Edit model card

-修改为相对位置 -对内容类型进行修改


from transformers import BertTokenizer, BertModel,BertConfig


Config = BertConfig.from_pretrained("napoler/chinese_roberta_L-2_H-512_relative_key_query_token_type_100")
tokenizer = BertTokenizer.from_pretrained('napoler/chinese_roberta_L-2_H-512_relative_key_query_token_type_100')
model = BertModel.from_pretrained("napoler/chinese_roberta_L-2_H-512_relative_key_query_token_type_100",config=Config)

修改方案

https://www.kaggle.com/terrychanorg/bert-notebook9525623d9e

Downloads last month
3
Hosted inference API
Fill-Mask
Examples
Examples
Mask token: [MASK]
This model can be loaded on the Inference API on-demand.