Keci embedding models for small benchmark datasets

This project provides embedding models for Drill. The embeddings are generated for 12 knowledge graphs Family, BioPax, Animal, Mutagenesis, Carcinogenesis, Lymphography, Nctrer, Premier League, Pyrimidine, Hepatitis, Mammographic and Suramin. Here are some more details: 12 knowledge graphs can be found here and here. The following table gives an overview of the parameters, that we use to calculate the Keci embeddings. Parameter Example Benchmark Dataset Model Keci Dimensions 32 Learning rate 0.1 Epochs 300 Batch size 32 Optimization function Adam Evaluation mode None Complete command dicee --dataset_dir KGs/example --model Keci --embedding_dim 32 --lr 0.1 --save_embeddings_as_csv --num_epochs 300 --batch_size 32 --optim Adam --scoring_technique NegSample --eval_mode None