llm_repl_7e6

The LLM REPL SEVEN MILLION (SevenMillion) application is a terminal-based chat loop / REPL for interacting with an ollama server.

It sends text to the model, prints the response, maintains a context until manually cleared or end of session, and handles tool calls in model response.

(In case I get distracted, we've only implemented tool calling on the ollama side, up to the point that we have extracted the function name and args out of squishy-space and JSON::Any jail and could reassemble that into a function call in deterministic-space.)

Installation

• Needs Crystal Lang installed. Built with v1.15.1

Usage

Usage: Run the script using crystal llm_repl_7e6.cr.

Commands:

• Type your prompt and press Enter to send it to the Ollama model.
• clear: Clears the conversation history (context).
• exit or quit: Ends the REPL session.

Contributing

1. Fork it (https://github.com/CameronCarroll/llm_repl_7e6/fork)
2. Create your feature branch (git checkout -b my-new-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin my-new-feature)
5. Create a new Pull Request

Contributors

• ieve - creator and maintainer

module LlmRepl7e6 VERSION = "0.1.0"