Skip to content
Densely Connected Search Space for More Flexible Neural Architecture Search
Branch: master
Clone or download
jiemin.fang
jiemin.fang modified
Latest commit 7245ecb Jun 25, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
configs The first commit. Jun 23, 2019
data The first commit. Jun 23, 2019
dataset update Jun 23, 2019
tools update Jun 23, 2019
.gitignore Initial commit Jun 21, 2019
LICENSE Initial commit Jun 21, 2019
README.md modified Jun 25, 2019
model_derived.py The first commit. Jun 23, 2019
operations.py The first commit. Jun 23, 2019
validation.py update Jun 23, 2019

README.md

DenseNAS

The evaluation code of the paper Densely Connected Search Space for More Flexible Neural Architecture Search

We propose a novel differentiable NAS method which can search for the width and the spatial resolution of each block simultaneously. We achieve this by constructing a densely connected search space and name our method as DenseNAS. Blocks with different width and spatial resolution combinations are densely connected to each other. The best path in the super network is selected by optimizing the transition probabilities between blocks. As a result the overall depth distribution of the network is optimized globally in a graceful manner.

Requirements

  • pytorch 1.0.1
  • python 3.6+

Results

DenseNAS obtains an architecture with 75.9% top-1 accuracy on ImageNet and the latency is as low as 24.3ms on a single TITAN-XP. The total search time is merely 23 hours on 4 GPUs. For the GPU latency, we measure all the models with the same setup (on one TITAN-XP with batch size of 32).

Our results on ImageNet are shown bellow.

Our pretrained models can be downloaded in the following:

Evaluate

  1. Download the related files of the pretrained model and put net_config and weights.pt into the model_path
  2. python validation.py --data_path 'The path to ImageNet data' --load_path 'The path you put the pretrained model'
You can’t perform that action at this time.