https://github.com/mathurutkarsh/keras-tuner
Keras Tuner for optimization
https://github.com/mathurutkarsh/keras-tuner
keras python tuner
Last synced: 2 months ago
JSON representation
Keras Tuner for optimization
- Host: GitHub
- URL: https://github.com/mathurutkarsh/keras-tuner
- Owner: MathurUtkarsh
- Created: 2022-10-05T02:52:53.000Z (about 3 years ago)
- Default Branch: main
- Last Pushed: 2023-01-26T05:58:43.000Z (over 2 years ago)
- Last Synced: 2025-06-23T04:35:09.216Z (4 months ago)
- Topics: keras, python, tuner
- Language: Jupyter Notebook
- Homepage:
- Size: 12.7 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Keras Tuner :)
## Hyper Parameter tuning using Keras-tuner
### Install the project
- Clone the repo
- Create a virtualenv and activate it:
```
conda create keras_tuner python=3.7
conda activate keras_tuner
```
- Install the requirements:
```
pip install requirements.txt
```
### Results
Tasks duration was measured on an RTX 2080 GPU
| Tuner | Search time | Best accuracy (%) |
|-----------------------|---------------|-------------------|
| Worst Baseline | 20min | 63.1 |
| Default Baseline | 20min | 74.5 |
| Random Search | 10h 59min | 76.8 |
| Hyperband | 10h 0min | 75.1 |
Here, the worst baseline is the worst accuracy obtained by a set of hyperparameters
during random search.
The default baseline is obtained by setting all hyperparameters to their default value.
### Run the baseline
```
python baseline.py
```
### Run the comparison
Available tuners :
- Random Search
- Hyperband
```
python tuner_comparison.py
```