Hey everyone, I am experiencing a rather weird situation.
Almost two months ago I built a model for my thesis (on a university server with 4 GPUs) - now I bought my own GPU and wanted to compare performance so I decided to rerun the code on my environment. The code “works” but gets stuck at the Dataloader sanity check (looks like it is exiting the train function and then restarts it once) and then freezes - has anybody experienced a similar thing before?
Any help is heavily appreciated
The code is quite a bulky mess but can be found here:
Script: Thesis_RelationExtraction_PoliticsNews/rebel.py at main · valentinwerner1/Thesis_RelationExtraction_PoliticsNews · GitHub
Full Repo: GitHub - valentinwerner1/Thesis_RelationExtraction_PoliticsNews: The repository for my master thesis
Error is posted as first comment to keep the Topic clean - thanks