ByteDance just dopped Doubao-1.5-pro tht uses sparse MoE architecture, it matches GPT 4o benchmarks while being 50x cheaper to run, and it's 5x cheaper than DeepSeek

☆ Yσɠƚԋσʂ ☆@lemmy.ml · 2 days ago

ByteDance just dopped Doubao-1.5-pro tht uses sparse MoE architecture, it matches GPT 4o benchmarks while being 50x cheaper to run, and it's 5x cheaper than DeepSeek

chiisana@lemmy.chiisana.net · 20 hours ago

Deepseek referred here seems to be v3, not r1. While the linked article didn’t seem to have info on parameter size, fact that they state it is sparse MoE architecture should suggest it is capable to run pretty quick (compared to other models of similar parameter space), so that’s cool.

uberstar@lemmy.ml · 2 days ago

Sweet! Where can I try it? for whatever reason im getting shitty search results looking for this

☆ Yσɠƚԋσʂ ☆@lemmy.ml · 2 days ago

not sure if you can sign up for it outside China https://team.doubao.com/en/special/doubao_1_5_pro

uberstar@lemmy.ml · edit-2 2 days ago

ty :), maybe some random VPN will do if so

☆ Yσɠƚԋσʂ ☆@lemmy.ml · 2 days ago

also worth noting that you can just run DeepSeek locally https://dev.to/shayy/run-deepseek-locally-on-your-laptop-37hl

ByteDance just dopped Doubao-1.5-pro tht uses sparse MoE architecture, it matches GPT 4o benchmarks while being 50x cheaper to run, and it's 5x cheaper than DeepSeek

ByteDance just dopped Doubao-1.5-pro tht uses sparse MoE architecture, it matches GPT 4o benchmarks while being 50x cheaper to run, and it's 5x cheaper than DeepSeek

ByteDance Releases Doubao Large Model 1.5 Pro, Performance Surpassing GPT-4o and Claude3.5Sonnet