3abiton

60 ms·

631.
by 3abiton 3y ago
This is an amazing experiment summary! I wonder if your APU conclusion still stands (CPU for bigger models otherwise APU), given all the advancements so far.
632.
by 3abiton 3y ago
One of the best application of LLM!
633.
by 3abiton 3y ago
Sounds like universal appimage
634.
by 3abiton 3y ago
Is he the person after the Yi LLM model?
635.
by 3abiton 3y ago
Doesn't certain licenses (like MIT) prevent exactly that?
636.
by 3abiton 3y ago
Interesting observation, that's experimental physics for you.
637.
by 3abiton 3y ago
Thanks for the clarification. I feel the defense against AI will be a major problem and a cat and an endless cat and mouse game.
638.
by 3abiton 3y ago
I wonder if that would be the case for graph based models too
639.
by 3abiton 3y ago
Anyone managed to get it working for AMD iGPU (APU) yet? I got the vega archi, still no luck running LLMs with either ZLUDA or ROCm backend.
640.
by 3abiton 3y ago
The problem is EU's winter. Pair this bad boy with a heat pump and watch it roar.
641.
by 3abiton 3y ago
This looks so promising! The only issue is that many MB don't have drivers available in the Linux kernel, makes reading temp control not possible yet.
642.
by 3abiton 3y ago
The issue is benchmarks for LLMs or models formats are tough to compare, as there are many factors at play. But beyond ooba's comparison, many other sources recommend GPTQ or AWQ for GPU inference as it gives better quality for the sam
643.
by 3abiton 3y ago
My "instead" meant instead of putting my info on LinkedIn (open for CV farmers), rather create a personal website/CV that you have more control over who can see it (more private), as you choose to share it only with specific
644.
by 3abiton 3y ago
Interesting, the main selling point is, pySpark but better as it is python native?
645.
by 3abiton 3y ago
I worry about privacy (CV farmers), and (current) employer snooping. Western Europe too, always get messages from recruiters asking for CVs even though I got everything on LinkedIn. I am curious if a personal website with an online CV is be
646.
by 3abiton 3y ago
Aren't TPU the first LLM-Native processors?
647.
by 3abiton 3y ago
This finding is so funny, it's like despite the interconnectedness in the 21st century things like this still happen.
648.
by 3abiton 3y ago
But usually AWQ get recommended for GPU inference over GGUF
649.
by 3abiton 3y ago
How will Mojo solve what is essentially a GPU driver issue (ROCm vs cuda)? AMD's track at supporting GPU computing is very poor to be trusted so far.
650.
by 3abiton 3y ago
Do you know any decent framework (ideally using python) to visualize such interactions?
651.
by 3abiton 3y ago
I never gave display port any attention, what's the appeal? Open-source?
652.
by 3abiton 3y ago
Glad to see some tutorials!
653.
by 3abiton 3y ago
I never thought maths was that cool.
654.
by 3abiton 3y ago
A bit sad hobbyist have to resort to such measures to get tinkering, not to mention the initial capital needed. We're all slaves to Nvidia's VRAM monopoly until AMD or Intel steps in and release a competitive alternative with beef
655.
by 3abiton 3y ago
Is the brain applying LoRa on our memory model?
656.
by 3abiton 3y ago
What makes it better than the current existing solutions though? For someone who is just interested in rapid prototyping?
657.
by 3abiton 3y ago
I agree, being a dev myself hosting few website, and a data enthusiast, it is very hard to balance both, except if all data was open source and released for all it's owners. All online platforms, especially social media relay on users
658.
by 3abiton 3y ago
Nintendo made it clear that this is also "illegal". You have to opt for the worse experience (using the switch) otherwise you're the bad guy ...
659.
by 3abiton 3y ago
How different is this compared to Jan.ai for example?
660.
by 3abiton 3y ago
The article content is misrepresented by the title, yet the subtitle offer the real answer: "Common Misconceptions about MLOps, LLMs and Machine Learning in production"

More ›