When i run the command line tira-run --input-dataset generative-ai-authorship-verification-panclef-2025/pan25-generative-ai-detection-val --image submission:latest --command ‘python ./modelo_final.py’ i get this error: Ensure that the input dataset pan25-generative-ai-detection-val is available.
Download: 78.9kiB [00:00, 2.18MiB/s]
Traceback (most recent call last):
File “/usr/bin/tira-run”, line 8, in
sys.exit(main())
File “/usr/lib/python3.8/site-packages/tira/tira_run.py”, line 412, in main
input_dir = tira.download_dataset(task, dataset)
File “/usr/lib/python3.8/site-packages/tira/rest_api_client.py”, line 675, in download_dataset
self.download_and_extract_zip_with_md5(url, target_dir + suffix, expected_md5, subdirectory, rename_to)
File “/usr/lib/python3.8/site-packages/tira/rest_api_client.py”, line 778, in download_and_extract_zip_with_md5
raise ValueError(
ValueError: MD5 is unexpected: I expected “fd12cbb06a882276278655acc949b91d” but got “8d73aaf98a42cb74bc6d97eadb5ad3c6” for URL “PAN'25 Generative AI Detection (Task 1): Voight-Kampff AI Detection Sensitivity”. Could it be that my token is not right ?
Hi, the problem is that the validation dataset is not public anymore.
Could you please use the --dataset pan25-generative-ai-detection-smoke-test-20250428-training?
This is described in the baseline for the task and should work: pan-code/clef25/generative-authorship-verification/pan25_genai_baselines at master · pan-webis-de/pan-code · GitHub
Best regards,
Maik