HackerTrans
トップ新着トレンドコメント過去質問紹介求人

tensorlibb

no profile record

投稿

25L Portable NV-linked Dual 3090 LLM Rig

reddit.com
139 ポイント·投稿者 tensorlibb·10 か月前·121 コメント

Anthropic admits they nerfed their Claude model in August

twitter.com
5 ポイント·投稿者 tensorlibb·10 か月前·6 コメント

コメント

tensorlibb
·9 か月前·議論
[dead]
tensorlibb
·9 か月前·議論
[dead]
tensorlibb
·9 か月前·議論
[dead]
tensorlibb
·9 か月前·議論
[dead]
tensorlibb
·9 か月前·議論
[dead]
tensorlibb
·10 か月前·議論
Any examples of face tracking a video and exporting mask coordinates to apply a filter in ffmpeg
tensorlibb
·10 か月前·議論
This hardware is fascinating - I've seen other examples of these farms Deployed in Ukraine but does anyone have more info on how these "servers" are orchestrated?
tensorlibb
·10 か月前·議論
Glad to hear!
tensorlibb
·10 か月前·議論
It's actually just wild - my last apt in Manhattan was built in 2018 and had a gas range. The "exhaust fan" just vented back into the kitchen and the only "exhaust" per se was a vent in the bathroom that "sucked" all the way to the roof with negative pressure.

If I ran my stove for more than 15 min my carbon monoxide / fire alarm would go off.
tensorlibb
·10 か月前·議論
I've found just opening a window does the most to make a difference - in most European / US homes (even with propane / gas service) the "exhaust" just blows into the same room or a cabinet!

That said, I personally use the Breathe Airmonitor Plus [1] - I kept having issues with calibration with the temtop unit. Mostly decided on this one since it uses an NDIR sensor similar to my Aranet which I carry with me all the time.

0 - https://amzn.to/423AAaj
tensorlibb
·10 か月前·議論
Curious if this will ever be open sourced?? Super cool project!
tensorlibb
·10 か月前·議論
I'm a huge fan of OpenRouter and their interface for solid LLM's but I recently jumped into fine tuning / modifying my own vision models for FPV drone detection (just for fun) and my daily workstation and it's 2080 just wasn't good enough.

Even in 2025 it's cool how solid a setup dual 3090's still are. nvlink is an absolute must but it's incredibly powerful. I'm able to run the latest Mistral thinking models and relatively powerful yolo based VLM's like the ones RoboFlow is based on.

Curious if anyone else is still using 3090's or has feedback for scaling up to 4-6 3090s.

Thanks everyone ;)
tensorlibb
·10 か月前·議論
Haha I've built something somewhat similar - glad I wasn't the only one who solved the stuck gpu job problem this way :)
tensorlibb
·10 か月前·議論
This is incredible! What gpu configs, budget to ultra high-end, would you recommend for local fine tuning?

Always curious to see what other ai enthusiasts are running!
tensorlibb
·10 か月前·議論
Agreed, but 95% of users simply don't see this or know to use it.
tensorlibb
·10 か月前·議論
They also never disclosed they were load balancing with nvidia GPUs and Amazon aws trainium inference chips until people complained.
tensorlibb
·10 か月前·議論
What's interesting is solar in cold environments on sunny days can actually OVER produce.

I've had a few friends run into this issue with systems almost frying inverters because they didn't allow enough headroom for open circuit voltage coming from the panels.
tensorlibb
·10 か月前·議論
I think this is a massively good step forward.

However, TikTok is still a brain rot slop machine and we would be right to question Ellison's motivations.
tensorlibb
·10 か月前·議論
Any idea where these copycat devs are from?
tensorlibb
·10 か月前·議論
This is cool - but I'm still highly skeptical of tools like this actually working at scale.

Have you done any substantive user studies to actually measure how effective this tool is vs not using an "ai tool" to learn?