The model is uploaded into the serve, and the path is /u/lchen63/voca. The python interpreter to run this model will need to use "source ~/.virtualenvs/voca/bin/activate" to activate. the file foler is too large to upload since there are pretrained model inside. The file Guo_training.py is the main training file, use this file to run the main model. All data_loader, networks and model files are in the utils folder, including the original TensorFlow model. The audio_loader.py is the data_loader file, the Batcher provides every batch for training and validation. The DataHandler finishes some preprocess steps. The Guo_network.py includes the voca model's networks, and the input dimension is slightly different from their original one since the implementation, but the operation and process should be the same. The Guo_model.py file is the main model creation file.