Negative Sample
qiushuigongchang opened this issue · 2 comments
qiushuigongchang commented
hi, How can I get a negative sample？If I want to get negative sample，Thanks！
intfloat commented
For training, given a triple (head entity, relation, tail entity), simply replacing the head or tail entity with a random one, then you can get a negative sample.
If you mean the samples that the model fails to predict correctly, please check out our model's predictions at https://github.com/intfloat/SimKGC/tree/main/predictions
qiushuigongchang commented
hi, I'm a little confused, What's the use of triplet_mask in models.py(in line 96)?