Ahmadzei's picture
added 3 more tables for large emb model
5fa1a76
Preprocess
For masked language modeling, the next step is to load a DistilRoBERTa tokenizer to process the text subfield:
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("distilbert/distilroberta-base")
You'll notice from the example above, the text field is actually nested inside answers.