Skip to content

Dataset · lm-provers

Olympiad problems with proofs from DeepSeek-Math-V2

lm-provers/FineProofs-SFT

7,777 examples of math olympiad problems with step-by-step reasoning and proofs from DeepSeek-Math-V2.

What's inside

Contains 7,777 English-language entries based on 4,300 unique problems from international math olympiads and the Art of Problem Solving (AoPS) community. Each problem includes thinking content and formal proofs distilled from the DeepSeek-Math-V2 model. Data is presented in tabular and text formats.

What you can do with it

  • SFT fine-tuning of models for solving complex math problems
  • Training LLMs to build rigorous step-by-step proofs
  • Evaluating the capabilities of reasoning models on olympiad math
Tasks
Text generation, Question answering
Languages
English
Size
10K–100K rows
Downloads this month
392
Likes
43

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("lm-provers/FineProofs-SFT")