Imperfect score-matching leads to a shift between the training and the sampling distribution of diffusion models. Due to the recursive nature of the generation process, errors in previous steps yield sampling iterates that drift away from the training distribution. Yet, the standard training objective via Denoising Score Matching (DSM) is only designed to optimize over non-drifted data. To train on drifted data, we propose to enforce a \emph{consistency} property which states that predictions of the model on its own generated data are consistent across time. Theoretically, we show that if the score is learned perfectly on some non-drifted points (via DSM) and if the consistency property is enforced everywhere, then the score is learned accurately everywhere. Empirically we show that our novel training objective yields state-of-the-art results for conditional and unconditional generation in CIFAR-10 and baseline improvements in AFHQ and FFHQ. We open-source our code and models: https://github.com/giannisdaras/cdm
翻译:由于生成过程的递归性质,以往步骤的错误导致抽样重复,偏离了培训分布。然而,通过Denooising分数匹配(DSM)的标准培训目标仅旨在优化非分散数据。关于漂流数据的培训,我们建议实施一个\emph{continente}属性,该属性指出,对模型本身生成的数据的预测在时间上是一致的。理论上,我们表明,如果在一些非分散点(通过DSM)上完全学到了分数,如果一致性属性在各地得到实施,那么分数就会在各地得到准确的。我们欣喜地表明,我们的新培训目标产生了在CIFAR-10中有条件和无条件生成的最新结果,以及AFHQ和FFHQ的基线改进。我们打开了我们的代码和模型:https://github.com/giannsdaras/cdm。