fix(swebench): build reports after Modal evaluations

[onatozmenn]

Fixes #623

Modal returns before SWE-bench writes its aggregate report, so the wrapper tries to move a file that was never created.

When that file is missing after a Modal run, this calls SWE-bench's own make_run_report. Non-Modal runs keep the existing path. The same fallback is used by SWE-bench and SWE-bench Multilingual.

Tested with pytest tests/test_swebench_eval_infer.py (10 passed), Ruff, pycodestyle, and Pyright.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[onatozmenn]

Hi maintainers — both the Pre-commit checks and Run tests workflows are awaiting maintainer approval on this fork PR. Could someone approve the current runs so CI can execute? I’ll address any failures once results are available.