Energy Considerations of Large Language Model Inference and Efficiency Optimizations | Read Paper on Bytez