Translations:Diffusion Models Are Real-Time Game Engines/44/en: Difference between revisions

Latest revision as of 03:06, 7 September 2024

Information about message (contribute)

This message has no documentation. If you know where or how this message is used, you can help other translators by adding documentation to this message.

Message definition (Diffusion Models Are Real-Time Game Engines)

Since we do observe degradation when using just a single sampling step, we also experimented with model distillation similarly to (Yin et al., [https://arxiv.org/html/2408.14837v1#bib.bib39 2024]; Wang et al., [https://arxiv.org/html/2408.14837v1#bib.bib36 2023]) in the single-step setting. Distillation does help substantially there (allowing us to reach 50 FPS as above), but still comes at some cost to simulation quality, so we opt to use the 4-step version without distillation for our method (see Appendix [https://arxiv.org/html/2408.14837v1#A1.SS4 A.4]). This is an interesting area for further research.

Since we do observe degradation when using just a single sampling step, we also experimented with model distillation similarly to (Yin et al., 2024; Wang et al., 2023) in the single-step setting. Distillation does help substantially there (allowing us to reach 50 FPS as above), but still comes at some cost to simulation quality, so we opt to use the 4-step version without distillation for our method (see Appendix A.4). This is an interesting area for further research.

Revision as of 05:28, 1 September 2024 (view source) FuzzyBot (talk \| contribs) (Importing a new version from external source)		Latest revision as of 03:06, 7 September 2024 (view source) FuzzyBot (talk \| contribs) (Importing a new version from external source)
Line 1:		Line 1:
	Since we do observe degradation when using just a single sampling step, we also experimented with model distillation similarly to (Yin ~~et al~~., [https://arxiv.org/html/2408.14837v1#bib.bib39 2024]; Wang ~~et al~~., [https://arxiv.org/html/2408.14837v1#bib.bib36 2023]) in the single-step setting. Distillation does help substantially there (allowing us to reach 50 FPS as above), but still comes at some cost to simulation quality, so we opt to use the 4-step version without distillation for our method (see Appendix [https://arxiv.org/html/2408.14837v1#A1.SS4 A.4]). This is an interesting area for further research.		Since we do observe degradation when using just a single sampling step, we also experimented with model distillation similarly to (Yin et al., [https://arxiv.org/html/2408.14837v1#bib.bib39 2024]; Wang et al., [https://arxiv.org/html/2408.14837v1#bib.bib36 2023]) in the single-step setting. Distillation does help substantially there (allowing us to reach 50 FPS as above), but still comes at some cost to simulation quality, so we opt to use the 4-step version without distillation for our method (see Appendix [https://arxiv.org/html/2408.14837v1#A1.SS4 A.4]). This is an interesting area for further research.