Llama_MindatWorkflow

Fine-tuning code and models for automating geoscience data analysis workflows using Llama 3.1.

Resources

Fine-tuned Model: llama_3.1_instruct_8b_openmindat

Training Dataset:

Synthetic Dataset Generation: alpaca_for_kg

LLaMA 3.1 Instruct 8B OpenMindat

This model is hosted on Ollama and is designed for AI-driven geoscience workflows, optimized for OpenMindat applications.

---

Model Overview

Attribute	Details
Size	4.9 GB
Architecture	llama
Parameters	8.03B
Quantization	Q4_K_M

---

How to Use the Model

Make sure you have Ollama installed. Then run:

ollama run gene21d4/llama_3.1_instruct_8b_openmindat

Citation

If you use this code in your research, please cite our work:

@misc{zhang2024fine,
  title={Fine-Tuning Small and Open {LLMs} to Automate Geoscience Data Analysis Workflows: A Scalable Approach},
  author={Zhang, Jiyin and Li, Wenjia and Que, Xiang and Chen, Weilin and Li, Chenhao and Ma, Xiaogang},
  howpublished={Available at SSRN},
  note={SSRN 5065689},
  year={2024},
  month={December},
  doi={10.2139/ssrn.5065689},
  keywords={Open LLM, Fine tuning, Mindat, Data analytics, Data science}
}

Paper: Fine-Tuning Small and Open LLMs to Automate Geoscience Data Analysis Workflows: A Scalable Approach