Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Can I run this LLM? (locally)

can-i-run-this-llm-blue.vercel.app

Post reply on HN
Print view View original

41–50 of 55 posts

Re: Show HN: Can I run this LLM? (locally)

#41

Post by thaumasiotes »

lucb1e wrote: post #26
codingdave wrote: post #2

And herein lies the problem with vibe coding - accuracy is wanting. I can absolutely run models that this site says cannot be run. Shared RAM is a thing - even with limited VRAM, shared RAM can compensate to run larger models. (Slowly, admittedly, but they work.)

New word for me: vibe coding > coined the term in February 2025 > Vibe coding is a new coding style [...] A programmer can describe a program in words and get an AI tool to generate working code, without requiring an understanding of the code. [...] [The programmer] surrenders to the "vibes" of the AI [without reading the resulting code.] When errors arise, he simply copies them into the system without further explan…

Austen Allred sold a group of investors on the idea that this was the future of everything.

https://www.gauntletai.com/

Re: Show HN: Can I run this LLM? (locally)

#42

Post by sudohackthenews »

do_not_redeem wrote: post #33
sudohackthenews wrote: post #30

Earlier quoted context omitted.

> > smaller distilled version Not technically the full R1 model, it’s talking about the distillations where Deepseek trained Qwen and Llama models based on R1 output

Then how about DeepSeek R1 GGUF: > Yes, you can run this model! Your system has sufficient resources (16GB RAM, 12GB VRAM) to run this model. No mention of distillations. This was definitely either made by AI, or someone picking numbers for the models totally at random.

Ok yeah that’s just weird

Re: Show HN: Can I run this LLM? (locally)

#43

Post by abujazar »

rkagerer wrote: post #39
abujazar wrote: post #35

Nice concept – but unfortunstely I found it to be incorrect in all of the examples I tried with my Mac. It'd also need to be much more precise in hardware specs and cover a lot more models and their variants to be actually useful. Grading the compatibilty is also an absolute requirement – it's rarely an absolute yes or no, but often a question of available GPU memory. There's a lot of other factors too which don't se…

I found it to be incorrect in all of the examples I tried Are you sure it's not powered by an LLM inside?

I believe it'd be more precise if it used an appropriately chosen and applied LLM in combination with web research – in contrast to juggling together some LLM generated code.

Re: Show HN: Can I run this LLM? (locally)

#46

Post by wbakst »

do_not_redeem wrote: post #28

> Can I Run DeepSeek R1 > Yes, you can run this model! Your system has sufficient resources (16GB RAM, 12GB VRAM) to run the smaller distilled version (likely 7B parameters or less) of this model. Last I checked DeepSeek R1 was a 671B model, not a 7B model. Was this site made with AI?

lol words out of my mouth

Re: Show HN: Can I run this LLM? (locally)

#49

Post by ggerules »

abujazar wrote: post #35

Nice concept – but unfortunstely I found it to be incorrect in all of the examples I tried with my Mac. It'd also need to be much more precise in hardware specs and cover a lot more models and their variants to be actually useful. Grading the compatibilty is also an absolute requirement – it's rarely an absolute yes or no, but often a question of available GPU memory. There's a lot of other factors too which don't se…

Confirmed. Nice idea but it doesn't really define"run". I can run some relatively large models compared to their choices. They just happen to be slow.

Re: Show HN: Can I run this LLM? (locally)

#50

Post by kennysoona »

jsheard wrote: post #34
do_not_redeem wrote: post #28

> Can I Run DeepSeek R1 > Yes, you can run this model! Your system has sufficient resources (16GB RAM, 12GB VRAM) to run the smaller distilled version (likely 7B parameters or less) of this model. Last I checked DeepSeek R1 was a 671B model, not a 7B model. Was this site made with AI?

> Was this site made with AI? OP said they "vibe coded" it, so yes. https://en.m.wikipedia.org/wiki/Vibe_coding

Goodness. I love getting older and see the ridiculousness of the next generation.
Post reply on HN