ale/aiexperiments-ai-duet
1
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
6da3c5d1c5
aiexperiments-ai-duet/server/magenta/models
History
python273 3af4766358
Add Dockerfile; fix predict.py and server.py
2016-11-17 07:33:16 +03:00
..
attention_rnn Add Dockerfile; fix predict.py and server.py 2016-11-17 07:33:16 +03:00
basic_rnn Add Dockerfile; fix predict.py and server.py 2016-11-17 07:33:16 +03:00
lookback_rnn Add Dockerfile; fix predict.py and server.py 2016-11-17 07:33:16 +03:00
shared Add Dockerfile; fix predict.py and server.py 2016-11-17 07:33:16 +03:00
__init__.py Add Dockerfile; fix predict.py and server.py 2016-11-17 07:33:16 +03:00
README.md Add Dockerfile; fix predict.py and server.py 2016-11-17 07:33:16 +03:00

README.md

Models

This directory contains the various Magenta models.

Generators

All generators for NoteSequence-based models should expose their generator functionality by implementing the BaseSequenceGenerator abstract class defined in lib/sequence_generator.py. This allows all generators to communicate with a standard protocol (defined in protobuf/genator.proto) and will make it easier for various interfaces (e.g., MIDI controllers) to communicate with any model.

Note that the Melody models share a common implementation of the BaseSequenceGenerator interface in MelodyRnnSequenceGenerator, which they then invoke using the generate scripts in each model's directory (e.g., attention_rnn_generate.py).