論文標題:DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter 提出機構(gòu)...
論文標題:DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter 提出機構(gòu)...
提出機構(gòu):阿里巴巴達摩院 論文鏈接:https://arxiv.org/pdf/1908.04577.pdf 作者認為Bert的預訓練任務忽略了語言結(jié)構(gòu)(language s...
機器學習講座總結(jié)-北航-互聯(lián)網(wǎng)應用下的大規(guī)模在線學習算法(四)-為什么要正則化 監(jiān)督機器學習問題無非就是“minimize your error while regulari...
<Paper Reading Series> 本文基于Facebook 2019的文章:Cross-lingual Language Model Pretraining 研究...
<Paper Reading Series> 本文基于文章:Neural Chinese Medical Named Entity Recognition with Dict...
<Paper Reading Series> 本文基于Facebook 2018-ICLR的文章:WORD TRANSLATION WITHOUT PARALLEL DATA...
<Paper Reading Series> 本文基于2018 Facebook AI Research的文章:Massively Multilingual Sentence...
<Paper Reading Series> 本文基于2017-ACL的文章:Weakly Supervised Cross-Lingual Named Entity Rec...
<Paper Reading Series> 本文基于2016 TACL的文章:Adversarial Deep Averaging Networks for Cross-L...