tsunrise/everybody-compose

Issues

• Tracking Issue: Use GAN

  #22 opened 2 years ago by tsunrise
  0

• Explore better model output

  #24 opened 2 years ago by tsunrise
  0

• Better sampling method

  #25 opened 2 years ago by tsunrise
  1

• Better post processing

  #26 opened 2 years ago by violetyao
  0

• Add speed normalization

  #31 opened 2 years ago by tsunrise
  1

• visualize note embeddings

  #32 opened 2 years ago by tsunrise
  0

• Add encoder/decoder LSTM

  #28 opened 2 years ago by violetyao
  0

• `quarterLength` is not seconds

  #38 opened 2 years ago by tsunrise
  0

• Explore more losses

  #23 opened 2 years ago by tsunrise
  1

• use bi-directional LSTM

  #21 opened 2 years ago by tsunrise
  0

• Training entire dataset on AWS

  #15 opened 2 years ago by tsunrise
  0

• Add Baseline: Vanilla RNN

  #19 opened 2 years ago by tsunrise
  0

• Add postprocessing

  #27 opened 2 years ago by tsunrise
  0

• [BLOCKING] Better DataLoader

  #18 opened 2 years ago by tsunrise
  0

• Add greedy search in predict_stream

  #17 opened 2 years ago by violetyao
  1

• Add Transformer

  #20 opened 2 years ago by tsunrise
  1

• Add tensorboard logger

  #13 opened 2 years ago by violetyao
  0

• AWS Deployment

  #5 opened 2 years ago by tsunrise
  0

• Support continue from snapshot

  #10 opened 2 years ago by tsunrise
  0

• Save checkpoint with lowest validation loss

  #11 opened 2 years ago by violetyao
  0

• Add train, dev split

  #9 opened 2 years ago by tsunrise
  0

• Cache preprocessed `numpy` array, implement `Dataset`, and use `DataLoader`

  #4 opened 2 years ago by tsunrise
  0

• Add Beats Generator

  #6 opened 2 years ago by tsunrise
  0

• Possibly Incorrect return value of `generate_sequences`

  #1 opened 2 years ago by tsunrise
  1