Hello LISA 2026 organizers, I am currently experiencing a repeated failure when trying to evaluate my Task 1b submission during the validation phase. From the latest log, the workflow successfully reaches the `score.cwl` step and launches the Task 1b scoring container: `Running scoring container with GPU: lisa2026 ['--task1b', ...]` However, the scoring script is killed before producing the expected output file: `/app/entrypoint.sh: line 22: 94 Killed python /app/scripts/task1b.py ...` and the workflow then fails with: `ERROR: results.json was not produced by the scoring container.` `Did not find output file with glob pattern: ['results.json']` I also noticed the following warning in the same job log: `Starting job ... with less than 10% of disk space remaining.` In addition, the scoring container downloads extra model weights during the scoring step, including the AlexNet weights and BRISQUE SVM weights, before the process is killed. Since there is no clear error message indicating a problem with my prediction files, such as an invalid file structure, missing prediction, or shape mismatch, I am not sure whether this is caused by my submission or by an infrastructure/resource issue on the evaluator side. The validation phase is ending in about one week, and I am concerned because I have not been able to obtain a single successful validation result for Task 1b due to this repeated scoring failure. Could you please investigate this as soon as possible and let me know whether the issue is on the evaluator/infrastructure side or with my submitted files? If the problem is caused by my submitted prediction files, I would really appreciate any details so that I can fix the submission before the validation phase closes. Thank you very much for your help.