Create README.md

Author: yul-nk97

README.md

@@ -0,0 +1,49 @@
+# Shift and Flip Invariant Convolutional Neural Networks for Thermo-Fluid Flow Properties Prediction of Laminar Channel Flows
+
+This repository contains the implementation of a shift and flip invariant convolutional neural network (CNN) designed for predicting thermo-fluid flow properties in laminar channel flows. The code enables robust and precise predictions by leveraging periodic boundary conditions.
+
+---
+
+## Conda Environment Setup
+
+To set up the environment for this project using Conda, follow these steps:
+
+1. Create the Conda environment from the provided `.yml` file:
+
+    ```bash
+    conda env create -f bwuni.yml
+    ```
+
+2. Activate the new environment:
+
+    ```bash
+    conda activate invariant_cnn
+    ```
+
+This environment contains all the required dependencies to run the project code, ensuring compatibility and reproducible experiments.
+
+---
+
+## Getting Started
+
+### Training and Testing the Model
+
+To train and test the periodic invariant CNN, you can execute the following command:
+
+```bash
+python3 iv_cnn_periodic_file.py --config train_config.yaml --testing False
+```
+
+To run the model for testing purposes only, use:
+
+```bash
+python3 iv_cnn_periodic_file.py --config train_config.yaml --testing True
+```
+
+### Configuration File
+
+The model's settings (such as hyperparameters and data paths) are defined in the YAML configuration file `train_config.yaml`. Adjust this file as necessary to fit your specific dataset and experiment settings.
+
+---
+
+