Instructions to use RUC-AIBOX/AEWM with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use RUC-AIBOX/AEWM with Transformers:

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

pipe = pipeline("text-generation", model="RUC-AIBOX/AEWM")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("RUC-AIBOX/AEWM")
model = AutoModelForMultimodalLM.from_pretrained("RUC-AIBOX/AEWM", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use RUC-AIBOX/AEWM with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/RUC-AIBOX/AEWM

SGLang

How to use RUC-AIBOX/AEWM 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 "RUC-AIBOX/AEWM" \
    --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": "RUC-AIBOX/AEWM",
		"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 "RUC-AIBOX/AEWM" \
        --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": "RUC-AIBOX/AEWM",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use RUC-AIBOX/AEWM with Docker Model Runner:
```
docker model run hf.co/RUC-AIBOX/AEWM
```

Add paper link

by nielsr HF Staff - opened 24 days ago

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

+11

-13

Files changed (1) hide show

README.md +11 -13

README.md CHANGED Viewed

@@ -1,26 +1,25 @@
 ---
-license: apache-2.0
 base_model: Qwen/Qwen3.5-35B-A3B
 library_name: transformers
 pipeline_tag: text-generation
-language:
-  - en
 tags:
-  - llm-agents
-  - world-models
-  - agent-editing
-  - action-judgment
-  - tool-use
-  - editact
 ---
 # AEWM: Agent-Editing World Model
 **AEWM is a world model that judges and edits an agent's decisions, rather than predicting tool responses.** Built on **Qwen3.5-35B-A3B**, it combines **Action Judge (AJ)** and **State Revision (SR)** to improve long-horizon reasoning and tool use across **Search, Terminal, and Software Engineering (SWE)**. Its inference framework, **EditAct**, integrates these capabilities into an agent's interaction with the real environment.
-<!-- The arXiv link is intentionally empty until the paper is available. -->
 <p align="center">
-  <a href=""><img src="https://img.shields.io/badge/Paper-arXiv-B31B1B?logo=arxiv&amp;logoColor=white" alt="Paper (arXiv link forthcoming)"></a>
   <a href="https://huggingface.co/RUC-AIBOX/AEWM"><img src="https://img.shields.io/badge/🤗_Hugging_Face-AEWM-FFD21E" alt="AEWM model on Hugging Face"></a>
   <a href="https://huggingface.co/datasets/RUC-AIBOX/AEWM-Action-Judge-Bench"><img src="https://img.shields.io/badge/🤗_Hugging_Face-Benchmark-FFD21E" alt="Action Judge Benchmark on Hugging Face"></a>
   <a href="https://github.com/RUCAIBox/Agent-Editing-World-Model"><img src="https://img.shields.io/badge/Code-GitHub-181717?logo=github&amp;logoColor=white" alt="Code on GitHub"></a>
@@ -113,5 +112,4 @@ export WM_API_KEY="EMPTY"
 `EMPTY` is only for an endpoint without authentication. Configure the agent and tools separately, following the repository's quick start. Use the provided domain-specific AJ/SR prompts and preserve reasoning and tool-call fields; a generic chat prompt is not equivalent to the evaluated workflow.
-The framework provides `editact_for_search`, `editact_terminal`, and `editact_for_swe` scaffolds. Search requires search and webpage-reading services; Terminal and SWE require an execution sandbox. The [standalone Action Judge evaluator](https://github.com/RUCAIBox/Agent-Editing-World-Model/tree/main/benchmarks/action_judge) instead classifies recorded decisions without executing tools.

 ---
 base_model: Qwen/Qwen3.5-35B-A3B
+language:
+- en
 library_name: transformers
+license: apache-2.0
 pipeline_tag: text-generation
 tags:
+- llm-agents
+- world-models
+- agent-editing
+- action-judgment
+- tool-use
+- editact
 ---
 # AEWM: Agent-Editing World Model
 **AEWM is a world model that judges and edits an agent's decisions, rather than predicting tool responses.** Built on **Qwen3.5-35B-A3B**, it combines **Action Judge (AJ)** and **State Revision (SR)** to improve long-horizon reasoning and tool use across **Search, Terminal, and Software Engineering (SWE)**. Its inference framework, **EditAct**, integrates these capabilities into an agent's interaction with the real environment.
 <p align="center">
+  <a href="https://huggingface.co/papers/2609.28416"><img src="https://img.shields.io/badge/Paper-arXiv-B31B1B?logo=arxiv&amp;logoColor=white" alt="Paper"></a>
   <a href="https://huggingface.co/RUC-AIBOX/AEWM"><img src="https://img.shields.io/badge/🤗_Hugging_Face-AEWM-FFD21E" alt="AEWM model on Hugging Face"></a>
   <a href="https://huggingface.co/datasets/RUC-AIBOX/AEWM-Action-Judge-Bench"><img src="https://img.shields.io/badge/🤗_Hugging_Face-Benchmark-FFD21E" alt="Action Judge Benchmark on Hugging Face"></a>
   <a href="https://github.com/RUCAIBox/Agent-Editing-World-Model"><img src="https://img.shields.io/badge/Code-GitHub-181717?logo=github&amp;logoColor=white" alt="Code on GitHub"></a>
 `EMPTY` is only for an endpoint without authentication. Configure the agent and tools separately, following the repository's quick start. Use the provided domain-specific AJ/SR prompts and preserve reasoning and tool-call fields; a generic chat prompt is not equivalent to the evaluated workflow.
+The framework provides `editact_for_search`, `editact_terminal`, and `editact_for_swe` scaffolds. Search requires search and webpage-reading services; Terminal and SWE require an execution sandbox. The [standalone Action Judge evaluator](https://github.com/RUCAIBox/Agent-Editing-World-Model/tree/main/benchmarks/action_judge) instead classifies recorded decisions without executing tools.