From Deterministic to Generative: Multi-Modal Stochastic RNNs for Video Captioning

Run the code

1. Train a model: python train_model.py
   
2. Test a trained model: python test.py
   

Notes

1. Please make sure the path in config.py is correct.
   
2. The structures of our trained models are listed in: model_*.py
   
3. The definition of our Stochastic layer is written in: stochastic_layer.py
   
4. Before running our codes, please make sure you installed the default deep learning libaries/framework, such as theano, lasagne, parmesa and so on.
   

=======