https://github.com/human-analysis/neural-architecture-transfer
Neural Architecture Transfer (Arxiv'20), PyTorch Implementation
https://github.com/human-analysis/neural-architecture-transfer
Last synced: about 2 months ago
JSON representation
Neural Architecture Transfer (Arxiv'20), PyTorch Implementation
- Host: GitHub
- URL: https://github.com/human-analysis/neural-architecture-transfer
- Owner: human-analysis
- Created: 2020-05-01T01:31:52.000Z (about 5 years ago)
- Default Branch: master
- Last Pushed: 2020-05-15T18:42:30.000Z (about 5 years ago)
- Last Synced: 2024-08-03T09:03:23.063Z (11 months ago)
- Language: Python
- Homepage: http://hal.cse.msu.edu/papers/neural-architecture-transfer/
- Size: 4.18 MB
- Stars: 150
- Watchers: 12
- Forks: 21
- Open Issues: 6
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- StarryDivineSky - human-analysis/neural-architecture-transfer
README
# Neural Architecture Transfer
[](https://paperswithcode.com/sota/neural-architecture-search-on-imagenet?p=neural-architecture-transfer)
[](https://paperswithcode.com/sota/architecture-search-on-cifar-10-image?p=neural-architecture-transfer)
[](https://paperswithcode.com/sota/neural-architecture-search-on-cifar-100-1?p=neural-architecture-transfer)
Code accompanying the paper.
> [Neural Architecture Transfer](https://arxiv.org/abs/2005.05859)
>
> [Zhichao Lu](https://www.zhichaolu.com), [Gautam Sreekumar](http://hal.cse.msu.edu/team/gautam-sreekumar/), [Erik Goodman](https://www.egr.msu.edu/~goodman/), [Wolfgang Banzhaf](http://www.cse.msu.edu/~banzhafw/), [Kalyanmoy Deb](https://www.egr.msu.edu/~kdeb/), and [Vishnu Boddeti](http://hal.cse.msu.edu/team/vishnu-boddeti/)
>
> *arXiv:2005.05859*
## Requirements
```
Python >= 3.7.x, PyTorch >= 1.4.0, timm >= 0.1.18
```## ImageNet Classification
``` shell
python evaluator.py --data /path/to/dataset --model subnets/imagenet/NAT-{M1,M2,M3,M4}/net.config
```## Architecture Transfer
``` shell
python evaluator.py \
--data /path/to/dataset \
--dataset {aircraft,cars,cifar10,cifar100,cinic10,dtd,flowers102,food101,pets,stl10} \
--model subnets/{dataset}/net-img@{xxx}-flops@{xxx}-top1@{xx.x}/net.config
```## Scalability to Objectives
#### Tri-Objective

#### 12-Objective
#### Architecture Visualization
## Acknowledgement
Codes are modified from [OnceForAll](https://github.com/mit-han-lab/once-for-all) and [pytorch-image-models](https://github.com/rwightman/pytorch-image-models)