Agentive_LLM_CL_Wiki

Head to the 'Wiki' tab to see more infor on the project!

A Tentative Design / Todo List

1. G: Setup GPT2-small for finetuning
2. D: Function to grab random doc from wiki pages (Function S)
3. D: Function to prep file for GPT-2 submission; Read this article on byte-pair word tokenizing.
4. Train on 1/x portion of doc
5. G: Function to eval perplexity
6. G: Eval continue vs. stop for a given txt (Function F)
7. D: function to jump back to either 4 & 5 or 2 (Function S)
8. D: Function to log visited files, perplexity levels