First commit of the bert-base-japanese-char-v2 model and tokenizer. 84dca01 singletongue commited on Mar 5, 2021