pytorch-densenet

A pytorch implementation of DenseNet (https://arxiv.org/abs/1608.06993)

Requirements

• Python 2.7
• Pytorch (newest version)

Usage

This implementation currently supports training on CIFAR-10 dataset.

Simply run:

python train.py --help

which will print:

usage: train.py [-h] [--gpu GPU] [--dataset DATASET] [--batch_size BATCH_SIZE]
                [--epoch EPOCH] [--seed SEED] [--checkpoint CHECKPOINT]
                [--resume_model RESUME_MODEL] [--lr LR] [--momentum MOMENTUM]

Densenet for Classification

optional arguments:
  -h, --help            show this help message and exit
  --gpu GPU
  --dataset DATASET
  --batch_size BATCH_SIZE
  --epoch EPOCH
  --seed SEED
  --checkpoint CHECKPOINT
  --resume_model RESUME_MODEL
  --lr LR
  --momentum MOMENTUM