Instructions to use luzimu/FullStack-Learn-LM-30B-A3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use luzimu/FullStack-Learn-LM-30B-A3B with Transformers:

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

pipe = pipeline("text-generation", model="luzimu/FullStack-Learn-LM-30B-A3B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("luzimu/FullStack-Learn-LM-30B-A3B")
model = AutoModelForCausalLM.from_pretrained("luzimu/FullStack-Learn-LM-30B-A3B")
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]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use luzimu/FullStack-Learn-LM-30B-A3B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "luzimu/FullStack-Learn-LM-30B-A3B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "luzimu/FullStack-Learn-LM-30B-A3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/luzimu/FullStack-Learn-LM-30B-A3B

SGLang

How to use luzimu/FullStack-Learn-LM-30B-A3B 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 "luzimu/FullStack-Learn-LM-30B-A3B" \
    --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": "luzimu/FullStack-Learn-LM-30B-A3B",
		"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 "luzimu/FullStack-Learn-LM-30B-A3B" \
        --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": "luzimu/FullStack-Learn-LM-30B-A3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use luzimu/FullStack-Learn-LM-30B-A3B with Docker Model Runner:
```
docker model run hf.co/luzimu/FullStack-Learn-LM-30B-A3B
```

nielsr HF Staff commited on Feb 4

Commit

79e528c

verified ·

1 Parent(s): b553002

Add library_name and citation, fix quick start links

Browse files

Hi! I'm Niels from the Hugging Face community science team.

I'm opening this pull request to improve the model card for **FullStack-Learn-LM-30B-A3B**:
- Added `library_name: transformers` to the metadata. This is based on the `config.json` which shows compatibility with the Transformers library.
- Corrected a duplicate link in the "Quick Start" section to properly point to the `FullStack-Bench` repository.
- Added a "Citation" section at the bottom using the BibTeX provided in your GitHub repository.

These changes will help users discover, use, and cite your work more effectively. Let me know if you have any questions!

Files changed (1) hide show

README.md +25 -6

README.md CHANGED Viewed

@@ -1,25 +1,28 @@
 ---
-license: apache-2.0
 language:
 - en
 metrics:
 - accuracy
-base_model:
-- Qwen/Qwen3-Coder-30B-A3B-Instruct
 pipeline_tag: text-generation
 tags:
 - code
 - agent
 ---
 # FullStack-Agent
 ### Overview
 This model is introduced in the paper ["FullStack-Agent: Enhancing Agentic Full-Stack Web Coding via Development-Oriented Testing and Repository Back-Translation"](https://arxiv.org/abs/2602.03798).
 In this paper, we introduce **FullStack-Agent**, a unified system that combines a multi-agent full-stack development framework equipped with efficient coding and debugging tools ([FullStack-Dev](https://github.com/mnluzimu/FullStack-Dev)), an iterative self-improvement method that improves the abilities of LLMs through repository augmentation and back-translation ([FullStack-Learn](https://github.com/mnluzimu/FullStack-Learn)), and a full-stack development benchmark that comprehensively evaluates frontend, backend, and database functionalities ([FullStack-Bench](https://github.com/mnluzimu/FullStack-Bench)).
 ![method](assets/method.png)
 ### Quick Start
@@ -30,7 +33,7 @@ Instructions for installation and running of the three components are in the fol
 - [FullStack-Learn](https://github.com/mnluzimu/FullStack-Learn)
-- [FullStack-Learn](https://github.com/mnluzimu/FullStack-Learn)
 ### Models
@@ -60,4 +63,20 @@ Using more templates result in better performance in most of the metrics, which
 Experimental results of FullStack-Learn tested on with FullStack-Dev on FullStack-Bench are as follows:
-![experiments](assets/experiments.png)

 ---
+base_model:
+- Qwen/Qwen3-Coder-30B-A3B-Instruct
 language:
 - en
+license: apache-2.0
 metrics:
 - accuracy
 pipeline_tag: text-generation
+library_name: transformers
 tags:
 - code
 - agent
 ---
 # FullStack-Agent
 ### Overview
 This model is introduced in the paper ["FullStack-Agent: Enhancing Agentic Full-Stack Web Coding via Development-Oriented Testing and Repository Back-Translation"](https://arxiv.org/abs/2602.03798).
 In this paper, we introduce **FullStack-Agent**, a unified system that combines a multi-agent full-stack development framework equipped with efficient coding and debugging tools ([FullStack-Dev](https://github.com/mnluzimu/FullStack-Dev)), an iterative self-improvement method that improves the abilities of LLMs through repository augmentation and back-translation ([FullStack-Learn](https://github.com/mnluzimu/FullStack-Learn)), and a full-stack development benchmark that comprehensively evaluates frontend, backend, and database functionalities ([FullStack-Bench](https://github.com/mnluzimu/FullStack-Bench)).
+Official Repository: [mnluzimu/FullStack-Agent](https://github.com/mnluzimu/FullStack-Agent)
 ![method](assets/method.png)
 ### Quick Start
 - [FullStack-Learn](https://github.com/mnluzimu/FullStack-Learn)
+- [FullStack-Bench](https://github.com/mnluzimu/FullStack-Bench)
 ### Models
 Experimental results of FullStack-Learn tested on with FullStack-Dev on FullStack-Bench are as follows:
+![experiments](assets/experiments.png)
+### Citation
+If you find our project helpful, please cite:
+```bibtex
+@misc{lu2026fullstackagentenhancingagenticfullstack,
+      title={FullStack-Agent: Enhancing Agentic Full-Stack Web Coding via Development-Oriented Testing and Repository Back-Translation},
+      author={Zimu Lu and Houxing Ren and Yunqiao Yang and Ke Wang and Zhuofan Zong and Mingjie Zhan and Hongsheng Li},
+      year={2026},
+      eprint={2602.03798},
+      archivePrefix={arXiv},
+      primaryClass={cs.SE},
+      url={https://arxiv.org/abs/2602.03798},
+}
+```