Here list all notable changes in GraphVite library.
- New model QuatE and its benchmarks on 5 knowledge graph datasets.
- Add an option to skip
faiss
in compilation. - Fix instructions for conda installation.
- New dataset
Wikidata5m
and its benchmarks, including TransE, DistMult, ComplEx, SimplE and RotatE. - Add interface for loading pretrained models and save hyperparameters.
- Add weight clip in asynchronous self-adversarial negative sampling.
- Add scalable multi-GPU prediction for node embedding and knowledge graph embedding. Evaluation on link prediction is 4.6x faster than v0.1.0.
- New demo dataset
math
and entity prediction evaluation for knowledge graph. - Support Kepler and Turing GPU architectures.
- Automatically choose the best episode size with regrad to RAM limit.
- Add template config files for applications.
- Change the update of global embeddings from average to accumulation. Fix a serious numeric problem in the update.
- Move file format settings from graph to application. Now one can customize formats and use comments in evaluation files. Add document for data format.
- Separate GPU implementation into training routines and models. Routines are in
include/instance/gpu/*
and models are ininclude/instance/model/*
.
- Multi-GPU training of large-scale graph embedding
- 3 applications: node embedding, knowledge graph embedding and graph & high-dimensional data visualization
- Node embedding
- Model: DeepWalk, LINE, node2vec
- Evaluation: node classification, link prediction
- Knowledge graph embedding
- Model: TransE, DistMult, ComplEx, SimplE, RotatE
- Evaluation: link prediction
- Graph & High-dimensional data visualization
- Model: LargeVis
- Evaluation: visualization(2D / 3D), animation(3D), hierarchy(2D)