LIMO: Less Is More for Reasoning 🚀

📌 Table of Contents

• Overview
• Key Results
• Model Zoo
• Datasets
• Quick Start
• Training
• Evaluation
• Citation

Overview

LIMO challenges the conventional wisdom in mathematical reasoning by demonstrating that models can achieve superior performance with significantly less but higher quality training data. Our approach:

• 🎯 Achieves SOTA with only 817 carefully curated training samples
• 🌟 Shows strong generalization across diverse problem types
• 🔬 Provides comprehensive evaluation on 10 benchmarks
• 📚 Releases high-quality datasets and evaluation tools

Key Results

Model	AIME24	MATH500	Training Samples
LIMO (Ours)	57.1%	94.8%	817
Previous SOTA	6.5%	59.2%	100k+

Click to see more detailed results

Benchmark	LIMO	Previous SOTA	Improvement
AIME24	57.1%	6.5%	+50.6%
MATH500	94.8%	59.2%	+35.6%
AMC23	92.0%	40.6%	+51.4%
OlympiadBench	66.8%	36.7%	+30.1%
CHMath	75.4%	11.2%	+64.2%
Gaokao	81.0%	49.4%	+31.6%
Kaoyan	73.4%	32.7%	+40.7%
GradeSchool	76.2%	36.2%	+40.0%
Minerva	44.9%	47.1%	-2.2%
GPQA	66.7%	73.3%	-6.6%

Model Zoo

Our LIMO model is available on Hugging Face 🤗:

Model	Backbone	Size	Link
LIMO	Qwen2.5-32B-Instruct	32B	🤗

Datasets

We release our datasets through Hugging Face 🤗:

Dataset	Description	Size	Link
LIMO	Training set used to train LIMO model	817	🤗

Note: We are gradually releasing additional datasets mentioned in our paper, including those used for comparative experiments, to facilitate reproducibility and further analysis by the research community. Stay tuned!

Quick Start

Our model is fine-tuned on Qwen2.5-32B-Instruct and is compatible with most mainstream frameworks like HF Transformers, VLLM, TensorRT-LLM and etc.

Start with HF Transformers

# Install required packages
pip install transformers

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

# Initialize model and tokenizer
model = AutoModelForCausalLM.from_pretrained(
    "GAIR/LIMO",
    torch_dtype="auto",
    trust_remote_code=True,
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("GAIR/LIMO", trust_remote_code=True)

# Prepare input messages (We use the following template and system prompt during training and inference)
messages = [
    {"role": "system", "content": "Please reason step by step, and put your final answer within \\boxed{}."},
    {"role": "user", "content": "What is the result of 1+1?"}
]

# Format input using chat template
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)

# Tokenize input
inputs = tokenizer(text, return_tensors="pt").to(model.device)

# Generate response
outputs = model.generate(
    **inputs,
    max_new_tokens=32768,
    temperature=0.7,
    top_p=0.95,
    do_sample=True
)

# Decode and print response
response = tokenizer.decode(outputs[0][inputs['input_ids'].shape[1]:], skip_special_tokens=True)
print(response)

Start with VLLM

# Install required packages
pip install vllm

from vllm import LLM, SamplingParams
from transformers import AutoTokenizer

# Initialize the model
llm = LLM(
    model="GAIR/LIMO",
    tensor_parallel_size=4,  # adjust based on available GPUs
    trust_remote_code=True,
    swap_space=60,
    gpu_memory_utilization=0.96,
)

# Prepare input messages (We use the following template and system prompt during training and inference)
messages = [
    {"role": "system", "content": "Please reason step by step, and put your final answer within \\boxed{}."},
    {"role": "user", "content": "What is the result of 1+1?"}
]

# Setup tokenizer
tokenizer = AutoTokenizer.from_pretrained("GAIR/LIMO", trust_remote_code=True)
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)

# Configure generation parameters
sampling_params = SamplingParams(
    temperature=0.7,
    max_tokens=32768,
    top_p=0.95,
)

# Generate response
output = llm.generate(text, sampling_params)
print(output[0].outputs[0].text)

License

This project is licensed under the MIT License - see the LICENSE file for details.

Citation

@misc{ye2025limoreasoning,
      title={LIMO: Less is More for Reasoning}, 
      author={Yixin Ye and Zhen Huang and Yang Xiao and Ethan Chern and Shijie Xia and Pengfei Liu},
      year={2025},
      eprint={2502.03387},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2502.03387}, 
}