2019-06-19 17:49:28 +02:00
|
|
|
## MUSIC GENERATION USING DEEP LEARNING
|
|
|
|
## AUTHOR: CEZARY PUKOWNIK
|
2019-05-29 10:50:00 +02:00
|
|
|
|
2019-06-19 17:49:28 +02:00
|
|
|
## How to use:
|
2019-10-24 14:01:43 +02:00
|
|
|
1. In folder ./midi_packs make folder with midi files you want train on
|
|
|
|
2. Use extract.py to export data from midis
|
|
|
|
> ./extract.py [str: midi_pack_name] [str: name_of_session] --b [int: seq_len] -a [analize data first]
|
|
|
|
3. Use train.py to train model
|
|
|
|
> ./train.py [str: name_of_session] --b [int: batch_size] --l [int: latent_space] --e [int: epochs] --i [str: instrument] -r [reset]
|
|
|
|
4. Use generate.py to generate music from models
|
|
|
|
> ./generate.py [str: name_of_session] --n [number of generations] --m [mode {'from_seq','from_state'}]
|
2019-05-29 10:50:00 +02:00
|
|
|
|