Our capstone project, a Python pipeline CLI library, offers a streamlined solution for generating captions and a general overview for images. By simply providing URL(s) or a directory containing images, our library delivers insightful captions. Embedded within are pre-trained models for image captioning and optical character recognition, seamlessly integrated with a language generation model (LLM) for cohesive output. Tailored to meet the requirements of our client, Ovative Group, this tool serves as an in-house asset for efficiently captioning advertisements, offering both conciseness and detail in its captions.
To install the required libraries with their specific versions from the requirements.txt file, run the following command:
pip install -r requirements.txt
You can run the project by running the command:
python -m runner