/TPA-LSTM-NotOrigional

Primary LanguagePython

Pytorch Implementation of TPA-LSTM.

The code was adapted from https://github.com/jingw2/demand_forecast This implementation is different from the paper.

Requirements

Please install Pytorch before run it, and

pip install -r requirements.txt

To excute

Use either of the command to train, evaluate and get the result of the specific dataset

./run_elec.sh
./run_exchange.sh
./run_solar.sh
./run_traffic.sh

The structure of the folder

  • ./data store the downloaded dataset
  • ./model store the trained model
  • ./result store the training, evaluation and testing result
  • main.py execute the training, evaluation and testing process
  • tpaLSTM.py has the model structure