Unable to load vocabulary from file

#7
by IgorSh - opened

hello
when i try to do
tokenizer = AutoTokenizer.from_pretrained('microsoft/deberta-v3-large')
I get error: OSError: Unable to load vocabulary from file. Please check that the provided vocabulary is accessible and not corrupted.
What I should do?

After trying to fix this problem for 3 days, I simply changed the location of the file and it worked. The space on a single file caused this error.

Hope this helps.

Thanks, but I don't understand what file you're talking about

I downloaded the model files to my PC. The file path of the model includes a folder with a space on its name. The space caused the error of "not being able to find the spm.model file. "

Sign up or log in to comment