Skip to content

Dataset · xlangai

DS-1000 code generation benchmark

xlangai/DS-1000

A simplified format of the popular benchmark with programming tasks comprising up to 10,000 examples.

What's inside

It contains text-based coding tasks (including the original fill-in-the-middle format). The dataset size ranges from 1 to 10 thousand examples. This is a reformatted and simplified version of the DS-1000 benchmark, associated with the Eval-Arena leaderboard.

What you can do with it

  • Evaluating the ability of language models to write and complete code
  • Comparing code LLMs using Eval-Arena metrics
Tasks
Text generation
Languages
Code
Size
1K–10K rows
Downloads this month
7K
Likes
39

How to load it

from datasets import load_dataset

ds = load_dataset("xlangai/DS-1000")