salesforce/MultiHopKG
Multi-hop knowledge graph reasoning learned via policy gradient with reward shaping and action dropout
Jupyter NotebookBSD-3-Clause
Issues
- 17
- 1
NELL-995 dataset
#20 opened by chrislouis0106 - 2
- 1
How to get the "raw.pgrk" file in the dataset?
#25 opened by victorup - 5
- 1
- 0
- 2
the meaning of NELL split
#22 opened by chrislouis0106 - 1
check the dev performance
#27 opened by wj1108114106 - 3
Some doubts about result
#26 opened by xlk369293141 - 0
About the code in learn_framework.py
#24 opened by victorup - 1
model:point.gc and gc
#23 opened by chrislouis0106 - 1
about the function of nell split
#21 opened by chrislouis0106 - 2
about the NELL dataset
#19 opened by chrislouis0106 - 0
dataset
#18 opened by chrislouis0106 - 0
Relation and entity dimension
#15 opened by msaebi1993 - 6
Problem with output model files
#14 opened - 2
About "margin" and "num_negative_samples" hyperparames when training embedding based models
#13 opened by David-Lee-1990 - 14
add Action Dropout on minerva WN18RR
#12 opened by David-Lee-1990 - 8
- 10
cannot replicate NELL-995 MINERVA result
#9 opened by gstoica27 - 0
NELL-995 bad data statistics
#10 opened by David-Lee-1990 - 1
Question about the input
#8 opened by BeerTai - 4
Can not reproduce the result on kinship dataset
#7 opened by Lee-zix - 1
about entity2id.txt
#6 opened by BeerTai - 1
Find a comment typo in pg.py
#5 opened by XCorder - 2
How to prepare page ranks?
#4 opened by Toni-Chan - 4
It seems that this code doesn't use negative triples to train embedding models.
#3 opened by davidlvxin - 1
A bug in pn.py
#2 opened by davidlvxin