code for paper "Boosting Entity-aware Image Captioning with Multi-modal Knowledge Graph"

• implemented using PyTorch and Fairseq, requires:
  • PyTorch 1.8.1
  • Fairseq 0.10.1
• working directory is src, use src/main/train.py to train model, and src/main/generate.py to generate captions
• data is in data directory