Users are excited by strong performance and clean designs in recent open-source releases like PrismML ternary models and Alibaba's Qwen3.8, praising efficient quantization and creative benchmarks.
Based on 5 visible X reactions from 13 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@mervenoyann @PrismML @thinkymachines @Kimi_Moonshot open model evals keep getting more creative — love seeing domain-specific benchmarks that actually stress test real capabilities
@mervenoyann @PrismML @thinkymachines @Kimi_Moonshot Ternary at 95% of fp16 is wild. That changes the deployment math fast
@mervenoyann @PrismML @thinkymachines @Kimi_Moonshot Just when you thought that's it, Qwen drops a banger
@mervenoyann Qwen3.8 yassssss
this week we had everything @PrismML new 27B ternary & binary models 🪴 10x smaller with 95% equally smart to fp16 🤯 @thinkymachines Inkling open sota multimodal reasoning Omni @Kimi_Moonshot K3 announcement girls only want one thing and it's open models 😄
just when I think it was over it just wasn't https://x.com/alibaba_qwen/status/2078759124914098291?s=46
Users are excited by strong performance and clean designs in recent open-source releases like PrismML ternary models and Alibaba's Qwen3.8, praising efficient quantization and creative benchmarks.
Based on 5 visible X reactions from 13 accounts; directional sample.
Ask a question below.
Published answers will appear here.