n5bytes

Updated the UI for rawsort.com rate it on 10, Whats feels off #buildinpublic #webdev

@iatnon Yep followed u back

@printfn5 Hi, I’m doing a Master in AI. Let’s connect!

Let’s connect👋 Solo founder building in AI + tech space. Looking to connect with people interested in tech, startups, AI, and building. If you’re working on something or just love talking about these topics drop a comment or DM

@mentor_krunal Yep followed u back

@printfn5 Awesome work! I've followed you, let's support each other.

@zubidavies Damn thats nice

@printfn5 At 58,000 page views since yesterday

@HyveMindx1 Followed back gl with ur project

@printfn5 Yeah definitely bud, Just followed you!✨

@HyveMindx1 Ah ok hope it goes well

Yes, definitely, although in this case, users also have the flexibility to experiment with even 8-bit quantized models depending on their hardware capabilities. The idea is to let users pick compatible GGUF-based models directly from Hugging Face and run them on-device. For devices powered by Snapdragon 8 Gen 2 and above, things become especially interesting because alongside standard CPU-based inference, I was also able to support inference acceleration through Qualcomm’s Hexagon NPU. So on supported Snapdragon devices, users essentially get two inference paths: • CPU-based inference • Hexagon NPU-accelerated inference

@HyveMindx1 Yea I heard mobiles can run 4bit quantified Gemma

@printfn5 Haha yes, and that’s exactly the core vision behind this 😄 Still a lot of optimization and experimentation going on, but it’s already running quantized GGUF models on device.✨

@zubidavies Useful as a utility tool for pdf but getting to the right users might take time

@printfn5 Right PDF Kit rightpdfkit.com

@Robyn_Social Ok follow back

@printfn5 LETS!

@udaykirancodes Ok fb

@printfn5 Lets connect

@AdityaBehe64144 Ok fb

@printfn5 Hey let's connect

@HyveMindx1 If u actually manage to run llms on mobile then this project is really gud

Cool, let’s connect! Btw, I’ve been building something called QuantaLLM, an experimental step toward running powerful LLMs directly on handheld devices through Android-native on-device inference. Would love your feedback and a visit if it resonates with you #buildinpublic #AI #GoogleAIEdge #llamacpp quanta-web-pi.vercel.app/?v=8/

@OpeyemiRighteo1 Ok fb

@printfn5 Let's connect

@chris_tcsc This is useful but might not be a daily tool tho

@printfn5 Building CleanMyList.io and some others - let’s #connect 👌

@Sonofpeace0001 Yep lets do it

@printfn5 Connecting with you all if you're into automation

@LaloLoops Yep lets do it

@printfn5 Hey also building with AI and Agents here. Happy to connect 🤝

@Bramha_01 Ok fb

@printfn5 Let's connect

@debojyoti452 @pocketbase @KeplarsHQ Seems useful but maybe for people who use complex workflows

Building keplars.com - an email instrastructure platform, simplified under 5 clicks. Recently launched Integrations with @pocketbase Here's a full blog on how to integrate @pocketbase with @KeplarsHQ for Production-Ready Email Workflows. Do give it a read : keplars.com/blog/pocketbas…

@sophiaanhnguyen Yep lets connect

@printfn5 Hey 👋 Also building in the AI space. Working on PixelPanda.ai, helping product-based brands create better visuals from simple photos, without needing a full studio setup. Would be great to connect with more AI/startup folks here.

@NirajAdhik52674 Thanks man

@printfn5 Looks great buddy

Updated the UI for rawsort.com rate it on 10, Whats feels off #buildinpublic #webdev