# Clone the repository
                  git clone https://github.com/yourgithub/pocket-racer.git
                  
                  # Navigate to the project directory
                  cd pocket-racer
                  
                  # Install required dependencies
                  pip install -r requirements.txt
                                  

Data Processing

To use the data_processor.py script for preparing and processing your datasets, run the following command:

python data_processor.py --n_stacked <number_of_stacked_images> --img_path <path_to_images> --csv_path <path_to_csv> --w <width> --h <height> --d <depth> --concatenate <concatenate_flag> --prediction_mode <mode>

Modify the parameters according to your dataset's requirements:

1. n_stacked: Number of images to stack for input data.
2. img_path: Path to the directory containing image files.
3. csv_path: Path to the CSV file containing metadata or labels.
4. w, h, d: Width, height, and depth of the processed images.
5. concatenate: Boolean flag to determine whether to concatenate data.
6. prediction_mode: Specifies the mode of prediction, e.g., 'linear' or 'categorical'.

python train_script.py --epochs 50 --batch_size 1000

Features

Our platform offers the following key features:

1. Model Training: Train vision transformers tailored for steering prediction.
2. Performance Metrics: Evaluate model performance with metrics such as MAE, RMSE, and R².
3. Customizable Training Options: Configure training epochs, batch size, and image preprocessing through command-line arguments.