Developed and integrated the MMLU benchmark and a baseline experiment pipeline into the microsoft/eureka-ml-insights repository, enabling comprehensive end-to-end model evaluation on the MMLU dataset. Leveraging Python and expertise in data processing and machine learning, the work introduced reusable utilities for handling MMLU data and established a reproducible configuration for running experiments. This addition allows for consistent comparison of model performance across a wide range of subjects, streamlining the benchmarking process within the repository. All changes were tracked and documented for transparency, reflecting a focused approach to enhancing model evaluation workflows and supporting ongoing research and development in machine learning benchmarking.