jarvis73 / iclib

Image Classification Library Using Supervised and Self-Supervised Methods.(PyTorch)

ICLib

Image Classification Library Using Deep Learning Methods. (PyTorch Implementation)

Requirements

1. Create conda environment

conda create -n torch python=3.7
source activate torch
conda install numpy=1.19.1
conda install pytorch=1.6.0 torchvision=0.7.0 cudatoolkit=10.1 -c pytorch 
conda install tqdm pymongo
pip install sacred

2. Setup MongoDB and Omniboard used for recording experiments by Sacred.

Self-Supervised Learning

The part of self-supervised learning is an PyTorch implement of SimCLR.

1. Usage

• Pretrain

python main.py pretrain with pretrain_cifar10

• Finetune

python main.py train_then_eval with train_then_eval_cifar10 ckpt_id=1

2. Results

Model	Resource	CIFAR-10
SimCLR (ResNet18)	5GB	92.43
SimCLR (ResNet50)	18GB	93.25
SimCLR (ResNet50), ImageNet-Pretrain		93.08

Acknowledgement

Thanks to google-research/simclr and sthalles/SimCLR .