fix double save results in vf-eval

[mikasenghaas]

Description

previously we would redundantly save the results in generate and run_evaluation (just once also pushing to HF hub). this makes it so that generate is entirely responsible for the saving and pushing to hub to avid unnecessary redundant disk writes.

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Test improvement

Testing

• All existing tests pass when running uv run pytest locally.
• New tests have been added to cover the changes

Checklist

• My code follows the style guidelines of this project as outlined in AGENTS.md
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• Any dependent changes have been merged and published

Additional Notes

---

Note

Fixes redundant saving of evaluation results.

• Removes the post-evaluation save_rollout_results call in run_evaluation so results are saved only by vf_env.evaluate when save_results is enabled
• No changes to evaluation flow, callbacks, or output formatting

^{Written by Cursor Bugbot for commit cc3af89. This will update automatically on new commits. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Bugbot Autofix is OFF. To automatically fix reported issues with Cloud Agents, enable Autofix in the Cursor dashboard.}