Instructions to use inclusionAI/AReaL-boba-2-8B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use inclusionAI/AReaL-boba-2-8B with Transformers:

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

pipe = pipeline("text-generation", model="inclusionAI/AReaL-boba-2-8B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("inclusionAI/AReaL-boba-2-8B")
model = AutoModelForCausalLM.from_pretrained("inclusionAI/AReaL-boba-2-8B", 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 inclusionAI/AReaL-boba-2-8B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "inclusionAI/AReaL-boba-2-8B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "inclusionAI/AReaL-boba-2-8B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/inclusionAI/AReaL-boba-2-8B

SGLang

How to use inclusionAI/AReaL-boba-2-8B 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 "inclusionAI/AReaL-boba-2-8B" \
    --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": "inclusionAI/AReaL-boba-2-8B",
		"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 "inclusionAI/AReaL-boba-2-8B" \
        --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": "inclusionAI/AReaL-boba-2-8B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use inclusionAI/AReaL-boba-2-8B with Docker Model Runner:
```
docker model run hf.co/inclusionAI/AReaL-boba-2-8B
```

Add metadata, link to Github

by nielsr HF Staff - opened Jun 4, 2025

base: refs/heads/main

←

from: refs/pr/2

Discussion Files changed

-8

Files changed (1) hide show

README.md +8 -8

README.md CHANGED Viewed

@@ -1,6 +1,9 @@
 ---
-{}
 ---
 <h1 align="center">
 <em>AReaL</em>: Ant Reasoning Reinforcement Learning for LLMs
 </h1>
@@ -9,6 +12,7 @@
 | <a href="https://arxiv.org/pdf/2505.24298"><b>Paper</b></a> | <a href="https://inclusionai.github.io/AReaL/"><b>Documentation</b></a> | <a href="https://deepwiki.com/inclusionAI/AReaL"><b>Ask DeepWiki</b></a> | <a href="https://huggingface.co/collections/inclusionAI/areal-boba-2-683f0e819ccb7bb2e1b2f2d5"><b>🤗 Models & Data</b></a> |
 </p>
 AReaL (Ant Reasoning RL) is an open-source **fully asynchronous reinforcement learning training system** for large reasoning models developed at **the RL Lab, Ant Research**. Built upon the open-source project [RealHF](https://github.com/openpsi-project/ReaLHF), we are fully committed to open-source by providing training details, data, and infrastructure required to reproduce results along with the model itself. AReaL aims to help everyone build their own AI agents easily and affordably. Our team loves milk tea because it's delicious, customizable, and affordable. We hope you enjoy our project just like how you enjoy real-world milk tea (cheers).
@@ -37,7 +41,7 @@ In our AReaL-boba² (A-ReaL-double-boba) release, we highlight the top 3 most im
 + Experimental support for **multi-turn** agentic RL training. Check our [complete example](https://inclusionai.github.io/AReaL/customization/agent.html).
-For the complete system design and more training details, please check [our v0.3 blog](/blog/AReaL_v0_3.md) and our [research paper](about:blank) for a more comprehensive presentation of our system design.
 ### Overview of Asynchronous RL Training
@@ -92,7 +96,7 @@ We highlight the [tutorials](https://inclusionai.github.io/AReaL/customization/d
 + [Streaming generation and reward computation](https://inclusionai.github.io/AReaL/developer/rollout/rollout_worker.html)
 + [Interruptible rollout](https://inclusionai.github.io/AReaL/developer/rollout/gserver.html)
 + [Data staleness control with the rollout controller](https://inclusionai.github.io/AReaL/developer/rollout/gserver.html)
-+ [The adoption of decoupled PPO loss](https://inclusionai.github.io/AReaL/customization/algorithm.html)
 ### RL Training for Multi-turn Agent
@@ -100,12 +104,8 @@ AReaL-boba² allows you to independently customize the [dataset](https://inclusi
 In particular, we show a simple example to develop a multi-turn math agent for RL training. Please see the learning curve below and reference the [step-by-step guide](https://inclusionai.github.io/AReaL/customization/agent.html) if you want to implement your own agentic RL project.
-**Multi-turn Agent Learning Curve**
 ## Getting Started
-### Quick Start
 Train Qwen3 1.7B locally:
 ```bash
@@ -214,4 +214,4 @@ We also appreciate all the pioneering works from the community, particularly the
       primaryClass={cs.LG},
       url={https://arxiv.org/abs/2505.24298},
 }
-```

 ---
+license: apache-2.0
+library_name: transformers
+pipeline_tag: text-generation
 ---
 <h1 align="center">
 <em>AReaL</em>: Ant Reasoning Reinforcement Learning for LLMs
 </h1>
 | <a href="https://arxiv.org/pdf/2505.24298"><b>Paper</b></a> | <a href="https://inclusionai.github.io/AReaL/"><b>Documentation</b></a> | <a href="https://deepwiki.com/inclusionAI/AReaL"><b>Ask DeepWiki</b></a> | <a href="https://huggingface.co/collections/inclusionAI/areal-boba-2-683f0e819ccb7bb2e1b2f2d5"><b>🤗 Models & Data</b></a> |
 </p>
+Code: https://github.com/inclusionAI/AReaL
 AReaL (Ant Reasoning RL) is an open-source **fully asynchronous reinforcement learning training system** for large reasoning models developed at **the RL Lab, Ant Research**. Built upon the open-source project [RealHF](https://github.com/openpsi-project/ReaLHF), we are fully committed to open-source by providing training details, data, and infrastructure required to reproduce results along with the model itself. AReaL aims to help everyone build their own AI agents easily and affordably. Our team loves milk tea because it's delicious, customizable, and affordable. We hope you enjoy our project just like how you enjoy real-world milk tea (cheers).
 + Experimental support for **multi-turn** agentic RL training. Check our [complete example](https://inclusionai.github.io/AReaL/customization/agent.html).
+For the complete system design and more training details, please check [our v0.3 blog](/blog/AReaL_v0_3.md) and our [research paper](https://arxiv.org/pdf/2505.24298).
 ### Overview of Asynchronous RL Training
 + [Streaming generation and reward computation](https://inclusionai.github.io/AReaL/developer/rollout/rollout_worker.html)
 + [Interruptible rollout](https://inclusionai.github.io/AReaL/developer/rollout/gserver.html)
 + [Data staleness control with the rollout controller](https://inclusionai.github.io/AReaL/developer/rollout/gserver.html)
++ [The adoption of decoupled PPO loss](https://inclusionai.github.io/AReaL/customization/algorithm.html#the-decoupled-ppo-loss)
 ### RL Training for Multi-turn Agent
 In particular, we show a simple example to develop a multi-turn math agent for RL training. Please see the learning curve below and reference the [step-by-step guide](https://inclusionai.github.io/AReaL/customization/agent.html) if you want to implement your own agentic RL project.
 ## Getting Started
 Train Qwen3 1.7B locally:
 ```bash
       primaryClass={cs.LG},
       url={https://arxiv.org/abs/2505.24298},
 }
+```