Instructions to use TIGER-Lab/VisCoder-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TIGER-Lab/VisCoder-7B with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="TIGER-Lab/VisCoder-7B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("TIGER-Lab/VisCoder-7B")
model = AutoModelForCausalLM.from_pretrained("TIGER-Lab/VisCoder-7B", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use TIGER-Lab/VisCoder-7B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TIGER-Lab/VisCoder-7B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TIGER-Lab/VisCoder-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/TIGER-Lab/VisCoder-7B

SGLang

How to use TIGER-Lab/VisCoder-7B with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "TIGER-Lab/VisCoder-7B" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TIGER-Lab/VisCoder-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "TIGER-Lab/VisCoder-7B" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TIGER-Lab/VisCoder-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use TIGER-Lab/VisCoder-7B with Docker Model Runner:
```
docker model run hf.co/TIGER-Lab/VisCoder-7B
```

VisCoder-7B

File size: 3,053 Bytes

---
base_model:
- Qwen/Qwen2.5-Coder-7B-Instruct
datasets:
- TIGER-Lab/VisCode-200K
language:
- en
license: apache-2.0
tags:
- code
library_name: transformers
pipeline_tag: text-generation
---

# VisCoder-7B

[🏠 Project Page](https://tiger-ai-lab.github.io/VisCoder) | [📖 Paper](https://arxiv.org/abs/2506.03930) | [💻 GitHub](https://github.com/TIGER-AI-Lab/VisCoder) | [🤗 VisCode-200K](https://huggingface.co/datasets/TIGER-Lab/VisCode-200K) | [🤗 VisCoder-3B](https://huggingface.co/TIGER-Lab/VisCoder-3B)

**VisCoder-7B** is a large language model fine-tuned for **Python visualization code generation and multi-turn self-correction**. It is trained on **VisCode-200K**, a large-scale instruction-tuning dataset that integrates validated executable code, natural language instructions, and revision supervision from execution feedback.


## 🧠 Model Description

**VisCoder-7B** is trained on **VisCode-200K**, a large-scale instruction-tuning dataset tailored for executable Python visualization tasks. It addresses a core challenge in data analysis: generating Python code that not only executes successfully but also produces **semantically meaningful plots** by aligning **natural language instructions**, **data structures**, and **visual outputs**.

We propose a **self-debug evaluation protocol** that simulates real-world developer workflows. In this setting, models are allowed to revise previously failed generations over multiple rounds with guidance from **execution feedback**.

## 📊 Main Results on PandasPlotBench

We evaluate VisCoder-7B on [**PandasPlotBench**](https://github.com/TIGER-AI-Lab/VisCoder/tree/main/eval), which tests executable visualization code generation across three major libraries. Our benchmark covers both standard generation and **multi-round self-debugging**.

![image/png](https://cdn-uploads.huggingface.co/production/uploads/64de37ee5e192985054be575/ZTicATvYEIVRe4OCj16GV.png)

> VisCoder-7B achieves over **90% execution pass rate** on both **Matplotlib** and **Seaborn** under the self-debug setting, outperforming open-source baselines and approaching GPT-4o performance.


## 📁 Training Details

- **Base model**: Qwen2.5-Coder-7B-Instruct
- **Framework**: [ms-swift](https://github.com/modelscope/swift)
- **Tuning method**: Full-parameter supervised fine-tuning (SFT)
- **Dataset**: [VisCode-200K](https://huggingface.co/datasets/TIGER-Lab/VisCode-200K), which includes:
  - 150K+ validated Python visualization samples with images
  - 45K+ multi-turn correction dialogues with execution feedback

## 📖 Citation

If you use VisCoder-7B or VisCode-200K in your research, please cite:

```bibtex
@article{ni2025viscoder,
  title={VisCoder: Fine-Tuning LLMs for Executable Python Visualization Code Generation},
  author={Ni, Yuansheng and Nie, Ping and Zou, Kai and Yue, Xiang and Chen, Wenhu},
  journal={arXiv preprint arXiv:2506.03930},
  year={2025}
}
```

For evaluation scripts and more information, see our [GitHub repository](https://github.com/TIGER-AI-Lab/VisCoder).