score-models 0.5.11

A simple pytorch interface for score model and basic diffusion.

A storage for score-based models. The ScoreModel interface gives access to the following utilities:

• Simple initialization of MLP, NCSN++ and DDPM neural network architectures

• A fit method to train the score model on a dataset using Denoising Score Matching (DSM).

• A sampling method based on an Euler-Maruyama discretization of an SDE.

• A simple interface to train an energy model using DSM

This repository is mainly intended for personal use. You might also want to refer to the original implementation at https://github.com/yang-song/score_sde.

Installation

To install the package, you can use pip:

pip install torch-score-models

Usage

ScoreModel

The ScoreModel class is the main interface for training and using score models, defined as

\begin{equation*} \mathbf{s}_\theta(t, \mathbf{x}) = \frac{1}{\sigma(t)} \nabla_\mathbf{x} \log p_t(\mathbf{x}) \end{equation*}

where \(\sigma(t)\) is the standard deviation of the perturbation kernel \(p_t(\mathbf{x} \mid \mathbf{x}_0)\) of an SDE.

The ScoreModel class extends the torch.nn.Module class. Example usage:

from score_models import ScoreModel, EnergyModel, NCSNpp, MLP, DDPM

# B is the batch size
# C is the input channels
# dimensions are the spatial dimensions (e.g. [28, 28] for MNIST)

# Create a ScoreModel instance with Yang Song's NCSN++ architecture and the VESDE
net = NCSNpp(channels=C, dimensions=len(dimensions), nf=128, ch_mult=[2, 2, 2, 2])
model = ScoreModel(model=net, sigma_min=1e-2, sigma_max=50, device="cuda")
# ... or the VPSDE
model = ScoreModel(model=net, beta_min=1e-2, beta_max=20, device="cuda")

# NN Architectures support a Unet with 1D convolutions for time series input data
net = NCSNpp(channels=1, nf=128, ch_mult=[2, 2, 2, 2], dimensions=1)
# ... or 3D convolutions for videos/voxels
net = NCSNpp(channels=1, nf=128, ch_mult=[2, 2, 2, 2], attention=False, dimensions=3)
# You can also use a simpler MLP architecture (dimensions=0)
net = MLP(dimensions=C, layers=4, units=100)
# ... or Jonathan Ho's DDPM architecture
net = DDPM(channels=1, dimensions=2, nf=128, ch_mult=[2, 2, 2, 2])

# Train the score model, and save its weight in checkpoints_directory
model.fit(dataset=your_dataset, epochs=100, batch_size=B, learning_rate=1e-4, checkpoints_directory=checkpoints_directory)

# Generate samples from the trained model (steps is the number of Euler-Maruyama steps)
samples = model.sample(shape=[B, *dimensions], steps=1000)

# Compute the score for a given input
score = model.score(t, x)

# Initialize the score model and its neural network from a path to a checkpoint directory
score = ScoreModel(checkpoints_directory=checkpoints_directory)

EnergyModel

The EnergyModel class works in pretty much the same way as ScoreModel, but implements the score via the automatic differentiation of an energy model

\begin{equation*} E_\theta(t, \mathbf{x}) = \frac{1}{2 \sigma(t)} \lVert \mathbf{x} - f_\theta(t, \mathbf{x}) \rVert_2^2 \end{equation*}

This is to say that the score is defined as

\begin{equation*} \mathbf{s}_\theta(t, \mathbf{x}) = - \nabla_\mathbf{x} E_\theta(t, \mathbf{x}) \end{equation*}

Note: When using the MLP architecture, the energy model can be constructed more efficiently as the output of the neural network by specifying nn_is_energy in the hyperparameters of the MLP, which will modify the neural network architecture to be a function \(f_\theta: \mathbb{R}^d \to \mathbb{R}\) instead of \(f_\theta: \mathbb{R}^d \to \mathbb{R}^d\). An output_activation like relu can also be specified to make the energy positive or bounded from below.

Training Parameters

When training the model using the fit method, you can provide various parameters to customize the training process. Some important parameters include:

• dataset: The training dataset (torch.utils.data.Dataset).

• epochs: The number of training epochs.

• learning_rate: The learning rate for the ADAM optimizer.

• batch_size: The batch size for training.

• checkpoints_directory: The directory to save model checkpoints (default: None).

• seed: The random seed for numpy and torch.

Refer to the method’s docstring or the class definition for more details on available parameters.

Citations

If you use this package in your research, please consider citing the following papers:

 @inproceedings{NEURIPS2020_4c5bcfec,
     author      = {Ho, Jonathan and Jain, Ajay and Abbeel, Pieter},
     booktitle   = {Advances in Neural Information Processing Systems},
     editor      = {H. Larochelle and M. Ranzato and R. Hadsell and M.F. Balcan and H. Lin},
     pages       = {6840--6851},
     publisher   = {Curran Associates, Inc.},
     title       = {Denoising Diffusion Probabilistic Models},
     url         = {https://proceedings.neurips.cc/paper/2020/file/4c5bcfec8584af0d967f1ab10179ca4b-Paper.pdf},
     volume      = {33},
     year        = {2020}
}

@inproceedings{song2021scorebased,
  title={Score-Based Generative Modeling through Stochastic Differential Equations},
  author={Yang Song and Jascha Sohl-Dickstein and Diederik P Kingma and Abhishek Kumar and Stefano Ermon and Ben Poole},
  booktitle={International Conference on Learning Representations},
  year={2021},
  url={https://openreview.net/forum?id=PxTIG12RRHS}
}

License

This package is licensed under the MIT License.

You can save the above content in an .rst file, such as README.rst, and include it in your project’s root directory.