Skip to content

Dataset · PrimeIntellect

Verified Python code fixing tasks

PrimeIntellect/Scale-SWE-Verified

17,202 cleaned tasks for solving real Python bugs with validated patches.

What's inside

This is a filtered English version of the Scale-SWE dataset containing 17,202 tasks for fixing issues in Python code. The authors removed nearly 15% of the original records (broken environments, non-working reference patches) to provide a clean end-to-end reward signal.

What you can do with it

  • Training code agents using reinforcement learning (RL)
  • SFT fine-tuning of models for automatic bug fixing in Python
  • Evaluating patch generation quality in software engineering
Tasks
Text generation
Languages
English
Size
10K–100K rows
Downloads this month
2.1K
Likes
6

How to load it

from datasets import load_dataset

ds = load_dataset("PrimeIntellect/Scale-SWE-Verified")