Go-UT-Bench: A Fine-Tuning Dataset for LLM-Based Unit Test Generation in Go
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866914615084974080 |
|---|---|
| author | Pipalani, Yashshi Raj, Hritik Ghosh, Rajat Bhargava, Vaishnavi Dutta, Debojyoti |
| author_facet | Pipalani, Yashshi Raj, Hritik Ghosh, Rajat Bhargava, Vaishnavi Dutta, Debojyoti |
| contents | Training data imbalance poses a major challenge for code LLMs. Most available data heavily over represents raw opensource code while underrepresenting broader software engineering tasks, especially in low resource languages like Golang. As a result, models excel at code autocompletion but struggle with real world developer workflows such as unit test generation. To address this gap, we introduce GO UT Bench, a benchmark dataset of 5264 pairs of code and unit tests, drawn from 10 permissively licensed Golang repositories spanning diverse domain. We evaluate its effectiveness as a fine tuning dataset across two LLM families i.e. mixture of experts and dense decoders. Our results show that finetuned models outperform their base counterparts on more than 75% of benchmark tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_10868 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Go-UT-Bench: A Fine-Tuning Dataset for LLM-Based Unit Test Generation in Go Pipalani, Yashshi Raj, Hritik Ghosh, Rajat Bhargava, Vaishnavi Dutta, Debojyoti Machine Learning E.0 Training data imbalance poses a major challenge for code LLMs. Most available data heavily over represents raw opensource code while underrepresenting broader software engineering tasks, especially in low resource languages like Golang. As a result, models excel at code autocompletion but struggle with real world developer workflows such as unit test generation. To address this gap, we introduce GO UT Bench, a benchmark dataset of 5264 pairs of code and unit tests, drawn from 10 permissively licensed Golang repositories spanning diverse domain. We evaluate its effectiveness as a fine tuning dataset across two LLM families i.e. mixture of experts and dense decoders. Our results show that finetuned models outperform their base counterparts on more than 75% of benchmark tasks. |
| title | Go-UT-Bench: A Fine-Tuning Dataset for LLM-Based Unit Test Generation in Go |
| topic | Machine Learning E.0 |
| url | https://arxiv.org/abs/2511.10868 |