In the orignal BERT paper: BERT: Pre-training of Deep Bidirectional Transformers for ... Use Multilingual cased if you connect the tokenizer block to a ...
確定! 回上一頁