This is a preliminar version of a small trained language model to follow orders as an assistant to test meta-prompts as a language model. The model was trained using 8,192 as context window.
