Welcome to the official repository for the oneAPI-GenAI-Hackathon-2023 organized by Hack2Skill!
To get started with the oneAPI-GenAI-Hackathon-2023 repository, follow these steps:
- Fork this repository
- Create a folder with your Team Name
- Upload all the code and necessary files in the created folder
- Upload a README.md file in your folder with the below-mentioned information.
- Generate a Pull Request with your Team Name. (Example: submission-XYZ_team)
Team Leader Email - manijb13@gmail.com
This project aims to develop a generative AI system that revolutionizes how customers interact with e-commerce platforms by introducing visual search capabilities leveraging state-of-the-art LLM technologies.
- This system will allow users to upload images of desired products or items, then identify similar products from the product catalog, enhancing the user's shopping experience.
- We develop a generative AI system that enables visual search, allowing customers to find products by uploading images rather than using text-based queries.
- LLM
- Python
- React
- Flask
- Intel Oneapi Developer cloud platform
- For visual search, we’ve encoded query images using the trained encoder and then found similarly encoded representations from our dataset.
- OpenAI's CLIP (Contrastive Language-Image Pretraining) model is used for embedding images.
- Each of these embeddings is then saved to Pinecone Vector Database for semantic search against the query.
- For product discovery, we generate new product images using the trained generator and display them to users based on their preferences.
- An IDC compute instance
- SSH access to the instance
- Miniconda (instructions included for installation)
- SSH into the IDC Compute Instance
- To start, SSH into your IDC compute instance. Replace with your actual instance IP address.
ssh username@<YourInstanceIP>
- Install Miniconda Once logged in, install Miniconda for simplified Python environment management. Run the following:
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh
Follow the on-screen instructions to complete the installation.
- Create and Activate a Conda Environment Create a new Conda environment with Python version 3.10.6:
conda create -n myenv python=3.10.6
conda activate myenv
This will create and activate a new environment named myenv.
- Install Required Modules Install the required modules specified in requirements.txt:
pip install -r requirements.txt
Ensure requirements.txt is present in your current directory.
- Accessing React UI, Flask API, and Port Forwarding To access the React UI, and Flask API, use SSH tunneling for port forwarding. Run the following command on your local machine:
python app.py
npm run dev
ssh -L 8501:localhost:8501 username@<YourInstanceIP>
Then, you can access the React UI, and Flask API by navigating to localhost:8501 and localhost:8502 in your web browser.
- Image to Image Search.
- Text to Image Search.
- Product Discovery (Create new types of products based on user requirements).
- Use generative AI to extract results from multiple E-commerce databases.
- Create an Android/iOS Application.
- Deploy the Application in AWS using EKS (Elastic Kubernetes Service)