ChipLingo: A Systematic Training Framework for Large Language Models in EDA

TL;DR

ChipLingo is a comprehensive training framework that adapts large language models specifically for electronic design automation, improving their domain expertise and retrieval capabilities.

Contribution

The paper introduces a systematic pipeline for domain-specific LLM training in EDA, including data curation, domain-adaptive pretraining, and instruction alignment, along with an internal benchmark.

Findings

ChipLingo-8B achieves 59.7% accuracy on EDA-Bench.

ChipLingo-32B reaches 70.02%, nearing commercial models.

QA augmentation and explicit RAG training improve domain performance and retrieval utilization.

Abstract

With the rapid advancement of semiconductor technology, Electronic Design Automation (EDA) has become an increasingly knowledge-intensive and document-driven engineering domain. Although large language models (LLMs) have shown strong general capabilities, applying them directly to EDA remains challenging due to limited domain expertise, cross-tool knowledge confusion, and degraded retrieval-augmented generation (RAG) performance after domain training. To address these issues, this paper presents ChipLingo, a systematic training pipeline for domain-adapted LLMs tailored to EDA scenarios. ChipLingo consists of three stages: domain corpus construction with multi-source data curation and QA augmentation, domain-adaptive pretraining with comparisons of different parameter training strategies, and instruction alignment with RAG scenario training under diverse retrieval conditions. We also curate an internal benchmark, EDA-Bench, covering representative EDA tool scenarios, with plans for public release. Experiments show that ChipLingo-8B achieves 59.7% accuracy on EDA-Bench, outperforming the same-scale base model and some larger general-purpose models. ChipLingo-32B reaches 70.02%, approaching leading closed-source commercial models. Further analysis shows that QA augmentation improves domain performance, Partial FT offers a better balance between adaptation and general capability retention than LoRA, and explicit RAG scenario training mitigates the decline in retrieval utilization after domain training. These results demonstrate the practical value of systematic domain training for knowledge-intensive EDA tasks and provide a foundation for future EDA agents and external-knowledge-driven systems.