tristan-jl / music-lstm

Melody generation using an LSTM

Music LSTM

Music LSTM generates a melody, given a seed, using a PyTorch LSTM. Inspired by this series of videos. Used Google Colab GPUs to train model.

Uses data from the ESAC Database.

Model Architecture

LSTM(
(lstm): LSTM(45, 256, batch_first=True)
(drop): Dropout(p=0.5, inplace=False)
(fc): Linear(in_features=256, out_features=45, bias=True)
)

License

MIT