Evaluation script working locally but not on server

[mariiak2021]

Hi,

I've followed the instructions to submit the challenge for EvalAI and evaluation script is working fine locally, but still has some errors on the server side:
"/tmp/tmp8biownt_/compute/challenge_data/challenge_2325/__init__.py", line 49, in <module> from .main import evaluate File "/tmp/tmp8biownt_/compute/challenge_data/challenge_2325/main.py", line 5, in <module> from phase_1.main import evaluate_accuracy ModuleNotFoundError: No module named 'phase_1'
What can we the problem? I even don't have these lines in the code now so I don't understand why does it use the old info.

Best,
Mariia

[JannikSchneider12]

Hey,

I have similar issues. I am getting errors in the worker logs from the web interface that state that a module can not be found although I uncluded it in the requirements and it worked locally. Did you find a solution?

[paularamo]

@mariiak2021 Did you find a solution for that? I am facing the same, and I cannot see any answer on the server. How could we get a faster response from the repo maintainers? Maybe @taranjeet @dexter1691 @deshraj @krtkvrm. Could you please point me to a solution? or maybe a person can help me with that? We will receive submissions once I launch the evaluation dataset tomorrow; I would like to see a real-time leaderboard. That was the main reason I selected Eval.AI over other platforms. Appreciate your help on this.