https://github.com/soham-shee/loadlens
A Load Forecasting Prediction System with Frontend
https://github.com/soham-shee/loadlens
forecasting-models forecasting-time-series gru gru-model load-forecasting
Last synced: about 2 months ago
JSON representation
A Load Forecasting Prediction System with Frontend
- Host: GitHub
- URL: https://github.com/soham-shee/loadlens
- Owner: soham-shee
- License: mit
- Created: 2024-07-11T21:34:44.000Z (11 months ago)
- Default Branch: main
- Last Pushed: 2025-02-09T19:55:35.000Z (4 months ago)
- Last Synced: 2025-03-26T09:51:10.647Z (2 months ago)
- Topics: forecasting-models, forecasting-time-series, gru, gru-model, load-forecasting
- Language: Jupyter Notebook
- Homepage: https://load-lens.streamlit.app
- Size: 5.56 MB
- Stars: 1
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Load Lens 😎
[](https://www.kaggle.com/code/sohamshee/gru-model-load-forecasting)\
This ML model is designed for load forecasting using Gated Recurrent Units (GRU). This user-friendly
app empowers users to input their past values data, specify the number of epochs, and set the
batch size for training a GRU-based model.
By leveraging the GRU architecture, the app efficiently captures temporal dependencies in the
data, making it ideal for accurate load forecasting. Once the model is trained, users can easily
download the trained model for future use, ensuring they have a reliable tool at their fingertips
for predicting load demand.
In addition to model training, this app offers a robust suite of features to enhance usability
and flexibility. Users can upload a previously trained model alongside a CSV file to retrain
the model, accommodating new data and improving prediction accuracy. This iterative approach
ensures the model remains up-to-date with the latest trends and patterns. Furthermore, the app
allows users to upload an existing model to forecast future values based on specified inputs,
providing quick and precise predictions. Whether you are training a new model, retraining with
additional data, or forecasting future values, this app offers a comprehensive solution for load
forecasting needs.
## Run Locally
Clone the project
```bash
git clone https://github.com/soham-shee/LoadLens.git
```
Install dependencies
```bash
pip install -r 'requirements.txt'
```
Start the server
```bash
streamlit run App.py
```
To directly start it (Alternative Method)
```bash
./run_app.sh
```
## Demo
https://load-lens.streamlit.app/
## Acknowledgements
- [Load Forecast Dataset (Panama Case Study)](https://www.kaggle.com/datasets/saurabhshahane/electricity-load-forecasting)