Andrew Peng & David Lin
Download the Oxford 102 dataset here and the captions here
Update variable in constants.py and run train.py
- How to get text embeddings
- RNN (DONE)
- Skip thought (DONE)
- BERT (DONE)
- GPT (DONE)
- Losses (DONE)
- Get it to train (DONE)
- Put on GCP / AWS, add GPU support (DONE)
- Benchmarks for accuracy (DONE)
- Loading and saving models (DONE)
- Generate image examples (DONE)