This repository contains a set of datasets in HVO_Sequence format as well as tools to use these datasets.
- This repository contains a number of datasets preprocessed into HVO_Sequence format. These datasets can be found in ./datasets_zipped subdirectory. Moreover, some analysis of the datasets can be found in ./datasets_analyzed/hvo_0.2.0/GrooveMIDI
Currently, the following sets are available:
- GrooveMIDI Preprocessed into HVO_Sequences (versions upto 0.3.0)
Found in ./datasets_zipped/GrooveMidi/
- ...
- A python script is also available for unzipping the datasets locally. After cloning the repository, run this script to extract datasets locally
Found in ./extract_locally.py
- A set of python classes to create multiple subsets of the complete sets with specific requirements for each subset.
Code and Documentations can be found in ./preprocessed_dataset/Subset_Creators/
