| DC Field | Value | Language |
|---|
| dc.contributor.author | Zhong Wu | - |
| dc.contributor.author | Rybak, V. A. | - |
| dc.coverage.spatial | Минск | en_US |
| dc.date.accessioned | 2024-08-16T06:56:23Z | - |
| dc.date.available | 2024-08-16T06:56:23Z | - |
| dc.date.issued | 2024 | - |
| dc.identifier.citation | Zhong Wu. Evaluation methods for code generation models / Zhong Wu, V. A. Rybak // Информационные технологии и управление : материалы 60-ой научной конференции аспирантов, магистрантов и студентов, Минск, 22–26 апреля 2024 года / Белорусский государственный университет информатики и радиоэлектроники ; редкол.: Л. Ю. Шилин [и др.]. – Минск, 2024. – С. 45–46. | en_US |
| dc.identifier.uri | https://libeldoc.bsuir.by/handle/123456789/56939 | - |
| dc.description.abstract | This article discusses evaluation methods for code generation models, focusing on BLEU scores and Pass@k metrics. BLEU scores are used to measure the similarity and consistency between generated code and reference code, while Pass@k metrics evaluate the pass rate of generated code on predefined test cases. Some other assessment methods are also presented.These assessment methods are important for improving and optimising code generation models. | en_US |
| dc.language.iso | en | en_US |
| dc.publisher | БГУИР | en_US |
| dc.subject | материалы конференций | en_US |
| dc.subject | optimising code | en_US |
| dc.subject | generation models | en_US |
| dc.subject | code generation models | en_US |
| dc.title | Evaluation methods for code generation models | en_US |
| dc.type | Article | en_US |
| Appears in Collections: | Информационные технологии и управление : материалы 60-й научной конференции аспирантов, магистрантов и студентов (2024)
|
