Turkish translation of the OpenAI HumanEval dataset.
164 Python programming problems with Turkish docstrings, translated from the original English HumanEval benchmark using claude-sonnet-4-5.
from datasets import load_dataset
ds = load_dataset("barandinho/HumanEval-TR", split="test")
task_id: Unique identifier (e.g., "HumanEval/0")
prompt: Function signature with Turkish docstring
canonical_solution:… See the full description on the dataset page:
https://huggingface.co/datasets/barandinho/HumanEval-TR.
