Convert Figma logo to code with AI

openai logojukebox

Code for the paper "Jukebox: A Generative Model for Music"

7,798
1,403
7,798
199

Top Related Projects

19,130

Magenta: Music and Art Generation with Machine Intelligence

Audiocraft is a library for audio processing and generation with deep learning. It features the state-of-the-art EnCodec audio compressor / tokenizer, along with MusicGen, a simple and controllable music generation LM with textual and melodic conditioning.

Implementation of MusicLM, Google's new SOTA model for music generation using attention networks, in Pytorch

4,498

Muzic: Music Understanding and Generation with Artificial Intelligence

1,826

An AI for Music Generation

Quick Overview

Jukebox is an open-source AI model developed by OpenAI for generating music, including raw audio waveforms with singing in various genres and styles. It uses a neural network to create novel songs with coherent melodies, lyrics, and vocals, pushing the boundaries of AI-generated music.

Pros

  • Generates high-quality, diverse music across multiple genres
  • Produces raw audio waveforms, including realistic vocals
  • Allows for fine-tuning and style transfer capabilities
  • Open-source, enabling further research and development in AI music generation

Cons

  • Requires significant computational resources for training and generation
  • Output quality can be inconsistent, with occasional artifacts or incoherence
  • Limited control over specific musical elements or structures
  • Potential copyright and ethical concerns regarding AI-generated music

Code Examples

# Load a pre-trained Jukebox model
vqvae, prior, _ = MODELS['5b']
model = load_model(vqvae, prior, 'cpu')

# Generate a new song
lyrics = "I'm walking on sunshine, woah"
artist = "Pop Star"
genre = "pop"
sample_rate = 44100
duration = 30  # seconds

metas = [dict(artist=artist, genre=genre, lyrics=lyrics)]
sample = generate_samples(model, metas, sample_rate, duration)
# Apply style transfer to an existing song
source_path = "path/to/source/audio.wav"
target_path = "path/to/target/audio.wav"

source = load_audio(source_path)
target = load_audio(target_path)

transferred = style_transfer(model, source, target)
save_audio(transferred, "output.wav")
# Fine-tune the model on custom data
dataset = load_custom_dataset("path/to/custom/data")
fine_tuned_model = fine_tune(model, dataset, epochs=10)

# Generate new samples using the fine-tuned model
new_samples = generate_samples(fine_tuned_model, metas, sample_rate, duration)

Getting Started

  1. Clone the repository:

    git clone https://github.com/openai/jukebox.git
    cd jukebox
    
  2. Install dependencies:

    pip install -r requirements.txt
    
  3. Download pre-trained models:

    from jukebox.utils.remote_utils import download_models
    download_models()
    
  4. Generate a sample:

    from jukebox.sample import sample_single_window
    from jukebox.utils.io import load_model
    
    vqvae, prior, _ = MODELS['5b']
    model = load_model(vqvae, prior, 'cpu')
    sample = sample_single_window(model, ['Pop Star'], ['pop'], ['Happy birthday to you'])
    

Competitor Comparisons

19,130

Magenta: Music and Art Generation with Machine Intelligence

Pros of Magenta

  • Broader scope, covering various music and art generation tasks
  • More active development and community support
  • Extensive documentation and tutorials for easier adoption

Cons of Magenta

  • Less focused on high-quality audio generation compared to Jukebox
  • May require more setup and configuration for specific tasks
  • Potentially steeper learning curve due to its wide range of features

Code Comparison

Magenta (TensorFlow-based):

import magenta.music as mm
from magenta.models.melody_rnn import melody_rnn_sequence_generator

generator = melody_rnn_sequence_generator.get_generator()
melody = generator.generate(steps_per_quarter=4, total_steps=32)

Jukebox (PyTorch-based):

import jukebox
from jukebox.sample import sample_partial_window

model = jukebox.load_model()
samples = sample_partial_window(model, prompt=None, sample_length=1024)

Both repositories focus on AI-driven music generation, but Magenta offers a more comprehensive toolkit for various creative tasks, while Jukebox specializes in high-quality audio generation. Magenta's code tends to be more modular and flexible, whereas Jukebox's approach is more streamlined for its specific use case. The choice between them depends on the user's specific requirements and level of expertise in machine learning and music generation.

Audiocraft is a library for audio processing and generation with deep learning. It features the state-of-the-art EnCodec audio compressor / tokenizer, along with MusicGen, a simple and controllable music generation LM with textual and melodic conditioning.

Pros of AudioCraft

  • More recent and actively maintained project (last update in 2023)
  • Supports multiple audio generation tasks: music, sound effects, and speech
  • Offers pre-trained models for immediate use

Cons of AudioCraft

  • Limited to 30-second audio clips
  • Requires more computational resources for training and inference

Code Comparison

AudioCraft:

import torch
from audiocraft.models import MusicGen

model = MusicGen.get_pretrained('medium')
wav = model.generate_unconditional(4, progress=True)

Jukebox:

import jukebox
from jukebox.sample import sample_partial_window

model = jukebox.make_model(5, 'small')
sample = sample_partial_window(model, ...)

Key Differences

  • AudioCraft focuses on shorter, high-quality audio generation across multiple domains
  • Jukebox specializes in long-form music generation with lyrics
  • AudioCraft offers easier implementation and use of pre-trained models
  • Jukebox provides more control over the generated music structure and style

Both projects contribute significantly to AI-driven audio generation, with AudioCraft being more accessible for general audio tasks and Jukebox excelling in complex music composition.

Implementation of MusicLM, Google's new SOTA model for music generation using attention networks, in Pytorch

Pros of MusicLM-pytorch

  • Implements the more recent MusicLM model, potentially offering improved audio generation capabilities
  • Written in PyTorch, which may be more accessible for some developers
  • Actively maintained with recent updates

Cons of MusicLM-pytorch

  • Less comprehensive documentation compared to Jukebox
  • Smaller community and fewer resources available
  • May require more computational resources for training and generation

Code Comparison

MusicLM-pytorch:

music = MusicLM(
    dim = 512,
    depth = 6,
    heads = 8,
    dim_head = 64,
    t5_name = 't5-base'
)

Jukebox:

vqvae = VQVAE(
    input_shape=(6144,),
    nb_levels=3,
    downsampling_factors=(2, 2, 2),
    nb_codebooks=2,
    codebook_size=2048,
    codebook_dim=64,
)

Both repositories provide implementations of advanced music generation models. Jukebox, developed by OpenAI, offers a more established and well-documented solution with a larger community. MusicLM-pytorch, on the other hand, implements a newer model and may offer improved generation capabilities, but with less comprehensive documentation and community support. The choice between the two depends on specific project requirements, familiarity with the frameworks, and desired model capabilities.

4,498

Muzic: Music Understanding and Generation with Artificial Intelligence

Pros of Muzic

  • Broader scope, covering various music AI tasks beyond generation
  • More active development and recent updates
  • Includes pre-trained models and datasets for easier experimentation

Cons of Muzic

  • Less focused on high-quality audio generation compared to Jukebox
  • May require more setup and configuration for specific tasks
  • Documentation could be more comprehensive for some components

Code Comparison

Jukebox (PyTorch):

vqvae = VQVAE(input_shape, levels, downs_t, strides_t, emb_width, l_bins)
prior = SimplePrior(prior_shape, bins, width, depth, heads, blocks)

Muzic (PyTorch):

model = MusicTransformer(
    embedding_dim=64,
    num_layers=6,
    num_heads=8,
    vocab_size=388,
    dropout=0.1,
    max_seq=2048
)

Both projects use PyTorch and implement transformer-based architectures, but Jukebox focuses on VQ-VAE and hierarchical priors, while Muzic offers a more general-purpose music transformer model.

1,826

An AI for Music Generation

Pros of MuseGAN

  • Focuses specifically on multi-track music generation
  • Provides a more lightweight and accessible implementation
  • Offers pre-trained models for quick experimentation

Cons of MuseGAN

  • Limited to MIDI-based music generation
  • Less advanced in terms of raw audio synthesis capabilities
  • Smaller scale project with potentially less ongoing development

Code Comparison

MuseGAN sample code:

from musegan.model import MuseGAN
model = MuseGAN()
generated_music = model.generate(n_bars=4, temperature=1.0)

Jukebox sample code:

import jukebox
model = jukebox.make_prior(hps)
samples, _, _ = model.sample(hps)
jukebox.save_wav(samples, hps)

Summary

MuseGAN is a more focused project for multi-track MIDI music generation, offering a simpler implementation and pre-trained models. However, it lacks the advanced audio synthesis capabilities of Jukebox. MuseGAN is better suited for those specifically interested in MIDI-based music generation, while Jukebox offers more comprehensive audio generation capabilities but with increased complexity.

Convert Figma logo designs to code with AI

Visual Copilot

Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.

Try Visual Copilot

README

Status: Archive (code is provided as-is, no updates expected)

Jukebox

Code for "Jukebox: A Generative Model for Music"

Paper Blog Explorer Colab

Install

Install the conda package manager from https://docs.conda.io/en/latest/miniconda.html

# Required: Sampling
conda create --name jukebox python=3.7.5
conda activate jukebox
conda install mpi4py=3.0.3 # if this fails, try: pip install mpi4py==3.0.3
conda install pytorch=1.4 torchvision=0.5 cudatoolkit=10.0 -c pytorch
git clone https://github.com/openai/jukebox.git
cd jukebox
pip install -r requirements.txt
pip install -e .

# Required: Training
conda install av=7.0.01 -c conda-forge 
pip install ./tensorboardX
 
# Optional: Apex for faster training with fused_adam
conda install pytorch=1.1 torchvision=0.3 cudatoolkit=10.0 -c pytorch
pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./apex

Sampling

Sampling from scratch

To sample normally, run the following command. Model can be 5b, 5b_lyrics, 1b_lyrics

python jukebox/sample.py --model=5b_lyrics --name=sample_5b --levels=3 --sample_length_in_seconds=20 \
--total_sample_length_in_seconds=180 --sr=44100 --n_samples=6 --hop_fraction=0.5,0.5,0.125
python jukebox/sample.py --model=1b_lyrics --name=sample_1b --levels=3 --sample_length_in_seconds=20 \
--total_sample_length_in_seconds=180 --sr=44100 --n_samples=16 --hop_fraction=0.5,0.5,0.125

The above generates the first sample_length_in_seconds seconds of audio from a song of total length total_sample_length_in_seconds. To use multiple GPU's, launch the above scripts as mpiexec -n {ngpus} python jukebox/sample.py ... so they use {ngpus}

The samples decoded from each level are stored in {name}/level_{level}. You can also view the samples as an html with the aligned lyrics under {name}/level_{level}/index.html. Run python -m http.server and open the html through the server to see the lyrics animate as the song plays.
A summary of all sampling data including zs, x, labels and sampling_kwargs is stored in {name}/level_{level}/data.pth.tar.

The hps are for a V100 GPU with 16 GB GPU memory. The 1b_lyrics, 5b, and 5b_lyrics top-level priors take up 3.8 GB, 10.3 GB, and 11.5 GB, respectively. The peak memory usage to store transformer key, value cache is about 400 MB for 1b_lyrics and 1 GB for 5b_lyrics per sample. If you are having trouble with CUDA OOM issues, try 1b_lyrics or decrease max_batch_size in sample.py, and --n_samples in the script call.

On a V100, it takes about 3 hrs to fully sample 20 seconds of music. Since this is a long time, it is recommended to use n_samples > 1 so you can generate as many samples as possible in parallel. The 1B lyrics and upsamplers can process 16 samples at a time, while 5B can fit only up to 3. Since the vast majority of time is spent on upsampling, we recommend using a multiple of 3 less than 16 like --n_samples 15 for 5b_lyrics. This will make the top-level generate samples in groups of three while upsampling is done in one pass.

To continue sampling from already generated codes for a longer duration, you can run

python jukebox/sample.py --model=5b_lyrics --name=sample_5b --levels=3 --mode=continue \
--codes_file=sample_5b/level_0/data.pth.tar --sample_length_in_seconds=40 --total_sample_length_in_seconds=180 \
--sr=44100 --n_samples=6 --hop_fraction=0.5,0.5,0.125

Here, we take the 20 seconds samples saved from the first sampling run at sample_5b/level_0/data.pth.tar and continue by adding 20 more seconds.

You could also continue directly from the level 2 saved outputs, just pass --codes_file=sample_5b/level_2/data.pth.tar. Note this will upsample the full 40 seconds song at the end.

If you stopped sampling at only the first level and want to upsample the saved codes, you can run

python jukebox/sample.py --model=5b_lyrics --name=sample_5b --levels=3 --mode=upsample \
--codes_file=sample_5b/level_2/data.pth.tar --sample_length_in_seconds=20 --total_sample_length_in_seconds=180 \
--sr=44100 --n_samples=6 --hop_fraction=0.5,0.5,0.125

Here, we take the 20 seconds samples saved from the first sampling run at sample_5b/level_2/data.pth.tar and upsample the lower two levels.

Prompt with your own music

If you want to prompt the model with your own creative piece or any other music, first save them as wave files and run

python jukebox/sample.py --model=5b_lyrics --name=sample_5b_prompted --levels=3 --mode=primed \
--audio_file=path/to/recording.wav,awesome-mix.wav,fav-song.wav,etc.wav --prompt_length_in_seconds=12 \
--sample_length_in_seconds=20 --total_sample_length_in_seconds=180 --sr=44100 --n_samples=6 --hop_fraction=0.5,0.5,0.125

This will load the four files, tile them to fill up to n_samples batch size, and prime the model with the first prompt_length_in_seconds seconds.

Training

VQVAE

To train a small vqvae, run

mpiexec -n {ngpus} python jukebox/train.py --hps=small_vqvae --name=small_vqvae --sample_length=262144 --bs=4 \
--audio_files_dir={audio_files_dir} --labels=False --train --aug_shift --aug_blend

Here, {audio_files_dir} is the directory in which you can put the audio files for your dataset, and {ngpus} is number of GPU's you want to use to train. The above trains a two-level VQ-VAE with downs_t = (5,3), and strides_t = (2, 2) meaning we downsample the audio by 2**5 = 32 to get the first level of codes, and 2**8 = 256 to get the second level codes.
Checkpoints are stored in the logs folder. You can monitor the training by running Tensorboard

tensorboard --logdir logs

Prior

Train prior or upsamplers

Once the VQ-VAE is trained, we can restore it from its saved checkpoint and train priors on the learnt codes. To train the top-level prior, we can run

mpiexec -n {ngpus} python jukebox/train.py --hps=small_vqvae,small_prior,all_fp16,cpu_ema --name=small_prior \
--sample_length=2097152 --bs=4 --audio_files_dir={audio_files_dir} --labels=False --train --test --aug_shift --aug_blend \
--restore_vqvae=logs/small_vqvae/checkpoint_latest.pth.tar --prior --levels=2 --level=1 --weight_decay=0.01 --save_iters=1000

To train the upsampler, we can run

mpiexec -n {ngpus} python jukebox/train.py --hps=small_vqvae,small_upsampler,all_fp16,cpu_ema --name=small_upsampler \
--sample_length=262144 --bs=4 --audio_files_dir={audio_files_dir} --labels=False --train --test --aug_shift --aug_blend \
--restore_vqvae=logs/small_vqvae/checkpoint_latest.pth.tar --prior --levels=2 --level=0 --weight_decay=0.01 --save_iters=1000

We pass sample_length = n_ctx * downsample_of_level so that after downsampling the tokens match the n_ctx of the prior hps. Here, n_ctx = 8192 and downsamples = (32, 256), giving sample_lengths = (8192 * 32, 8192 * 256) = (65536, 2097152) respectively for the bottom and top level.

Learning rate annealing

To get the best sample quality anneal the learning rate to 0 near the end of training. To do so, continue training from the latest checkpoint and run with

--restore_prior="path/to/checkpoint" --lr_use_linear_decay --lr_start_linear_decay={already_trained_steps} --lr_decay={decay_steps_as_needed}

Reuse pre-trained VQ-VAE and train top-level prior on new dataset from scratch.

Train without labels

Our pre-trained VQ-VAE can produce compressed codes for a wide variety of genres of music, and the pre-trained upsamplers can upsample them back to audio that sound very similar to the original audio. To re-use these for a new dataset of your choice, you can retrain just the top-level

To train top-level on a new dataset, run

mpiexec -n {ngpus} python jukebox/train.py --hps=vqvae,small_prior,all_fp16,cpu_ema --name=pretrained_vqvae_small_prior \
--sample_length=1048576 --bs=4 --aug_shift --aug_blend --audio_files_dir={audio_files_dir} \
--labels=False --train --test --prior --levels=3 --level=2 --weight_decay=0.01 --save_iters=1000

Training the small_prior with a batch size of 2, 4, and 8 requires 6.7 GB, 9.3 GB, and 15.8 GB of GPU memory, respectively. A few days to a week of training typically yields reasonable samples when the dataset is homogeneous (e.g. all piano pieces, songs of the same style, etc).

Near the end of training, follow this to anneal the learning rate to 0

Sample from new model

You can then run sample.py with the top-level of our models replaced by your new model. To do so,

  • Add an entry my_model=("vqvae", "upsampler_level_0", "upsampler_level_1", "small_prior") in MODELS in make_models.py.
  • Update the small_prior dictionary in hparams.py to include restore_prior='path/to/checkpoint'. If you you changed any hps directly in the command line script (eg:heads), make sure to update them in the dictionary too so that make_models restores our checkpoint correctly.
  • Run sample.py as outlined in the sampling section, but now with --model=my_model

For example, let's say we trained small_vqvae, small_prior, and small_upsampler under /path/to/jukebox/logs. In make_models.py, we are going to declare a tuple of the new models as my_model.

MODELS = {
    '5b': ("vqvae", "upsampler_level_0", "upsampler_level_1", "prior_5b"),
    '5b_lyrics': ("vqvae", "upsampler_level_0", "upsampler_level_1", "prior_5b_lyrics"),
    '1b_lyrics': ("vqvae", "upsampler_level_0", "upsampler_level_1", "prior_1b_lyrics"),
    'my_model': ("my_small_vqvae", "my_small_upsampler", "my_small_prior"),
}

Next, in hparams.py, we add them to the registry with the corresponding restore_paths and any other command line options used during training. Another important note is that for top-level priors with lyric conditioning, we have to locate a self-attention layer that shows alignment between the lyric and music tokens. Look for layers where prior.prior.transformer._attn_mods[layer].attn_func is either 6 or 7. If your model is starting to sing along lyrics, it means some layer, head pair has learned alignment. Congrats!

my_small_vqvae = Hyperparams(
    restore_vqvae='/path/to/jukebox/logs/small_vqvae/checkpoint_some_step.pth.tar',
)
my_small_vqvae.update(small_vqvae)
HPARAMS_REGISTRY["my_small_vqvae"] = my_small_vqvae

my_small_prior = Hyperparams(
    restore_prior='/path/to/jukebox/logs/small_prior/checkpoint_latest.pth.tar',
    level=1,
    labels=False,
    # TODO For the two lines below, if `--labels` was used and the model is
    # trained with lyrics, find and enter the layer, head pair that has learned
    # alignment.
    alignment_layer=47,
    alignment_head=0,
)
my_small_prior.update(small_prior)
HPARAMS_REGISTRY["my_small_prior"] = my_small_prior

my_small_upsampler = Hyperparams(
    restore_prior='/path/to/jukebox/logs/small_upsampler/checkpoint_latest.pth.tar',
    level=0,
    labels=False,
)
my_small_upsampler.update(small_upsampler)
HPARAMS_REGISTRY["my_small_upsampler"] = my_small_upsampler

Train with labels

To train with you own metadata for your audio files, implement get_metadata in data/files_dataset.py to return the artist, genre and lyrics for a given audio file. For now, you can pass '' for lyrics to not use any lyrics.

For training with labels, we'll use small_labelled_prior in hparams.py, and we set labels=True,labels_v3=True. We use 2 kinds of labels information:

  • Artist/Genre:
    • For each file, we return an artist_id and a list of genre_ids. The reason we have a list and not a single genre_id is that in v2, we split genres like blues_rock into a bag of words [blues, rock], and we pass atmost max_bow_genre_size of those, in v3 we consider it as a single word and just set max_bow_genre_size=1.
    • Update the v3_artist_ids and v3_genre_ids to use ids from your new dataset.
    • In small_labelled_prior, set the hps y_bins = (number_of_genres, number_of_artists) and max_bow_genre_size=1.
  • Timing:
    • For each chunk of audio, we return the total_length of the song, the offset the current audio chunk is at and the sample_length of the audio chunk. We have three timing embeddings: total_length, our current position, and our current position as a fraction of the total length, and we divide the range of these values into t_bins discrete bins.
    • In small_labelled_prior, set the hps min_duration and max_duration to be the shortest/longest duration of audio files you want for your dataset, and t_bins for how many bins you want to discretize timing information into. Note min_duration * sr needs to be at least sample_length to have an audio chunk in it.

After these modifications, to train a top-level with labels, run

mpiexec -n {ngpus} python jukebox/train.py --hps=vqvae,small_labelled_prior,all_fp16,cpu_ema --name=pretrained_vqvae_small_prior_labels \
--sample_length=1048576 --bs=4 --aug_shift --aug_blend --audio_files_dir={audio_files_dir} \
--labels=True --train --test --prior --levels=3 --level=2 --weight_decay=0.01 --save_iters=1000

For sampling, follow same instructions as above but use small_labelled_prior instead of small_prior.

Train with lyrics

To train in addition with lyrics, update get_metadata in data/files_dataset.py to return lyrics too. For training with lyrics, we'll use small_single_enc_dec_prior in hparams.py.

  • Lyrics:
    • For each file, we linearly align the lyric characters to the audio, find the position in lyric that corresponds to the midpoint of our audio chunk, and pass a window of n_tokens lyric characters centred around that.
    • In small_single_enc_dec_prior, set the hps use_tokens=True and n_tokens to be the number of lyric characters to use for an audio chunk. Set it according to the sample_length you're training on so that its large enough that the lyrics for an audio chunk are almost always found inside a window of that size.
    • If you use a non-English vocabulary, update text_processor.py with your new vocab and set n_vocab = number of characters in vocabulary accordingly in small_single_enc_dec_prior. In v2, we had a n_vocab=80 and in v3 we missed + and so n_vocab=79 of characters.

After these modifications, to train a top-level with labels and lyrics, run

mpiexec -n {ngpus} python jukebox/train.py --hps=vqvae,small_single_enc_dec_prior,all_fp16,cpu_ema --name=pretrained_vqvae_small_single_enc_dec_prior_labels \
--sample_length=786432 --bs=4 --aug_shift --aug_blend --audio_files_dir={audio_files_dir} \
--labels=True --train --test --prior --levels=3 --level=2 --weight_decay=0.01 --save_iters=1000

To simplify hps choices, here we used a single_enc_dec model like the 1b_lyrics model that combines both encoder and decoder of the transformer into a single model. We do so by merging the lyric vocab and vq-vae vocab into a single larger vocab, and flattening the lyric tokens and the vq-vae codes into a single sequence of length n_ctx + n_tokens. This uses attn_order=12 which includes prime_attention layers with keys/values from lyrics and queries from audio. If you instead want to use a model with the usual encoder-decoder style transformer, use small_sep_enc_dec_prior.

For sampling, follow same instructions as above but use small_single_enc_dec_prior instead of small_prior. To also get the alignment between lyrics and samples in the saved html, you'll need to set alignment_layer and alignment_head in small_single_enc_dec_prior. To find which layer/head is best to use, run a forward pass on a training example, save the attention weight tensors for all prime_attention layers, and pick the (layer, head) which has the best linear alignment pattern between the lyrics keys and music queries.

Fine-tune pre-trained top-level prior to new style(s)

Previously, we showed how to train a small top-level prior from scratch. Assuming you have a GPU with at least 15 GB of memory and support for fp16, you could fine-tune from our pre-trained 1B top-level prior. Here are the steps:

  • Support --labels=True by implementing get_metadata in jukebox/data/files_dataset.py for your dataset.
  • Add new entries in jukebox/data/ids. We recommend replacing existing mappings (e.g. rename "unknown", etc with styles of your choice). This uses the pre-trained style vectors as initialization and could potentially save some compute.

After these modifications, run

mpiexec -n {ngpus} python jukebox/train.py --hps=vqvae,prior_1b_lyrics,all_fp16,cpu_ema --name=finetuned \
--sample_length=1048576 --bs=1 --aug_shift --aug_blend --audio_files_dir={audio_files_dir} \
--labels=True --train --test --prior --levels=3 --level=2 --weight_decay=0.01 --save_iters=1000

To get the best sample quality, it is recommended to anneal the learning rate in the end. Training the 5B top-level requires GPipe which is not supported in this release.

Citation

Please cite using the following bibtex entry:

@article{dhariwal2020jukebox,
  title={Jukebox: A Generative Model for Music},
  author={Dhariwal, Prafulla and Jun, Heewoo and Payne, Christine and Kim, Jong Wook and Radford, Alec and Sutskever, Ilya},
  journal={arXiv preprint arXiv:2005.00341},
  year={2020}
}

License

Noncommercial Use License

It covers both released code and weights.