# AlexMIDIDataSet

**Repository Path**: oatmeal3000/AlexMIDIDataSet

## Basic Information

- **Project Name**: AlexMIDIDataSet
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2020-10-17
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# AlexMIDIDataSet
All of my musical works (~2000 MIDI files) in one nice Data-Set for Music AI Purposes. Please CC BY-NC-SA :) Otherwise, enjoy and have fun! :)

Benefits of using this dataset:

1) Specifically designed for Music AI development/testing/creation.
2) Multi-instrumental with drums
3) Multiple styles and many familiar melodies
4) Free sample analysis/evaluation by myself if you have any problems/questions as I know all of the dataset's compositions by heart :)
5) I do not care if you will plagiarize my work so copyright is not an issue with this set at all.

Description of the Set/Files:

1) AlexMIDIDataSet-CC-BY-NC-SA.zip ORIGINAL DATASET (~430 Multi-Instrumental MIDIs with Drums and Dynamics)
2) AlexMIDIDataSet-CC-BY-NC-SA-Piano-Only.zip Reduction from the ORIGINAL DATASET to Piano Only (~400 Piano MIDIs with Dynamics).
3) AlexMIDIDataSet-CC-BY-NC-SA-All-Drafts-Piano-Only.zip Reduction from all of my musical drafts/scores to Piano Only (~1500 Piano MIDIs with Dynamics)

Please note that many of these MIDIs were co-composed with Artificial Intelligence systems/models (MuseNet/AIVA/Magenta) and some of the MIDIs are continuations or harmonizations of well known compositions of other artists. Please excerise care with the resulting models/output and please respect the license terms because we do not want to offend anyone whos work might've been over-fitted or plagiarized by mentioned music AI systems/models.

Huge Thanks and Credits go out to the following people:

Ravi. A for assisting with the code of the reducer/parsing/prepping Piano Only Dataset :) Thanks a lot, bro! :)

1) OpenAI and Christine Payne for MuseNet Music AI model
2) Google Magenta Team for Google Magenta Piano Transformer Model
3) AIVA Team for exclusive beta access that helped to create some of the tracks in the DataSet
4) AMPER Team for free-trial of AMPER that helped to create one of the tracks in the DataSet.
5) MAESTRO Creators for your awesome Piano Dataset :)
6) Everyone else who contributed or developed the original code/AI models/AI systems :)))

Tegridy Code 2020 Aleksandr Sigalov