AmpliGraph

AmpliGraph copied to clipboard

Migrate to tensor flow 2

2

Some environments and libraries are now working with tf 2. It is worth to upgrade Ampligraph to run over TF 2.X as described in https://www.tensorflow.org/guide/migrate/migrate_tf2

rdelaguila

Implement RotatE

10

**Description** Implement *RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space* https://openreview.net/forum?id=HkgEQnRqYQ

lukostaz

### Description The function generate_candidates(X, strategy='graph_degree', target_rel='y', max_candidates=3) generates different output then shown in the example in the file, which is not correct as it is generating 3x the same...

smejkka3

When training ConvE with early stopping, the following error gets thrown: Traceback (most recent call last): File "ConvE(1-n)_fb15k.py", line 47, in early_stopping_params=early_stopping_params File "/home/ramab/anaconda3/envs/ampligraph/lib/python3.7/site-packages/ampligraph/latent_features/models/ConvE.py", line 753, in fit raise e...

ramesesz

'ConvKB' object has no attribute 'corr_batch_size'

1

### Description After completion of training ConvKB model i tried to predict the score of the triples using the following lines of code : ``` # predict the score of...

WarisBunglawala

### Description The spelling mistake in `ampligraph/latent_features/models/EmbeddingModel.py` or documentation causes the the strategy of generating corruptions is allways `default`('s,o'). ### Actual Behavior I want to build a `TransE` model, when...

Faker-lz

https://github.com/Accenture/AmpliGraph/blob/d4bf44559cb7178039f21203780be4eb946ea4eb/experiments/IJCAI-21/experiments.py#L12

1

There is an obvious bug, how can you get the experimental results in your IJCAI paper with this wrong code? def compare_vanilla_focusE_models(dataset, model_name, vanilla_params, focusE_params): if dataset == 'onet20k': X...

Diison

Continuous Training / Model tuning / Initialize with pre-trained Model

2

I want to continue/resume training a pre-trained model using new triples. We may assume that there are no new entities or relations but new instants. What is the best way...

mhmgad

Hi How to train GraphSAGE model for prediction tasks, like **TransE, RotateE, ComplEx models** etc. **For example:** After training GraphSAGE on KG predict **head** or **tail** entities using `query_topn` Thanks

geo47

Is query_topn computes as same as evaluate_performance

1

### Description I'm trying to evaluate the recommendation result of a tool. I have trained a model using ampligraph's model, however, the problem appears when I tried to evaluate the...

SuShu19