Skip to content

Latest commit

 

History

History

gMLP

README.md

Pay Attention to MLPs, arxiv

PaddlePaddle training/validation code and pretrained models for gMLP.

The 3rd party pytorch implementation is here.

This implementation is developed by PaddleViT.

drawing

gMLP Model Overview

Update

  • Update (2022-03-30): Code is refactored.
  • Update (2021-09-27): Model FLOPs and # params are uploaded.
  • Update (2021-08-11): Code is released and ported weights are uploaded.

Models Zoo

Model Acc@1 Acc@5 #Params FLOPs Image Size Crop_pct Interpolation Link
gmlp_s16_224 79.64 94.63 19.4M 4.5G 224 0.875 bicubic google/baidu

*The results are evaluated on ImageNet2012 validation set.

Note: gMLP weights are ported from timm

Data Preparation

ImageNet2012 dataset is used in the following file structure:

│imagenet/
├──train_list.txt
├──val_list.txt
├──train/
│  ├── n01440764
│  │   ├── n01440764_10026.JPEG
│  │   ├── n01440764_10027.JPEG
│  │   ├── ......
│  ├── ......
├──val/
│  ├── n01440764
│  │   ├── ILSVRC2012_val_00000293.JPEG
│  │   ├── ILSVRC2012_val_00002138.JPEG
│  │   ├── ......
│  ├── ......
  • train_list.txt: list of relative paths and labels of training images. You can download it from: google/baidu
  • val_list.txt: list of relative paths and labels of validation images. You can download it from: google/baidu

Usage

To use the model with pretrained weights, download the .pdparam weight file and change related file paths in the following python scripts. The model config files are located in ./configs/.

For example, assume weight file is downloaded in ./gmlp_s16_224.pdparams, to use the gmlp_s16_224 model in python:

from config import get_config
from gmlp import build_gmlp as build_model
# config files in ./configs/
config = get_config('./configs/gmlp_s16_224.yaml')
# build model
model = build_model(config)
# load pretrained weights
model_state_dict = paddle.load('./gmlp_s16_224.pdparams')
model.set_state_dict(model_state_dict)

Evaluation

To evaluate model performance on ImageNet2012, run the following script using command line:

sh run_eval_multi.sh

or

CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 \
python main_multi_gpu.py \
-cfg='./configs/gmlp_s16_224.yaml' \
-dataset='imagenet2012' \
-batch_size=256 \
-data_path='/dataset/imagenet' \
-eval \
-pretrained='./gmlp_s16_224.pdparams' \
-amp

Note: if you have only 1 GPU, change device number to CUDA_VISIBLE_DEVICES=0 would run the evaluation on single GPU.

Training

To train the model on ImageNet2012, run the following script using command line:

sh run_train_multi.sh

or

CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 \
python main_multi_gpu.py \
-cfg='./configs/gmlp_s16_224.yaml' \
-dataset='imagenet2012' \
-batch_size=256 \
-data_path='/dataset/imagenet' \
-amp

Note: it is highly recommanded to run the training using multiple GPUs / multi-node GPUs.

Reference

@article{liu2021pay,
  title={Pay attention to MLPs},
  author={Liu, Hanxiao and Dai, Zihang and So, David and Le, Quoc},
  journal={Advances in Neural Information Processing Systems},
  volume={34},
  year={2021}
}