Datasets
| Dataset | Tasks |
|---|---|
| Dataset | Tasks |
|---|---|
SWE-Atlas - Codebase QnA is a benchmark of deep codebase comprehension and QnA problems for coding agents. Checkout https://github.com/scaleapi/SWE-Atlas/ for instructions on running it.
harbor run -d scale-ai/swe-atlas-qnaA benchmark of deep codebase comprehension and QnA problems for coding agents.
See the repository for setup and usage instructions: