F

Fairyr1 32B

Developed by PKU-DS-LAB
FairyR1-32B is an efficient large language model based on DeepSeek-R1-Distill-Qwen-32B, optimized through distillation and merging processes, excelling in mathematical and programming tasks.
Downloads 372
Release Time : 5/23/2025

Model Overview

FairyR1-32B is a high-efficiency language model focused on mathematical and programming tasks. Through task-oriented fine-tuning and model merging techniques, it significantly reduces model size and inference costs while maintaining high performance.

Model Features

High Efficiency
Using only about 5% of the parameters, it performs comparably or better than larger models in mathematical and programming tasks.
Optimized Distillation Process
Through multi-stage data filtering and reconstruction, high-quality training samples are selected to enhance model performance.
Model Merging Technique
Using the AcreeFusion tool to merge multiple expert models into a unified model, simplifying architecture and reducing costs.

Model Capabilities

Mathematical Problem Solving
Code Generation
Scientific Q&A
Text Generation

Use Cases

Education
Math Competition Problem Solving
Used to solve math competition problems such as AIME, providing detailed solution steps.
Performed excellently in AIME 2024 and 2025 tests.
Programming
Code Generation and Optimization
Generates and optimizes programming code, supporting multiple programming languages.
Outperformed similar models in LiveCodeBench tests.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Ā© 2025AIbase