Training and inference code for audio generation models

The library can be installed from PyPI with:

$ pip install stable-audio-tools

To run the training scripts or inference code, you'll want to clone this repository, navigate to the root, and run:

$ pip install .

Requires PyTorch 2.0 or later for Flash Attention support

Development for the repo is done in Python 3.8.10

Before starting your training run, you'll need a model config file, as well as a dataset config file. For more information about those, refer to the Configurations section below

The training code also requires a Weights & Biases account to log the training outputs and demos. Create an account and log in with:

$ wandb login

To start a training run, run the train.py script in the repo root with:

$ python3 ./train.py --dataset-config /path/to/dataset/config --model-config /path/to/model/config --name harmonai_train

The --name parameter will set the project name for your Weights and Biases run.

stable-audio-tools uses PyTorch Lightning to facilitate multi-GPU and multi-node training.

When a model is being trained, it is wrapped in a "training wrapper", which is a pl.LightningModule that contains all of the relevant objects needed only for training. That includes things like discriminators for autoencoders, EMA copies of models, and all of the optimizer states.

The checkpoint files created during training include this training wrapper, which greatly increases the size of the checkpoint file.

unwrap_model.py in the repo root will take in a wrapped model checkpoint and save a new checkpoint file including only the model itself.

That can be run with from the repo root with:

$ python3 ./unwrap_model.py --model-config /path/to/model/config --ckpt-path /path/to/wrapped/ckpt --name model_unwrap

Unwrapped model checkpoints are required for:

• Inference scripts
• Using a model as a pretransform for another model (e.g. using an autoencoder model for latent diffusion)
• Fine-tuning a pre-trained model with a modified configuration (i.e. partial initialization)

Fine-tuning a model involves continuning a training run from a pre-trained checkpoint.

To continue a training run from a wrapped model checkpoint, you can pass in the checkpoint path to train.py with the --ckpt-path flag.

To start a fresh training run using a pre-trained unwrapped model, you can pass in the unwrapped checkpoint to train.py with the --pretrained-ckpt-path flag.

Additional optional flags for train.py include:

• --config-file
  • The path to the defaults.ini file in the repo root, required if running train.py from a directory other than the repo root
• --pretransform-ckpt-path
  • Used in various model types such as latent diffusion models to load a pre-trained autoencoder. Requires an unwrapped model checkpoint.
• --save-dir
  • The directory in which to save the model checkpoints
• --checkpoint-every
  • The number of steps between saved checkpoints.
  • Default: 10000
• --batch-size
  • Number of samples per-GPU during training. Should be set as large as your GPU VRAM will allow.
  • Default: 8
• --num-gpus
  • Number of GPUs per-node to use for training
  • Default: 1
• --num-nodes
  • Number of GPU nodes being used for training
  • Default: 1
• --accum-batches
  • Enables and sets the number of batches for gradient batch accumulation. Useful for increasing effective batch size when training on smaller GPUs.
• --strategy
  • Multi-GPU strategy for distributed training. Setting to deepspeed will enable DeepSpeed ZeRO Stage 2.
  • Default: ddp if --num_gpus > 1, else None
• --precision
  • floating-point precision to use during training
  • Default: 16
• --num-workers
  • Number of CPU workers used by the data loader
• --seed
  • RNG seed for PyTorch, helps with deterministic training

Training and inference code for stable-audio-tools is based around JSON configuration files that define model hyperparameters, training settings, and information about your training dataset.

The model config file defines all of the information needed to load a model for training or inference. It also contains the training configuration needed to fine-tune a model or train from scratch.

The following properties are defined in the top level of the model configuration:

• model_type
  • The type of model being defined, currently limited to one of "autoencoder", "diffusion_uncond", "diffusion_cond", "diffusion_cond_inpaint", "diffusion_autoencoder", "musicgen".
• sample_size
  • The length of the audio provided to the model during training, in samples. For diffusion models, this is also the raw audio sample length used for inference.
• sample_rate
  • The sample rate of the audio provided to the model during training, and generated during inference, in Hz.
• audio_channels
  • The number of channels of audio provided to the model during training, and generated during inference. Defaults to 2. Set to 1 for mono.
• model
  • The specific configuration for the model being defined, varies based on model_type
• training
  • The training configuration for the model, varies based on model_type. Provides parameters for training as well as demos.

stable-audio-tools currently supports two kinds of data sources: local directories of audio files, and WebDataset datasets stored in Amazon S3. More information can be found in the dataset config documentation

• Add documentation for different model types
• Add documentation for Gradio interface
• Add troubleshooting section
• Add contribution guidelines