Instructions to use THU-KEG/SIRI-7B-low with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use THU-KEG/SIRI-7B-low with Transformers:

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

pipe = pipeline("text-generation", model="THU-KEG/SIRI-7B-low")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("THU-KEG/SIRI-7B-low")
model = AutoModelForCausalLM.from_pretrained("THU-KEG/SIRI-7B-low")
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

vLLM

How to use THU-KEG/SIRI-7B-low with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/THU-KEG/SIRI-7B-low

SGLang

How to use THU-KEG/SIRI-7B-low 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 "THU-KEG/SIRI-7B-low" \
    --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": "THU-KEG/SIRI-7B-low",
		"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 "THU-KEG/SIRI-7B-low" \
        --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": "THU-KEG/SIRI-7B-low",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use THU-KEG/SIRI-7B-low with Docker Model Runner:
```
docker model run hf.co/THU-KEG/SIRI-7B-low
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

SIRI: Scaling Iterative Reinforcement Learning with Interleaved Compression

📃 Paper • 📝 Wandb

🔍 Overview

SIRI (Scaling Iterative Reinforcement Learning with Interleaved Compression) is a reinforcement-learning–based framework designed to improve the efficiency and accuracy of Large Reasoning Models (LRMs).

Traditional RL training often causes overthinking and long, redundant reasoning traces. Prior methods that compress outputs (length penalties, pruning, or skipping thought tokens) improve efficiency but hurt accuracy.

SIRI solves this trade-off by iteratively alternating between compression and expansion of the reasoning budget, controlled by a cosine length scheduler. This approach dynamically balances concise reasoning with long-horizon exploration.

pareto_front

🚀 Key Features

Interleaved Compression–Expansion:
- Compression phase: forces concise, high-density reasoning by limiting rollout length.
- Expansion phase: restores longer rollouts to encourage exploration and planning.
Token Efficiency without Accuracy Loss: Unlike previous methods, SIRI improves accuracy while reducing average token usage.
Iterative RL Training: Built on GRPO with modifications from DAPO (clip-high/low decoupling, KL removal).
Generalization Across Model Sizes: Validated on both 1.5B and 7B models.

📊 Benchmarks

📝 Citation

@misc{wen2025siriscalingiterativereinforcement,
      title={SIRI: Scaling Iterative Reinforcement Learning with Interleaved Compression}, 
      author={Haoming Wen and Yushi Bai and Juanzi Li and Jie Tang},
      year={2025},
      eprint={2509.25176},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2509.25176}, 
}