Official original HumanEval.jsonl.gz, project README, MIT license text, and SHA-256 manifest. Please use an isolated environment before running generated code.
164 Python programming tasks. The chart below is based on actual delivered file statistics, not sample questions; the complete data is available for download after purchase.
The entry function in this original record is has_close_elements. Use the prompt to generate the function, then use the test in the same record to check the output; canonical_solution is the reference implementation and should not be used as input to the model under test. Place model-generated code in an isolated sandbox before execution.
{
"task_id": "HumanEval/0",
"entry_point": "has_close_elements",
"prompt": "from typing import List\n\n\ndef has_close_elements(numbers: List[float], threshold: float) -> bool:\n \"\"\" Check if in given list of numbers, are any two numbers closer to each other than\n given threshold.\n >>> has_close_elements([1.0, 2.0, 3.0], 0.5)\n False\n >>> has_close_elements([1.0, 2.8, 3.0, 4.0, 5.0, 2.0], 0.3)\n True\n \"\"\""
}Official original HumanEval.jsonl.gz, project README, MIT license text, and SHA-256 manifest. Please use an isolated environment before running generated code.
Official test split only; does not include a model runtime environment or automated evaluation service.
Each record includes task_id, prompt, entry_point, canonical_solution, and test.
Includes original files, source information, license text or publisher license statement, and SHA-256 manifest.
After purchase, obtain a private ZIP download link valid for 600 seconds through the platform's Dataset app. Anonymous visitors cannot access the original files.
Filename: humaneval-reviewed-attributed.zip
Size: 48,948 bytes
SHA-256: 801fb5e51f7b2e330ad7a6196346749e06558a00f31bdca1a19e51bf56417fb4
MIT. Please comply with the attribution, reposting, and applicable licensing terms specified by the original publisher.
View the original license text or publisher statement
View the original release page. The public preview displays only one verified sample from the original package and aggregate statistics; the complete data must be downloaded after purchase.
Review the delivery scope and licensing terms, then select the standard-price plan; private file downloads can only be requested after purchase is complete.