Welcome to PyTorch-BigGraph’s documentation!¶

Contents:

  • Data model
  • From entity embeddings to edge scores
    • Embeddings
    • Global embeddings
    • Operators
    • Comparators
    • Bias
    • Coherent sets of configuration parameters
    • Interpreting the scores
  • I/O format
    • Entity and relation types
    • Entities
    • Edges
    • Checkpoint
  • Batch preparation
  • Distributed mode
    • Setup
    • Communication protocols
  • Loss calculation
    • Negative sampling
    • Loss functions
    • Optimizers
  • Evaluation
    • Offline evaluation
    • Evaluation during training
  • Dynamic relations
  • Featurized entities
  • Configuration
    • Schema
  • FAQ & Troubleshooting
    • Frequently Asked Questions
    • Common issues
  • Related works
    • TransE
    • RESCAL
    • DistMult
    • ComplEx
    • Reciprocal Relations
  • Downstream tasks
    • Parsing the output data
    • Using the embeddings
  • Pre-trained embeddings
    • Wikidata

Indices and tables¶

  • Index
  • Module Index
  • Search Page

Legal¶

  • Terms of Use
  • Privacy Policy

PyTorch-BigGraph

Navigation

Contents:

  • Data model
  • From entity embeddings to edge scores
  • I/O format
  • Batch preparation
  • Distributed mode
  • Loss calculation
  • Evaluation
  • Dynamic relations
  • Featurized entities
  • Configuration
  • FAQ & Troubleshooting
  • Related works
  • Downstream tasks
  • Pre-trained embeddings

Related Topics

  • Documentation overview
    • Next: Data model

Quick search

©2019, Facebook AI Research. | Powered by Sphinx 1.8.6 & Alabaster 0.7.12 | Page source