Semi-Autoregressive Energy Flows: Exploring Likelihood-Free Training of Normalizing Flows
This work addresses computational bottlenecks in generative modeling for researchers and practitioners, offering an incremental alternative to maximum likelihood training.
The paper tackles the challenge of training normalizing flow generative models by proposing a likelihood-free energy objective that avoids computationally expensive Jacobian determinants, enabling flexible architectures like semi-autoregressive energy flows with competitive sample quality and generation speed, though log-likelihood estimates are poor.
Training normalizing flow generative models can be challenging due to the need to calculate computationally expensive determinants of Jacobians. This paper studies the likelihood-free training of flows and proposes the energy objective, an alternative sample-based loss based on proper scoring rules. The energy objective is determinant-free and supports flexible model architectures that are not easily compatible with maximum likelihood training, including semi-autoregressive energy flows, a novel model family that interpolates between fully autoregressive and non-autoregressive models. Energy flows feature competitive sample quality, posterior inference, and generation speed relative to likelihood-based flows; this performance is decorrelated from the quality of log-likelihood estimates, which are generally very poor. Our findings question the use of maximum likelihood as an objective or a metric, and contribute to a scientific study of its role in generative modeling.