https://github.com/wallarm/nascell-automl
https://github.com/wallarm/nascell-automl
automl machine-learning neural-network
Last synced: about 1 year ago
JSON representation
- Host: GitHub
- URL: https://github.com/wallarm/nascell-automl
- Owner: wallarm
- License: mit
- Created: 2017-12-11T19:53:35.000Z (over 8 years ago)
- Default Branch: master
- Last Pushed: 2021-05-30T08:04:26.000Z (about 5 years ago)
- Last Synced: 2024-03-26T19:07:44.951Z (about 2 years ago)
- Topics: automl, machine-learning, neural-network
- Language: Python
- Size: 10.8 MB
- Stars: 252
- Watchers: 14
- Forks: 79
- Open Issues: 6
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# nascell-automl
This code belongs to the "Simple implementation of Neural Architecture Search with Reinforcement Learning
" blog post.
An original blog post with all the details (step-by-step guide):
https://lab.wallarm.com/the-first-step-by-step-guide-for-implementing-neural-architecture-search-with-reinforcement-99ade71b3d28
# Requirements
- Python 3
- Tensorflow > 1.4
# Training
Print parameters:
```
python3 train.py --help
```
```
optional arguments:
-h, --help show this help message and exit
--max_layers MAX_LAYERS
```
Train:
```
python3 train.py
```
# For evaluate architecture
Print parameters:
```
$ cd experiments/
$ python3 train.py --architecture "61, 24, 60, 5, 57, 55, 59, 3"
```