| # Contributing |
|
|
| Contributions that improve reproducibility, portability, evaluation, calibration, |
| or documentation are welcome. |
|
|
| Please keep changes auditable: |
|
|
| 1. Explain the experimental claim being changed. |
| 2. Include the exact model/data/code version used. |
| 3. Add or update tests when changing loading or scoring behavior. |
| 4. Do not replace measured results with estimates. |
| 5. Keep public benchmark rows out of training and checkpoint selection. |
| 6. Report regressions and failed experiments instead of deleting them from the record. |
|
|
| For model-quality changes, include row-level or task-level evidence sufficient to |
| compare against the frozen public 1.0 baseline. |
|
|
