AceReason-1.1-SFT Mathematical Code Reasoning Dataset
Date
Organization
Publish URL
AceReason-1.1-SFT is a diverse and high-quality supervised fine-tuning (SFT) dataset released by NVIDIA in 2025, focusing on mathematical and code reasoning. The related paper results are:AceReason-Nemotron 1.1: Advancing Math and Code Reasoning through SFT and RL Synergy", which aims to train SFT models that focus on mathematical and code reasoning.
This dataset serves as a mathematical and code reasoning model AceReason-Nemotron-1.1-7B SFT training data of , all answers in the dataset are generated by DeepSeek-R1.
The AceReason-1.1-SFT dataset contains 2,668,741 math samples and 1,301,591 code samples, covering data from OpenMathReasoning, NuminaMath-CoT, OpenCodeReasoning, MagicoderEvolInstruct, opc-sft-stage2, leetcode, TACO, and apps. The dataset is cleaned and samples with 9-gram overlap with any test samples in math and coding benchmarks are filtered.