This project is a Blog Generator web application built using Streamlit and LangChain. The app uses the Ollama LLM (Large Language Model) to generate professional, informative blogs based on the user's input.
- Generate blogs for different target audiences such as:
- Common people
- Data Scientists
- Software Engineers
- Teachers
- Body Builders
- Customizable blog topics and word counts.
- Professional and informative tone with no invalid references or links.
- Simple and clean UI built with Streamlit.
- Fast response time with real-time blog generation.
- Enter a blog topic.
- Specify the number of words.
- Choose the target audience from the options.
- Click Generate to create a blog in seconds.
- The app shows the blog along with the time taken for generation.
- Streamlit for building the user interface.
- LangChain for managing language model prompts.
- Ollama LLM (
gemma2:2bmodel) for blog generation. - Python for backend logic and integration.
-
Clone this repository:
git clone https://github.com/your-username/blog-generator.git cd blog-generator -
Install dependencies:
pip install -r requirements.txt
-
Download Ollama setup from Ollama's Github, install it and then download a model (if not already installed). I used the gemma2:2b model for this application. The model size is 1.6GB.
If you've downloaded any other model, make sure to change the line no: 18 of the main.py file to: llm = load_ollama_llm(YOUR_DOWNLOADED_MODEL_NAME_HERE)
- Run the app:
streamlit run app.py
- app.py: Main application logic and Streamlit UI.
- src/module.py: Contains the
generate_blogfunction that handles the blog generation process. - requirements.txt: List of dependencies to run the project.
- Allow users to select different writing styles.
- Improve customization for blog tone and structure.
- Add more audience options.
This project is licensed under the MIT License.