NLP/bert-utils
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

修改数据集

Browse Source
This commit is contained in:
joe 2019-07-01 10:22:40 +08:00
parent 6730f0067f
commit 91b0cc6b60
4 changed files with 120006 additions and 102479 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
README.md
View File

@ -2,10 +2,13 @@
本文基于Google开源的[BERT](https://github.com/google-research/bert)代码进行了进一步的简化，方便生成句向量与做文本分类 本文基于Google开源的[BERT](https://github.com/google-research/bert)代码进行了进一步的简化，方便生成句向量与做文本分类
---
***** New July 1st, 2019 ***** ***** New July 1st, 2019 *****
+ 修改句向量`graph`文件的生成方式，提升句向量启动速度。不再每次以临时文件的方式生成，首次执行extract_feature.py时会创建`tmp/result/graph`， + 修改句向量`graph`文件的生成方式，提升句向量启动速度。不再每次以临时文件的方式生成，首次执行extract_feature.py时会创建`tmp/result/graph`，
再次执行时直接读取该文件，如果`args.py`文件内容有修改，需要删除`tmp/result/graph`文件 再次执行时直接读取该文件，如果`args.py`文件内容有修改，需要删除`tmp/result/graph`文件
+ 修复同时启动两个进程生成句向量时代码报错的bug + 修复同时启动两个进程生成句向量时代码报错的bug
+ 修改文本匹配数据集为QA_corpus，该份数据相比于蚂蚁金服的数据更有权威性
--- ---

42479
data/dev.csv
View File

File diff suppressed because it is too large Load Diff

10001
data/test.csv Normal file
View File

File diff suppressed because it is too large Load Diff

170002
data/train.csv
View File

File diff suppressed because it is too large Load Diff