Benchmarking

[TomLucidor]

After taking a look at benchmarks doing comparisons, wondering if there are ways to show that Sana is really flexible or not

• Flux https://blackforestlabs.ai/announcing-black-forest-labs/
• Playground v3 https://arxiv.org/html/2409.10695v2
• SD3 https://stability.ai/news/stable-diffusion-3-research-paper
• PixArt Sigma https://arxiv.org/html/2403.04692v2
• Open arena to compare model output https://artificialanalysis.ai/text-to-image/arena

[lawrence-cj]

Do you mean we should add a comparison on leaderboard or something else?

[TomLucidor]

Yeah, comparison or leaderboard or something that demo their performance vs your SOTA (preferably against SDXL tweaks, Flux, and AuraFlow WIPs)

[lawrence-cj]

Do you know how to get performance ranking from here: https://artificialanalysis.ai/text-to-image/arena?

[TomLucidor]

@lawrence-cj you mean the user preference platform? Yeah their contact form is open to accepting new models for testing so would like to see how things will pan out