Skip to content

Dataset · MathLLMs

Dataset for solving math problems via code

MathLLMs/MathCodeInstruct-Plus

English text data for training LLMs in mathematical reasoning with programming code integration.

What's inside

The dataset is associated with the MathCoder open model family (based on Llama-2 and Code Llama) built for solving math problems. The corpus contains 10 to 100 thousand English text entries. Data is intended for text generation and question-answering tasks.

What you can do with it

  • SFT fine-tuning of language models to solve math using code generation
  • Training LoRA adapters to enhance LLM logical and computational skills
  • Testing model capabilities to combine text reasoning and programming
Tasks
Question answering, Text generation
Languages
English
Size
10K–100K rows
Downloads this month
528
Likes
19

How to load it

from datasets import load_dataset

ds = load_dataset("MathLLMs/MathCodeInstruct-Plus")