Llama 3 Context-Obedient Models in a 3 x MoE configuration. The models experts are split into Understanding and summarizing input, following the format provided, and outputting only context relevant answers as the three experts involved

Update README.md

initial commit