project_in_ai

A repository containing source code for the course "Project in AI", offered on the Msc programme of engineering in Robotics at SDU.

If you want to try out the code, follow the below steps to get started.

Content

The hand-in-folder contains 3 folders and 4 files. The 3 folders are:

/bertles 		The bertles language model, trained on
			the dataset beatles.txt.
			
/experiments 		All results from the experiments, 
			described in the report.
			
/melodies		All the melody fragments generated
			randomly using shorturl.at/svwFP.

The 4 files are:

ai-env.yml		The conda environment containing all
			dependencies necessary to run the
			source code.
	
beatles.txt		A dataset containing lyrics from all 
			online available song lyrics by The 
			Beatles, with removed repititions
			and non-English words.
	
harmony.py		The source code for harmony 
			generation.
	
rhyme.py		The source code for song lyric 
			generation.

Set Up

To add the repository to a folder on your PC, you need to type the following in the terminal:

$	git clone https://github.com/lauranolling/project_in_ai.git

All dependencies are installed in the conda-environment added in the re- pository. To use the environment, you need to install Anaconda on your PC: https://docs.anaconda.com/anaconda/install/.

Then, in the terminal go to the location of the source code and type:

$ 	conda env create -f ai-env.yml 

Then, activate the environment:

$	conda activate ai-env 

RUN THE CODE

To run the source code for harmony generation, type:

$ 	python3 harmony.py  

Likewise, to run the source code for lyric generation, type:

$ 	python3 rhyme.py