A stable and functional version of an automated writing script that utilizes LLaMA 3.1 via LM Studio. This script generates structured HTML articles from keywords provided in a TXT file. It's designed for SEO optimization and use on platforms like WordPress, though it is highly versatile.
- Generates HTML articles with a title, introduction, body sections 1, 2, and 3, and a conclusion.
- Utilizes section-specific prompts to enhance content quality.
- Advanced CSV handling that avoids overwriting headers or duplicating data.
- Each generated article is logged as a new line in the
articulos.csvfile, facilitating tracking and management. - Runs 100% locally using LM Studio.
-
Install LM Studio:
- Download LM Studio from https://lmstudio.ai/
- Install by following the instructions for your operating system.
-
Set up LLaMA 3.1:
- Open LM Studio.
- Go to the "Browse" tab.
- Search for "llama 3.1 8b instruct" and download the Meta-Llama-3.1-8B-Instruct model.
- This process may take some time depending on your internet connection.
-
Run the LM Studio Server:
- Once the model is downloaded, select it from your local library.
- Click on "Start Server" (or the equivalent in your LM Studio version, often under a "Local Server" tab).
- Configure the server to listen on
http://localhost:1234/v1. - Start the server.
Clone the repository:
git clone https://github.com/R4F405/Redactor-Llama3.1.gitCreate a .env file based on the example:
# On Windows
copy .env.example .env
# On Linux/Mac
cp .env.example .envInstall dependencies:
pip install -r requirements.txt-
Prepare a TXT file named
tematica.txtwhere each line will be the basis for an article:artificial intelligence in education sustainable air conditioning domestic robots in 2025
-
Ensure LM Studio is running with the LLaMA 3.1 model loaded and the server active.
-
Run the main script:
python GeneradorArticulos.py
-
The articles will be logged in an
articulos.csvfile.
Redactor-Llama3.1/
├── GeneradorArticulos.py # Main script
├── utils.py # Helper functions (LLM connection, CSV saving, etc.)
├── tematica.txt # File with keywords to process
├── articulos.csv # Log of generated results
├── .env.example # Example of required environment variables
├── requirements.txt # Project dependencies
└── README.md # This file
The .env file should contain:
LM_STUDIO_API_URL=http://localhost:1234/v1
You can adjust this URL if you have configured LM Studio to use a different port.
<h1>The Future of Artificial Intelligence in Education</h1>
<p><b>Introduction:</b> Artificial intelligence is transforming teaching methods...</p>
<h2>Current AI Applications in Education</h2>
<p>From virtual tutors to student performance analysis, AI systems are...</p>
<p><b>Conclusion:</b> The effective implementation of AI in education can make a significant difference...</p>- Connection Error: Ensure LM Studio is running and the API server is active at the configured URL.
- Generation Errors: Verify that the Meta-Llama-3.1-8B-Instruct model is correctly loaded in LM Studio.
- Slow Responses: Adjust LM Studio settings according to your system's resources.
This project is distributed under the GPL-3.0 license. See the LICENSE file for more details.
Have suggestions, improvements, or want to collaborate? Feel free to open an issue or get in touch!
- Larger models like Llama-3.1-70B will require more system resources.
- You can modify the prompts in
GeneradorArticulos.pyto customize the article style. - The generated HTML format is optimized for WordPress but can be used on any system.
- Basic Graphical User Interface (GUI)
- Direct export to WordPress via API
- Multi-language support
- Improved control over content tone and style