You can choose jupyter notebook file or python file both are avaliabe.

you should place all data under audios/clips

• For jupyter notebook just simplely run hw1.ipynb
• For python file run main.py file:

  python main.py --epochs 1000 \
                 --input_len 32 \
                 --input_dim 313 \
                 --hidden_dim 32 \
                 --sr 16000 \
                 --mode train \
  

  argument: hyper-parameter settings
  • epochs : training epochs
  • input_len : sequence length
  • input_dim : input dimension
  • hidden_dim : hidden dimension
  • sr : sample rate
  • mode : choose train or test mode