
    • added timm/models/
    • modified timm/models/
  • Added “OneCycle” LR Schedule
    • added timm/scheduler/
    • modified timm/scheduler/
    • modified timm/scheduler/
    • modified timm/scheduler/
    • modified (added two lines to support this LR schedule)

We are confident that the use of the OneCycle schedule here is not critical, and one could likely just as well
train ConvMixers with the built-in cosine schedule.


We provide some model weights below:

Model Name Kernel Size Patch Size File Size
ConvMixer-1536/20 9 7 207MB
ConvMixer-768/32* 7 7 85MB
ConvMixer-1024/20 9 14 98MB

* Important: ConvMixer-768/32 here uses ReLU instead of GELU, so you would have to change accordingly (we will fix this later).

You can evaluate ConvMixer-1536/20 as follows:

python --model convmixer_1536_20 --b 64 --num-classes 1000 --checkpoint [/path/to/convmixer_1536_20_ks9_p7.pth.tar] [/path/to/ImageNet1k-val]

You should get a 81.37% accuracy.


If you had a node with 10 GPUs, you could train a ConvMixer-1536/20 as follows (these are exactly the settings we used):

sh 10 [/path/to/ImageNet1k] 
    --train-split [your_train_dir] 
    --val-split [your_val_dir] 
    --model convmixer_1536_20 
    -b 64 
    -j 10 
    --opt adamw 
    --epochs 150 
    --sched onecycle 
    --input-size 3 224 224
    --lr 0.01 
    --aa rand-m9-mstd0.5-inc1 
    --cutmix 0.5 
    --mixup 0.5 
    --reprob 0.25 
    --remode pixel 
    --num-classes 1000 
    --warmup-epochs 0 
    --clip-grad 1.0

We also included a ConvMixer-768/32 in timm/models/ (though it is simple to add more ConvMixers). We trained that one with the above settings but with 300 epochs instead of 150 epochs.

Note: If you are training on CIFAR-10 instead of ImageNet-1k, we recommend setting --scale 0.75 1.0 as well, since the default value of 0.08 1.0 does not make sense for 32×32 inputs.

The tweetable version of ConvMixer, which requires from torch.nn import *:

def ConvMixer(h,d,k,p,n):
 S,C,A=Sequential,Conv2d,lambda x:S(x,GELU(),BatchNorm2d(h))
 R=type('',(S,),{'forward':lambda s,x:s[0](x)+x})
 return S(A(C(3,h,p,p)),*[S(R(A(C(h,h,k,groups=h,padding=k//2))),A(C(h,h,1))) for i in range(d)],AdaptiveAvgPool2d(1),Flatten(),Linear(h,n))


View Github