Instructions to use PKU-ML/G1-3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use PKU-ML/G1-3B with Transformers:

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

pipe = pipeline("text-generation", model="PKU-ML/G1-3B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("PKU-ML/G1-3B")
model = AutoModelForCausalLM.from_pretrained("PKU-ML/G1-3B", 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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use PKU-ML/G1-3B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "PKU-ML/G1-3B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PKU-ML/G1-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/PKU-ML/G1-3B

SGLang

How to use PKU-ML/G1-3B 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 "PKU-ML/G1-3B" \
    --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": "PKU-ML/G1-3B",
		"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 "PKU-ML/G1-3B" \
        --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": "PKU-ML/G1-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use PKU-ML/G1-3B with Docker Model Runner:
```
docker model run hf.co/PKU-ML/G1-3B
```

Update pipeline tag and add project page link

by nielsr HF Staff - opened Aug 20, 2025

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

+26

-18

Files changed (1) hide show

README.md +26 -18

README.md CHANGED Viewed

@@ -1,21 +1,20 @@
 ---
-license: apache-2.0
 datasets:
 - PKU-ML/Erdos
 language:
 - en
 metrics:
 - accuracy
-base_model:
-- Qwen/Qwen2.5-3B-Instruct
-pipeline_tag: text-generation
 tags:
 - graph
 - chat
-library_name: transformers
 ---
 # G1-3B
 ## Introduction
@@ -25,19 +24,20 @@ We apply Group Relative Policy Optimization (GRPO) for reinforcement learning wi
 G1 brings the following improvements:
-- **Significant improvement on graph reasoning**: G1 models achieve up to 46% improvement over baselines on Erdős, with the 7B variant matching OpenAI’s o3-mini and the 3B model surpassing Qwen2.5-72B-Instruct by notable margins.
-- **Strong Generalization to unseen graph tasks**: G1 exhibits zero-shot generalization on unseen graph tasks, improving performance on *other graph reasoning benchmarks* (GraphWiz, GraphArena) and *real-world graphs* (Cora, PubMed).
-- **NO Compromise on general reasoning**: Crucially, G1 preserves general reasoning ability (GSM8K, MATH, MMLU-Pro), proving its versatility.
 **This repo contains the G1-3B model**, which has the following features:
-- Type: Causal Language Models
-- Training Stage: SFT & RL
-- Architecture: the same with Qwen2.5-Instruct
-- Number of Parameters: 3.09B
-- Context Length: Full 32,768 tokens and generation 8192 tokens
 For more details, please refer to our [paper](https://arxiv.org/pdf/2505.18499) and [GitHub](https://github.com/PKU-ML/G1/tree/main).
 ## Requirements
@@ -72,10 +72,18 @@ model = AutoModelForCausalLM.from_pretrained(
 )
 tokenizer = AutoTokenizer.from_pretrained(model_name)
-prompt = "The task is to determine the degree centrality of a node in the graph.\n\n"\
-        "Degree centrality for a node is the fraction of nodes it is connected to.\n\n"\
-        "Here is an undirected graph containing nodes from 1 to 15. The edges are: (1, 15), (15, 11), (2, 3), (2, 6), (3, 6), (3, 7), (6, 7), (6, 8), (7, 8), (7, 14), (4, 10), (10, 5), (10, 12), (8, 14), (8, 9), (12, 11), (12, 13).\n\n"\
-        "Question: What is the degree centrality of node 2 in the graph?\n\n"\
         "You need to format your answer as a float number."
 messages = [
     {"role": "user", "content": INSTRUCTION_TEMPLATE.format(instruction=prompt)}

 ---
+base_model:
+- Qwen/Qwen2.5-3B-Instruct
 datasets:
 - PKU-ML/Erdos
 language:
 - en
+library_name: transformers
+license: apache-2.0
 metrics:
 - accuracy
+pipeline_tag: graph-ml
 tags:
 - graph
 - chat
 ---
 # G1-3B
 ## Introduction
 G1 brings the following improvements:
+-   **Significant improvement on graph reasoning**: G1 models achieve up to 46% improvement over baselines on Erdős, with the 7B variant matching OpenAI’s o3-mini and the 3B model surpassing Qwen2.5-72B-Instruct by notable margins.
+-   **Strong Generalization to unseen graph tasks**: G1 exhibits zero-shot generalization on unseen graph tasks, improving performance on *other graph reasoning benchmarks* (GraphWiz, GraphArena) and *real-world graphs* (Cora, PubMed).
+-   **NO Compromise on general reasoning**: Crucially, G1 preserves general reasoning ability (GSM8K, MATH, MMLU-Pro), proving its versatility.
 **This repo contains the G1-3B model**, which has the following features:
+-   Type: Causal Language Models
+-   Training Stage: SFT & RL
+-   Architecture: the same with Qwen2.5-Instruct
+-   Number of Parameters: 3.09B
+-   Context Length: Full 32,768 tokens and generation 8192 tokens
 For more details, please refer to our [paper](https://arxiv.org/pdf/2505.18499) and [GitHub](https://github.com/PKU-ML/G1/tree/main).
+Project page: [Hugging Face Collection](https://huggingface.co/collections/PKU-ML/g1-683d659e992794fc99618cf2)
 ## Requirements
 )
 tokenizer = AutoTokenizer.from_pretrained(model_name)
+prompt = "The task is to determine the degree centrality of a node in the graph.
+"\
+        "Degree centrality for a node is the fraction of nodes it is connected to.
+"\
+        "Here is an undirected graph containing nodes from 1 to 15. The edges are: (1, 15), (15, 11), (2, 3), (2, 6), (3, 6), (3, 7), (6, 7), (6, 8), (7, 8), (7, 14), (4, 10), (10, 5), (10, 12), (8, 14), (8, 9), (12, 11), (12, 13).
+"\
+        "Question: What is the degree centrality of node 2 in the graph?
+"\
         "You need to format your answer as a float number."
 messages = [
     {"role": "user", "content": INSTRUCTION_TEMPLATE.format(instruction=prompt)}