Comparing Llama-2 and GPT-3 LLMs for HPC kernels generation
Oak Ridge National Laboratory, Oak Ridge, TN, 37830, USA
arXiv:2309.07103 [cs.SE], (12 Sep 2023)
@misc{valerolara2023comparing,
title={Comparing Llama-2 and GPT-3 LLMs for HPC kernels generation},
author={Pedro Valero-Lara and Alexis Huante and Mustafa Al Lail and William F. Godoy and Keita Teranishi and Prasanna Balaprakash and Jeffrey S. Vetter},
year={2023},
eprint={2309.07103},
archivePrefix={arXiv},
primaryClass={cs.SE}
}
We evaluate the use of the open-source Llama-2 model for generating well-known, high-performance computing kernels (e.g., AXPY, GEMV, GEMM) on different parallel programming models and languages (e.g., C++: OpenMP, OpenMP Offload, OpenACC, CUDA, HIP; Fortran: OpenMP, OpenMP Offload, OpenACC; Python: numpy, Numba, pyCUDA, cuPy; and Julia: Threads, CUDA.jl, AMDGPU.jl). We built upon our previous work that is based on the OpenAI Codex, which is a descendant of GPT-3, to generate similar kernels with simple prompts via GitHub Copilot. Our goal is to compare the accuracy of Llama-2 and our original GPT-3 baseline by using a similar metric. Llama-2 has a simplified model that shows competitive or even superior accuracy. We also report on the differences between these foundational large language models as generative AI continues to redefine human-computer interactions. Overall, Copilot generates codes that are more reliable but less optimized, whereas codes generated by Llama-2 are less reliable but more optimized when correct.
September 17, 2023 by hgpu
