BlenderLLM: Training Large Language Models for Computer-Aided Design with Self-improvement

TL;DR

BlenderLLM introduces a self-improving framework for training large language models tailored to CAD tasks, significantly enhancing accuracy and functionality in generating CAD scripts through minimal fine-tuning and iterative improvement.

Contribution

The paper presents BlenderLLM, a novel self-improvement approach for training LLMs specifically for CAD, supported by a new dataset and evaluation suite, advancing CAD automation.

Findings

Existing models have limited CAD script accuracy.

Self-improvement markedly improves model performance.

BlenderLLM outperforms baseline models in CAD tasks.

Abstract

The application of Large Language Models (LLMs) in Computer-Aided Design (CAD) remains an underexplored area, despite their remarkable advancements in other domains. In this paper, we present BlenderLLM, a novel framework for training LLMs specifically for CAD tasks leveraging a self-improvement methodology. To support this, we developed a bespoke training dataset, BlendNet, and introduced a comprehensive evaluation suite, CADBench. Our results reveal that existing models demonstrate significant limitations in generating accurate CAD scripts. However, through minimal instruction-based fine-tuning and iterative self-improvement, BlenderLLM significantly surpasses these models in both functionality and accuracy of CAD script generation. This research establishes a strong foundation for the application of LLMs in CAD while demonstrating the transformative potential of self-improving models in advancing CAD automation. We encourage further exploration and adoption of these methodologies to drive innovation in the field. The dataset, model, benchmark, and source code are publicly available at https://github.com/FreedomIntelligence/BlenderLLM