Abstract: Embodiments of a generative framework comprise cooperative learning of two generative flow models, in which the two models are iteratively updated based on the jointly synthesized examples. In one or more embodiments, the first flow model is a normalizing flow that transforms an initial simple density into a target density by applying a sequence of invertible transformations, and the second flow model is a Langevin flow that runs finite steps of gradient-based MCMC toward an energy-based model. In learning iterations, synthesized examples are generated by using a normalizing flow initialization followed by a short-run Langevin flow revision toward the current energy-based model. Then, the synthesized examples may be treated as fair samples from the energy-based model and the model parameters are updated, while the normalizing flow directly learns from the synthesized examples by maximizing the tractable likelihood. Also provided are both theoretical and empirical justifications for the embodiments.

Abstract: This application provides a program operating method and a related apparatus. The method includes obtaining first operating data of the program from a server, the first operating data comprising at least an instruction for starting the program; executing the first operating data, and displaying a first page of the program according to the first operating data; determining, in response to a second triggering instruction of a control in the first page, a second page identifier corresponding to the control; obtaining second operating data from the server according to the second page identifier if operating data corresponding to the second page identifier is not in the first operating data, the second operating data comprising the operating data corresponding to the second page identifier; and executing the second operating data, and displaying a second page according to the second operating data, the second operating data comprising portions of the program.