Instructions to use xx18/Composition-RL-4B-Depth1_2_3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use xx18/Composition-RL-4B-Depth1_2_3 with Transformers:

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

pipe = pipeline("text-generation", model="xx18/Composition-RL-4B-Depth1_2_3")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("xx18/Composition-RL-4B-Depth1_2_3")
model = AutoModelForCausalLM.from_pretrained("xx18/Composition-RL-4B-Depth1_2_3")
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 xx18/Composition-RL-4B-Depth1_2_3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "xx18/Composition-RL-4B-Depth1_2_3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "xx18/Composition-RL-4B-Depth1_2_3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/xx18/Composition-RL-4B-Depth1_2_3

SGLang

How to use xx18/Composition-RL-4B-Depth1_2_3 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 "xx18/Composition-RL-4B-Depth1_2_3" \
    --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": "xx18/Composition-RL-4B-Depth1_2_3",
		"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 "xx18/Composition-RL-4B-Depth1_2_3" \
        --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": "xx18/Composition-RL-4B-Depth1_2_3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use xx18/Composition-RL-4B-Depth1_2_3 with Docker Model Runner:
```
docker model run hf.co/xx18/Composition-RL-4B-Depth1_2_3
```

Composition-RL-8B

This repository contains the 8B model checkpoint for Composition-RL, introduced in the paper Composition-RL: Compose Your Verifiable Prompts for Reinforcement Learning of Large Language Models.

Overview

Composition-RL is a data-efficient Reinforcement Learning with Verifiable Rewards (RLVR) approach. It addresses the challenge of "too-easy" prompts (where the pass rate reaches 1) by automatically composing multiple verifiable problems into a single, harder yet still-verifiable prompt. This ensures that RL training continues to receive informative signals as the model's reasoning capabilities improve.

Model Details

Base Model: Qwen3-8B-Base
Training Dataset: MATH-Composition-199K
Framework: Composition-RL
Paper: Composition-RL: Compose Your Verifiable Prompts for Reinforcement Learning of Large Language Models
Code: GitHub - XinXU-USTC/Composition-RL

Citation

If you find this work helpful for your research, please consider citing:

@article{xu2026composition-rl,
  title={Composition-RL: Compose Your Verifiable Prompts for Reinforcement Learning of Large Language Models},
  author={Xu, Xin and Bai, Clive and Yang, Kai and Chen, Tianhao and Chen, Yangkun and Liu, Weijie and Chen, Hao and Wang, Yang and Yang, Saiyong and Yang, Can},
  journal={arXiv preprint arXiv:2602.12036},
  year={2026}
}

Downloads last month: 21

Collection including xx18/Composition-RL-4B-Depth1_2_3

Composition-RL

Collection

Datasets and trained checkpoints of Composition-RL: https://github.com/XinXU-USTC/Composition-RL • 13 items • Updated Mar 19 • 1

Paper for xx18/Composition-RL-4B-Depth1_2_3

Composition-RL: Compose Your Verifiable Prompts for Reinforcement Learning of Large Language Models

Paper • 2602.12036 • Published Feb 12 • 93