WebAI

This web application features a range of AI functionalities, including data loading, processing, training, and prediction. Originally developed for estimating subsurface CO2 flow simulations, the package is currently being expanded to accommodate a variety of applications.

Function API can be found here.

Usage

---

1. Clone the repository:

git clone https://github.com/acse-efk23/webai.git

2. Change directory:

cd webai

3. Create an environment (-m means run as a module):

python -m venv .venv

4. Activate the environment:

.venv\Scripts\activate

5. Install the requirements:

pip install -r requirements.txt

6. Install the deepdown package:

pip install .

7. Install the web plugins package:

cd web_plugins
pip install .

8. Select the browser to view the web app:

cd ..
webviz preferences --browser chrome

9. Run the application (viewed in local host):

webviz build configuration.yaml

Web Application

---

Deep Learning Architecture

---

Comparison with Physics-Based Simulation

---

Acknowledgements

---

The web application is built through an open source Python package developed by Equinor which can be found here.