Instructions to use xxwu/Agent-STAR-SFT-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use xxwu/Agent-STAR-SFT-7B with Transformers:

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

pipe = pipeline("text-generation", model="xxwu/Agent-STAR-SFT-7B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("xxwu/Agent-STAR-SFT-7B")
model = AutoModelForCausalLM.from_pretrained("xxwu/Agent-STAR-SFT-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]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use xxwu/Agent-STAR-SFT-7B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "xxwu/Agent-STAR-SFT-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": "xxwu/Agent-STAR-SFT-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/xxwu/Agent-STAR-SFT-7B

SGLang

How to use xxwu/Agent-STAR-SFT-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 "xxwu/Agent-STAR-SFT-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": "xxwu/Agent-STAR-SFT-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 "xxwu/Agent-STAR-SFT-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": "xxwu/Agent-STAR-SFT-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use xxwu/Agent-STAR-SFT-7B with Docker Model Runner:
```
docker model run hf.co/xxwu/Agent-STAR-SFT-7B
```

Add paper link and improve model card for Agent-STAR SFT 7B

by nielsr HF Staff - opened Mar 25

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

+28

-16

Files changed (1) hide show

README.md +28 -16

README.md CHANGED Viewed

@@ -1,34 +1,34 @@
 ---
 library_name: transformers
 license: other
-base_model: Qwen/Qwen2.5-7B-Instruct
 tags:
 - llama-factory
 - full
 - generated_from_trainer
 model-index:
 - name: TravelPlanner_Instruct_7B_1128_1K_4Epoch
   results: []
 ---
-<!-- This model card has been generated automatically according to the information the Trainer had access to. You
-should probably proofread and complete it, then remove this comment. -->
 # TravelPlanner_Instruct_7B_1128_1K_4Epoch
-This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the travelplanner_1108_5k_filter_1128_1k dataset.
 ## Model description
-More information needed
-## Intended uses & limitations
-More information needed
-## Training and evaluation data
-More information needed
 ## Training procedure
@@ -44,18 +44,30 @@ The following hyperparameters were used during training:
 - gradient_accumulation_steps: 4
 - total_train_batch_size: 32
 - total_eval_batch_size: 64
-- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_ratio: 0.1
 - num_epochs: 4.0
-### Training results
 ### Framework versions
 - Transformers 4.57.1
 - Pytorch 2.9.0+cu128
 - Datasets 4.0.0
 - Tokenizers 0.22.1

 ---
+base_model: Qwen/Qwen2.5-7B-Instruct
 library_name: transformers
 license: other
+pipeline_tag: text-generation
 tags:
 - llama-factory
 - full
 - generated_from_trainer
+- agent
+- tool-use
 model-index:
 - name: TravelPlanner_Instruct_7B_1128_1K_4Epoch
   results: []
 ---
 # TravelPlanner_Instruct_7B_1128_1K_4Epoch
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) specialized for long-horizon tool-use tasks. It is part of the **Agent-STAR** project, introduced in the paper [Demystifying Reinforcement Learning for Long-Horizon Tool-Using Agents: A Comprehensive Recipe](https://huggingface.co/papers/2603.21972).
 ## Model description
+This model corresponds to the **STAR-SFT-7B** checkpoint. It was fine-tuned from the Qwen2.5-Instruct base using **1K successful trajectories** generated by DeepSeek-V3 on synthetic queries within the TravelPlanner environment. This model serves as the Supervised Fine-Tuning (SFT) foundation for the reinforcement learning recipes explored in the paper.
+- **Paper:** [Demystifying Reinforcement Learning for Long-Horizon Tool-Using Agents: A Comprehensive Recipe](https://huggingface.co/papers/2603.21972)
+- **Repository:** [https://github.com/WxxShirley/Agent-STAR](https://github.com/WxxShirley/Agent-STAR)
+- **Dataset:** [Agent-STAR-TravelDataset](https://huggingface.co/datasets/xxwu/Agent-STAR-TravelDataset)
+## Intended uses & limitations
+The model is designed to act as an autonomous agent capable of long-horizon planning and tool orchestration. It is specifically optimized for the TravelPlanner benchmark, which requires satisfying multifaceted constraints (commonsense and hard constraints) across multi-turn interactions.
 ## Training procedure
 - gradient_accumulation_steps: 4
 - total_train_batch_size: 32
 - total_eval_batch_size: 64
+- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_ratio: 0.1
 - num_epochs: 4.0
 ### Framework versions
 - Transformers 4.57.1
 - Pytorch 2.9.0+cu128
 - Datasets 4.0.0
 - Tokenizers 0.22.1
+## Citation
+If you find this model helpful, please cite:
+```bibtex
+@misc{wu2026agentstar,
+      title={Demystifying Reinforcement Learning for Long-Horizon Tool-Using Agents: A Comprehensive Recipe},
+      author={Xixi Wu and Qianguo Sun and Ruiyang Zhang and Chao Song and Junlong Wu and Yiyan Qi and Hong Cheng},
+      year={2026},
+      eprint={2603.21972},
+      archivePrefix={arXiv},
+      primaryClass={cs.LG},
+      url={https://arxiv.org/abs/2603.21972},
+}
+```