https://github.com/jmisilo/classification-playground
https://github.com/jmisilo/classification-playground
Last synced: about 1 year ago
JSON representation
- Host: GitHub
- URL: https://github.com/jmisilo/classification-playground
- Owner: jmisilo
- Created: 2022-09-09T12:53:16.000Z (almost 4 years ago)
- Default Branch: main
- Last Pushed: 2022-09-11T08:18:43.000Z (almost 4 years ago)
- Last Synced: 2025-02-09T20:36:10.445Z (over 1 year ago)
- Language: Jupyter Notebook
- Size: 363 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: readme.md
Awesome Lists containing this project
README
## Classification Playground
The goal of the project was to implement a very simple model for the classification of datasets from sklearn. [W&B](wandb.ai) was used to optimize hyperparameters. The model has been implemented using [PyTorch](https://pytorch.org/). I used datasets generated with `make_blobs` and `make_moons` methods. Both were trained at the same time, using only my Notebook with CPU. Times:
- **Moons** - 12min 28s
- **Blobs** - 11min 56s
Models' max accuracy on the validation sets:
- **Moons** - 99.33%
- **Blobs** - 91.33%
Detailed results of training process:
- [**Moons runs**](https://wandb.ai/jmisilo/classification-playground-moons-sweep?workspace=user-jmisilo)
- [**Blobs runs**](https://wandb.ai/jmisilo/classification-playground-blobs-sweep?workspace=user-jmisilo)