PKU-ML
/

G1-Direct-SFT-7B

@@ -1,199 +1,122 @@
 ---
 library_name: transformers
-tags: []
 ---
-# Model Card for Model ID
-<!-- Provide a quick summary of what the model is/does. -->
-## Model Details
-### Model Description
-<!-- Provide a longer summary of what this model is. -->
-This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
-- **Developed by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Model type:** [More Information Needed]
-- **Language(s) (NLP):** [More Information Needed]
-- **License:** [More Information Needed]
-- **Finetuned from model [optional]:** [More Information Needed]
-### Model Sources [optional]
-<!-- Provide the basic links for the model. -->
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
-## Uses
-<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
-### Direct Use
-<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
-[More Information Needed]
-### Downstream Use [optional]
-<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
-[More Information Needed]
-### Out-of-Scope Use
-<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
-[More Information Needed]
-## Bias, Risks, and Limitations
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
-[More Information Needed]
-### Recommendations
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
-Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
-## How to Get Started with the Model
-Use the code below to get started with the model.
-[More Information Needed]
-## Training Details
-### Training Data
-<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
-[More Information Needed]
-### Training Procedure
-<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
-#### Preprocessing [optional]
-[More Information Needed]
-#### Training Hyperparameters
-- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
-#### Speeds, Sizes, Times [optional]
-<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
-[More Information Needed]
-## Evaluation
-<!-- This section describes the evaluation protocols and provides the results. -->
-### Testing Data, Factors & Metrics
-#### Testing Data
-<!-- This should link to a Dataset Card if possible. -->
-[More Information Needed]
-#### Factors
-<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
-[More Information Needed]
-#### Metrics
-<!-- These are the evaluation metrics being used, ideally with a description of why. -->
-[More Information Needed]
-### Results
-[More Information Needed]
-#### Summary
-## Model Examination [optional]
-<!-- Relevant interpretability work for the model goes here -->
-[More Information Needed]
-## Environmental Impact
-<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
-Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
-- **Hardware Type:** [More Information Needed]
-- **Hours used:** [More Information Needed]
-- **Cloud Provider:** [More Information Needed]
-- **Compute Region:** [More Information Needed]
-- **Carbon Emitted:** [More Information Needed]
-## Technical Specifications [optional]
-### Model Architecture and Objective
-[More Information Needed]
-### Compute Infrastructure
-[More Information Needed]
-#### Hardware
-[More Information Needed]
-#### Software
-[More Information Needed]
-## Citation [optional]
-<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
-**BibTeX:**
-[More Information Needed]
-**APA:**
-[More Information Needed]
-## Glossary [optional]
-<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
-[More Information Needed]
-## More Information [optional]
-[More Information Needed]
-## Model Card Authors [optional]
-[More Information Needed]
-## Model Card Contact
-[More Information Needed]

 ---
+license: apache-2.0
+datasets:
+- PKU-ML/Erdos
+language:
+- en
+metrics:
+- accuracy
+base_model:
+- Qwen/Qwen2.5-7B-Instruct
+pipeline_tag: text-generation
+tags:
+- graph
+- chat
 library_name: transformers
 ---
+# G1-Direct-SFT-7B
+## Introduction
+G1 is the series of large language models trained on our benchmark [Erdos](https://huggingface.co/datasets/PKU-ML/Erdos) for solving graph reasoning tasks, based on Qwen2.5-Instruct.
+We apply Group Relative Policy Optimization (GRPO) for reinforcement learning with supervised finetuning as a prelimary step.
+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-Direct-SFT-7B model**, which has the following features:
+- Type: Causal Language Models
+- Training Stage: SFT
+- Architecture: the same with Qwen2.5-Instruct
+- Number of Parameters: 7.62B
+- 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
+The model is trained based on Qwen/Qwen2.5-7B-Instruct. The code of Qwen2.5 has been in the latest Hugging face `transformers` and we advise you to use the latest version of `transformers`.
+With `transformers<4.37.0`, you will encounter the following error:
+```
+KeyError: 'qwen2'
+```
+## Quickstart
+Here provides a code snippet with `apply_chat_template` to show you how to load the tokenizer and model and how to generate contents.
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+INSTRUCTION_TEMPLATE = """
+    {instruction}
+    Solve the above problem efficiently and clearly. The last line of your response should be of the following format: 'Therefore, the final answer is: $\\boxed{{ANSWER}}$. I hope it is correct' (without quotes) where ANSWER is just the final number or expression that solves the problem. Think step by step before answering.
+    """.strip()
+model_name = "PKU-ML/G1-Direct-SFT-7B"
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    torch_dtype="auto",
+    device_map="auto"
+)
+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)}
+]
+text = tokenizer.apply_chat_template(
+    messages,
+    tokenize=False,
+    add_generation_prompt=True
+)
+model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
+generated_ids = model.generate(
+    **model_inputs,
+    max_new_tokens=4096,
+    top_p=0.95,
+    top_k=30,
+    temperature=0.6
+)
+generated_ids = [
+    output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
+]
+response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
+print(response)
+```
+## Evaluation & Performance
+Detailed evaluation results are reported in this [📑 paper](https://arxiv.org/pdf/2505.18499).
+## Citation
+If you find our work helpful, feel free to give us a cite.
+```
+@article{guo2025g1,
+  title={G1: Teaching LLMs to Reason on Graphs with Reinforcement Learning},
+  author={Guo, Xiaojun and Li, Ang and Wang, Yifei and Jegelka, Stefanie and Wang, Yisen},
+  journal={arXiv preprint arXiv:2505.18499},
+  year={2025}
+}
+```