Fast Sampling via Discrete Non-Markov Diffusion Models with Predetermined Transition Time
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2023
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910729164029952 |
|---|---|
| author | Chen, Zixiang Yuan, Huizhuo Li, Yongqian Kou, Yiwen Zhang, Junkai Gu, Quanquan |
| author_facet | Chen, Zixiang Yuan, Huizhuo Li, Yongqian Kou, Yiwen Zhang, Junkai Gu, Quanquan |
| contents | Discrete diffusion models have emerged as powerful tools for high-quality data generation. Despite their success in discrete spaces, such as text generation tasks, the acceleration of discrete diffusion models remains under-explored. In this paper, we propose discrete non-Markov diffusion models (DNDM), which naturally induce the predetermined transition time set. This enables a training-free sampling algorithm that significantly reduces the number of function evaluations (i.e., calls to the neural network), making the sampling process much faster. Furthermore, we study the transition from finite to infinite step sampling, offering new insights into bridging the gap between discrete and continuous-time processes for discrete diffusion models. Extensive experiments on natural language generation and machine translation tasks demonstrate the superior performance of our method in terms of both generation speed and sample quality compared to existing methods for discrete diffusion models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_09193 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Fast Sampling via Discrete Non-Markov Diffusion Models with Predetermined Transition Time Chen, Zixiang Yuan, Huizhuo Li, Yongqian Kou, Yiwen Zhang, Junkai Gu, Quanquan Machine Learning Artificial Intelligence Discrete diffusion models have emerged as powerful tools for high-quality data generation. Despite their success in discrete spaces, such as text generation tasks, the acceleration of discrete diffusion models remains under-explored. In this paper, we propose discrete non-Markov diffusion models (DNDM), which naturally induce the predetermined transition time set. This enables a training-free sampling algorithm that significantly reduces the number of function evaluations (i.e., calls to the neural network), making the sampling process much faster. Furthermore, we study the transition from finite to infinite step sampling, offering new insights into bridging the gap between discrete and continuous-time processes for discrete diffusion models. Extensive experiments on natural language generation and machine translation tasks demonstrate the superior performance of our method in terms of both generation speed and sample quality compared to existing methods for discrete diffusion models. |
| title | Fast Sampling via Discrete Non-Markov Diffusion Models with Predetermined Transition Time |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2312.09193 |