Instructions to use PersonalAILab/A2FM-32B-rl with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use PersonalAILab/A2FM-32B-rl with Transformers:

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

pipe = pipeline("text-generation", model="PersonalAILab/A2FM-32B-rl")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("PersonalAILab/A2FM-32B-rl")
model = AutoModelForCausalLM.from_pretrained("PersonalAILab/A2FM-32B-rl")
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 PersonalAILab/A2FM-32B-rl with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "PersonalAILab/A2FM-32B-rl"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PersonalAILab/A2FM-32B-rl",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/PersonalAILab/A2FM-32B-rl

SGLang

How to use PersonalAILab/A2FM-32B-rl 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 "PersonalAILab/A2FM-32B-rl" \
    --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": "PersonalAILab/A2FM-32B-rl",
		"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 "PersonalAILab/A2FM-32B-rl" \
        --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": "PersonalAILab/A2FM-32B-rl",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use PersonalAILab/A2FM-32B-rl with Docker Model Runner:
```
docker model run hf.co/PersonalAILab/A2FM-32B-rl
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

🧠 A²FM: Adaptive Agent Foundation Model for Tool-Aware Hybrid Reasoning

A²FM (Adaptive Agent Foundation Model) unifies reasoning-centric and agentic paradigms into a single framework that adaptively selects among three execution modes — instant, reasoning, and agentic.
It follows a route-then-align training principle and introduces Adaptive Policy Optimization (APO) to jointly optimize accuracy and efficiency.

A²FM achieves state-of-the-art performance on major reasoning and agentic benchmarks:

13.4% on BrowseComp (agentic)
70.4% on AIME25 (reasoning)
16.7% on HLE (general)

Notably, its adaptive execution achieves a cost of pass of only $0.00487 per correct answer, cutting cost by 45.2% vs. reasoning and 33.5% vs. agentic, delivering substantially higher cost efficiency while maintaining comparable accuracy.

📄 Paper
💻 GitHub

🔑 Key Highlights

⚙️ Unified reasoning & agentic modeling
Integrates direct reasoning, chain-of-thought, and tool-augmented actions within a single backbone.
🔄 Route-then-Align supervised fine-tuning
Trains task-aware routing followed by mode-aligned trajectory learning.
🧩 Adaptive Policy Optimization (APO)
Reinforcement learning with adaptive sampling and cost-regularized reward for efficiency–accuracy balance.
💡 Substantially lower inference cost
Adaptive routing cuts redundant reasoning/tool use while preserving correctness.

📘 Citation

@article{chen2025textsuperscript,
  title={A$\backslash$textsuperscript $\{$2$\}$ FM: An Adaptive Agent Foundation Model for Tool-Aware Hybrid Reasoning},
  author={Chen, Qianben and Cao, Jingyi and Zhang, Jiayu and Qin, Tianrui and Li, Xiaowan and Zhu, King and Shi, Dingfeng and Zhu, He and Liu, Minghao and Liang, Xiaobo and others},
  journal={arXiv preprint arXiv:2510.12838},
  year={2025}
}

Downloads last month: 2

Safetensors

Model size

33B params

Tensor type

BF16

Model tree for PersonalAILab/A2FM-32B-rl

Quantizations

2 models

Paper for PersonalAILab/A2FM-32B-rl

A^2FM: An Adaptive Agent Foundation Model for Tool-Aware Hybrid Reasoning

Paper • 2510.12838 • Published Oct 13, 2025 • 25