I’m having what seems to be permission issues with huggingface.
I’m training/finetuning a DistilBERT model in a notebook, the first epoch trains fine (and can be used) but as soon as the second epoch finishes I get a bunch of ‘notadirectory’ and permission errors. See log pastebin below.
The code is directly from Orilley’s transformers for NLP textbook so I am convinced that the code is not the issue… but maybe i’m being naive. Issue is also the same even after restarting PC and making sure nothing else has the file open
Logs look like it’s something up with huggingface_hub library.
Any help would be appreciated!