Skip to content

Dataset · IFM

Data for reasoning code generation

IFM/Code-Reasoning

Coding tasks with reasoning traces, direct answers, and task synthesis in parquet.

What's inside

The dataset contains data for solving coding tasks, including subsets with reasoning traces, direct answers, and task synthesis. It is part of the K2 Horizon collection; each subset has a train split in parquet shards, supporting browsing and streaming.

What you can do with it

  • Training code models to reason before answering
  • Comparing reasoning and direct-answer modes
  • Generating synthetic tasks for code SFT
Tasks
Text generation
Size
100M–1B rows
Downloads this month
43.9K
Likes
105

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("IFM/Code-Reasoning")