amazon-science/recode

Releasing code for "ReCode: Robustness Evaluation of Code Generation Models"

35
/ 100
Emerging

This project helps evaluate how well code-generating AI models perform when presented with slightly altered inputs. It takes your existing code generation models and datasets, applies subtle changes to docstrings, function names, or code syntax, and then measures the model's ability to still produce correct code. This tool is for AI researchers and engineers who develop or deploy code generation models and need to understand their reliability.

No commits in the last 6 months.

Use this if you need to thoroughly test the practical robustness of your code generation models against common, subtle variations in input.

Not ideal if you are looking to test general code quality, functional correctness, or performance of your models under normal, unperturbed conditions.

AI-model-evaluation code-generation-AI model-robustness natural-language-processing software-engineering-AI
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 8 / 25
Maturity 16 / 25
Community 11 / 25

How are scores calculated?

Stars

58

Forks

6

Language

Python

License

Apache-2.0

Last pushed

Mar 20, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/amazon-science/recode"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.