I am running into issues finetuning a masked language model on a custom corpus using the run_mlm.py script. The error is caused by the line of code metric = evaluate.load(“accuracy”), the full error message returns a FileNotFoundError (see below for full output). It appears that for some reason the machine my script is running on is unable to locate the accuracy metric stored in file accuracy.py · evaluate-metric/accuracy at main. Is there a flag or argument that must be set or passed to ensure that the proper location is being searched?
Thank you for any help that can be offered!
