File size: 1,316 Bytes
b34f680
 
 
 
426d6ed
8318a67
 
 
 
 
 
426d6ed
 
 
 
 
 
 
b34f680
 
05c93ca
b34f680
8318a67
b34f680
095b8f8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
---
license: apache-2.0
base_model:
- Qwen/Qwen2.5-7B-Instruct
pipeline_tag: text-to-3d
datasets:
- FreedomIntelligence/BlendNet
metrics:
- code_eval
tags:
- code
- render
- CAD
- 3D
- Modeling
- LLM
- bpy
- Blender
---

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

**BlenderLLM** is built using **Qwen2.5-Coder-7B-Instruct** as the base model. It has been fine-tuned on the **BlendNet** training dataset and further optimized through **Self-improvement** techniques to achieve the best performance.  

For more details, please visit our [GitHub repository](https://github.com/FreedomIntelligence/BlenderLLM) or refer to our [arXiv paper](https://www.arxiv.org/abs/2412.14203).

## 📖 Citation
```angular2
@misc{du2024blenderllmtraininglargelanguage,
      title={BlenderLLM: Training Large Language Models for Computer-Aided Design with Self-improvement}, 
      author={Yuhao Du and Shunian Chen and Wenbo Zan and Peizhao Li and Mingxuan Wang and Dingjie Song and Bo Li and Yan Hu and Benyou Wang},
      year={2024},
      eprint={2412.14203},
      archivePrefix={arXiv},
      primaryClass={cs.HC},
      url={https://arxiv.org/abs/2412.14203}, 
}
```

We are from the School of Data Science (SDS), the Chinese University of Hong Kong, Shenzhen (CUHKSZ).