HLS-Eval
收藏资源简介:
HLS-Eval是一个为评估大型语言模型在高级综合设计任务中的性能而创建的基准和框架。该数据集由94个独特的设计组成,来源于社区现有的高级综合基准和新型来源。通过半自动化流程,每个案例都被调整为'LLM准备就绪',并补充了自然语言描述和相应的测试台。HLS-Eval不仅提供了设计基准,还提供了一个自动化、并行评估本地和托管LLM的框架,支持各种高级综合设计任务。
HLS-Eval is a benchmark and framework developed for evaluating the performance of large language models (LLMs) in high-level synthesis (HLS) design tasks. This dataset consists of 94 unique designs sourced from both existing community high-level synthesis benchmarks and novel sources. Through a semi-automated workflow, each case has been adjusted to be "LLM-ready", and supplemented with natural language descriptions and corresponding testbenches. HLS-Eval not only provides design benchmarks, but also offers a framework for automated, parallel evaluation of both local and hosted LLMs, supporting a wide range of high-level synthesis design tasks.

- 1HLS-Eval: A Benchmark and Framework for Evaluating LLMs on High-Level Synthesis Design Tasks乔治亚理工学院 · 2025年



