wuwuwuwuwuwuwu/testone

RoBERTa-base-ch 模型是哈工大讯飞联合实验室（HFL）开源的RoBERTa-wwm-ext的中文base 版。 RoBERTa-wwm-ext 中文模型是基于 RoBERTa 用全词Mask 方法预训练出的模型。

The RoBERTa-base-ch model is the chinese version of RoBERTa-wwm-ext which is open sourced by the Harbin Institute of Technology Xunfei Lab (HFL). RoBERTa-wwm-ext chinese model is pre-trained based on the RoBERTa model with whole word mask which is proposed by Yiming Cui Wanxiang Che Ting Liu Bing Qin Ziqing Yang.