Command line tool to summarize text files using AI with a local Ollama instance. The project is mainly developed with AI itself using Visual Studio Code, the Continue.dev extension, local Ollama instance and Python. So it's more a playground for developing with AI assistant (as agent and chat), using AI tools and learning about AI itself.

• Summarize text files using AI
  • --model MODEL argument to select model (default llama3.1:8b)
  • --file FILE argument to select file
  • --text "a text" argument to provide a text
  • --question QUESTION argument to select question (default Summarize the following text:)
  • --chunked argument to chunk the input file/text based on (currently hardcoded) chunk size
• List all available models with --list-models/-lsm
• Set Ollama host with optional --host HOST argument. Default is http://127.0.0.1:11434

To summarize a text file using the default settings, you can run:

• python src/main.py --file example/python-ai.md
• python src/main.py --file example/python-ai.md --question "Count how many times AI is mentioned in this text:"
• python src/main.py --file example/python-ai.md --model qwen2.5
• python src/main.py --text "strawberry" --question "How many letter 'R' are in this text:"

List all available models:

• python src/main.py --list-models

1. Install Python (3.8 or higher)
2. Clone this repository
3. Run pip install -r requirements.txt
4. Run python src/main.py

• Ollama installed and running locally
• Optional: Default model is llama3.1:8b. Otherwise it's necessary to specifiy the model with the --model argument.

Feel free to contribute via pull requests.

The project is licensed by the Apache 2 license.