BPMN Model Generation with Instruction Tuned LLM

BPMN-IT: Instruction-Tuned Language Model and Dataset for Business Process Modeling This repository contains the supplementary materials for the paper: Generating Business Process Models with Open Source Large Language Models using Instruction Tuning (link to paper). Overview We introduce an instruction-tuned large language model fine-tuned on a custom dataset for generating and understanding BPMN (Business Process Model and Notation) diagrams. We instruction tune Qwen3-4B model as our base model and compare the results of this model with our prior tuned Gemma2-9b model, other open source models and proprietary LLMs. Resources 🧠 Instruction-tuned Qwen model on Hugging Face 📦 BPMN-IT Dataset on Hugging Face 👤 Hugging Face profile containing other trained models Original Paper for the Source Dataset: Li, X., Ni, L., Li, R., Liu, J. and Zhang, M., 2023, June. MaD: A Dataset for Interview-based BPM in Business Process Management. In 2023 International Joint Conference on Neural Networks (IJCNN) (pp. 1–8). IEEE.