natas_hw
/
sempre
mirror of https://github.com/percyliang/sempre
0
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Added the run command to README

This commit is contained in:
Panupong Pasupat 2017-09-04 15:39:42 -07:00
parent 1fc5b82030
commit f9807e8fec
2 changed files with 17 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/edu/stanford/nlp/sempre/tables/features/PhraseInfo.java
View File

@ -25,7 +25,7 @@ public class PhraseInfo {
@Option(gloss = "Fuzzy match predicates")
public boolean computeFuzzyMatchPredicates = false;
@Option(gloss = "Do not produce lexicalized features if the phrase begins or ends with a stop word")
public boolean forbidBorderStopWordInLexicalizedFeatures = false;
public boolean forbidBorderStopWordInLexicalizedFeatures = true;
}
public static Options opts = new Options();

16
tables/README.md
View File

@ -41,6 +41,22 @@ Running the code
Other available sets include `u-2`, ..., `u-5` (four other development splits)
and `test` (actual train-test split).
Other options
-------------
### Macro Grammar (Experimental)
Macro grammar can be used to significantly speed up the parser.
To turn on macro grammar, run the following:
./run @mode=tables @data=u-1 @feat=more @parser=cprune @grammar=extended @fuzzy=editdist-fuzzy @train=1
Please refer to the following paper for more information:
> Yuchen Zhang, Panupong Pasupat, Percy Liang.
> Macro Grammars and Holistic Triggering for Efficient Semantic Parsing
> Empirical Methods on Natural Language Processing (EMNLP), 2017.
Official evaluation
-------------------