revamp data preprocessing

zhvng / open-musiclm

Implementation of MusicLM, a text to music model published by Google Research, with a few modifications.

https://arxiv.org/abs/2301.11325

MIT License

522 stars 59 forks source link

Closed zhvng closed 1 year ago

zhvng commented 1 year ago

compute entire token sequences before cropping, which eliminates the influence of different context windows, and reduces redundant computation
save/load processed data from database
pass in train_config and model_config into single stage trainer