Skip to content

Dataset · KodCode

Synthetic coding tasks with tests

KodCode/KodCode-V1-SFT-R1

Large corpus of English synthetic coding tasks with verifiable solutions and tests for SFT and RL.

What's inside

Contains 100k to 1M English examples split across 12 subsets. Fully synthetic data featuring tasks of varying difficulty, from basic exercises and library work to competitive programming and interview questions. Each task includes a verifiable solution and tests.

What you can do with it

  • SFT fine-tuning of code generation models
  • RL training of models with solution verification via autotests
  • Evaluating programming skills on tasks of varying difficulty
Tasks
Question answering
Languages
English
Size
100K–1M rows
Downloads this month
61.1K
Likes
40

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("KodCode/KodCode-V1-SFT-R1")