TESS: Text-to-Text Self-Conditioned Simplex Diffusion

doi:doi:10.57702/bvpixosy

TESS: Text-to-Text Self-Conditioned Simplex Diffusion

Diffusion models have emerged as a power-ful paradigm for generation, obtaining strong performance in various continuous domains. However, applying continuous diffusion models to natural language remains challenging due to its discrete nature and the need for a large number of diffusion steps to generate text, making diffusion-based generation expensive.

BibTex:

@dataset{Rabeeh_Karimi_Mahabadi_and_James_Henderson_and_Iz_Beltagy_and_Hamish_Ivison_and_Matthew_E_Peters_and_Jaesung_Tae_and_Arman_Cohan_2024,
    abstract = {Diffusion models have emerged as a power-ful paradigm for generation, obtaining strong performance in various continuous domains. However, applying continuous diffusion models to natural language remains challenging due to its discrete nature and the need for a large number of diffusion steps to generate text, making diffusion-based generation expensive.},
    author = {Rabeeh Karimi Mahabadi and James Henderson and Iz Beltagy and Hamish Ivison and Matthew E. Peters and Jaesung Tae and Arman Cohan},
    doi = {10.57702/bvpixosy},
    institution = {No Organization},
    keyword = {'diffusion models', 'natural language processing', 'text generation'},
    month = {dec},
    publisher = {TIB},
    title = {TESS: Text-to-Text Self-Conditioned Simplex Diffusion},
    url = {https://service.tib.eu/ldmservice/dataset/tess--text-to-text-self-conditioned-simplex-diffusion},
    year = {2024}
}