This is a lightweight interactive app built with Streamlit and Google's Gemini 1.5 Flash model. It allows you to type in prompts, experiment with model parameters like temperature and top_p, and instantly see the AI's response. Ideal for learning, prototyping, and exploring how large language models behave.
- 🔐 Secure Google API key entry
- 💬 Dynamic prompt input
- 🎛️ Adjustable generation parameters:
- Temperature
- Top-p (nucleus sampling)
- Top-k
- Max output tokens
- 📜 Prompt & response history (scrollable)
- 🧪 Built on Gemini 1.5 Flash (
gemini-1.5-flash-latest) - ⚡ Runs entirely on your local machine
- Visit: https://aistudio.google.com/app/apikey
- Sign in with your Google account
- Click “Create API key”
- Copy the key (it starts with
AIza...) - You’ll paste this key into the app when it runs
git clone https://github.com/Bayero-abdul/Few-shot_learning_app.git
cd few-shot_learning_apppython -m venv venv
source venv/bin/activate # macOS/Linux
venv\Scripts\activate # Windowspip install streamlit google-generativeaistreamlit run app.pyThen go to http://localhost:8501 in your browser.
| File | Description |
|---|---|
app.py |
The main Streamlit application |
README.md |
This file |
Gemini is Google DeepMind’s family of large language models. This app uses the Gemini 1.5 Flash version — optimized for fast, cost-effective inference. Great for chat, summarization, coding, and more.
Created by Bayero Abdul
This tool was built as part of the MLC Nigeria study group project to learn how to prototype with large language models and explore prompt engineering.
MIT License