arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.03375cs.AIcs.PL
收藏

BetterV:基于判别性引导的可控 Verilog 生成

BetterV: Controlled Verilog Generation with Discriminative Guidance

The Chinese University of Hong Kong(香港中文大学);Noah’s Ark Lab, Huawei(华为诺亚方舟实验室)

Zehua Pei, Hui-Ling Zhen, Mingxuan Yuan, Yu Huang, Bei Yu

AI总结:

BetterV 通过领域数据微调大语言模型并引入生成式判别器引导,在 VerilogEval 上超越 GPT-4,显著提升 EDA 下游任务性能。

摘要:

由于现代集成电路(ICs)日益增长的复杂性,亟需自动化的电路设计方法。近年来,硬件设计语言生成的研究不断增多,以促进设计流程。本文提出了一种 Verilog 生成框架 BetterV,该框架在经处理的领域特定数据集上微调大语言模型(LLMs),并引入生成式判别器以针对特定设计需求提供引导。Verilog 模块从互联网上收集、过滤和处理,形成干净且丰富的数据集。专门设计了指令微调方法来微调 LLMs,使其理解 Verilog 相关知识。此外,通过数据增强丰富训练集,并利用增强数据在特定下游任务上训练生成式判别器,从而为 LLMs 优化 Verilog 实现提供引导。BetterV 能够生成语法和功能正确的 Verilog,在 VerilogEval 基准上可超越 GPT-4。借助任务特定的生成式判别器,BetterV 在多种电子设计自动化(EDA)下游任务上取得了显著改进,包括综合中的网表节点缩减以及利用布尔可满足性(SAT)求解的验证运行时间缩减。

英文摘要:

Due to the growing complexity of modern Integrated Circuits (ICs), there is a need for automated circuit design methods. Recent years have seen rising research in hardware design language generation to facilitate the design process. In this work, we propose a Verilog generation framework, BetterV, which fine-tunes the large language models (LLMs) on processed domain-specific datasets and incorporates generative discriminators for guidance on particular design demands. The Verilog modules are collected, filtered and processed from internet to form a clean and abundant dataset. Instruct-tuning methods are specially designed to fine-tune the LLMs to understand the knowledge about Verilog. Furthermore, data are augmented to enrich the training set and also used to train a generative discriminator on particular downstream task, which leads a guidance for the LLMs to optimize the Verilog implementation. BetterV has the ability to generate syntactically and functionally correct Verilog, which can outperform GPT-4 on the VerilogEval benchmark. With the help of task-specific generative discriminator, BetterV can achieve remarkable improvement on various electronic design automation (EDA) downstream tasks, including the netlist node reduction for synthesis and verification runtime reduction with Boolean Satisfiability (SAT) solving.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑