This project was created as part of the implementation of the competition task on the 🔗Kaggle platform.
The task was to present an interesting application of Google Gemini's LLM model using its long context window, which opens up new possibilities in the world of data science.
The goal was to demonstrate the model's potential for processing large amounts of data and analyzing complex sets of information. With a long context window, ✨Gemini LLM allows users to work on extensive datasets in a single step, significantly speeding up the analysis process and enabling new applications in data science.
Your future Kaggle notebook work center consists of two key sections:
- 🎉 About - a current page with a greeting and summary instructions on how the site works
- 📓 Notebook Creator - a page where you can generate a whole notebook based on a selected dataset
- 💬 AI Chat - the place where you can talk to Gemini about your chosen notebook project
- ⚙️ Settings - before you start work, choose the model and parameters
- 🗝️ Kaggle API Key - it is needed in order to be able to extract the necessary data, which are the selected notebooks. The entire process of obtaining the key is very simple and is well described in the 🔗Kaggle repository
- 🗝️ Gemini API Key - the key to Gemini API is necessary in order to use selected models and functions of Google Gemini and for the proper functioning of this project. The process of obtaining the access key is possible via 🔗AI Studio.
Download a repository
> git clone https://github.com/mateo252/AI-Kaggle-Assistant.git
> cd AI-Kaggle-Assistant
Create a virtual environment and install requirements (require Python <= 3.12)
> python -m venv venv
> venv\Scripts\activate
(venv) > pip install -r requirements.txt
Create .env file and add Gemini API Key like GEMINI_API=...
Finally run a project
(venv) > cd src
(venv) > streamlit run main.py
Apache License 2.0