mdp2021

3 ms·

1.
EU KIDS Act Won't Keep the Internet Accountable and Trustworthy(eff.org)
4 points by mdp2021 5d ago|0 comments
2.
Scanning Light Field Display [3D rotating display] [video](youtube.com)
1 points by mdp2021 5d ago|0 comments
3.
Preschoolers Who Blink More Often Show Stronger Self-Control, Study Finds(studyfinds.com)
2 points by mdp2021 6d ago|0 comments
4.
by mdp2021 6d ago
That suggests that 16GB RAM will not be enough. In fact, like it appears in the reports above, it is "7b" as in > 7B parameters in its visual generation component It seems they calibrated the size to fill a 16GB VRAM near the
5.
by mdp2021 6d ago
> I mostly think that that's a weird collection of things to all be at once to a murdered son Were you joking, or you never heard it? «My name is Maximus Decimus Meridius, Commander of the Armies of the North, General of the Felix
6.
by mdp2021 6d ago
You could try attaching other images as references (I think you can attach a maximum of 10 images). If the attachments can be blurred or sketchy or generic enough, they could be used for generalization.
7.
by mdp2021 6d ago
Or trying to install the whole of CUDA on machines that do not even have a GPU (not Nvidia, not anything past the embedded)... Yep, that's (also) what I meant ;) Lean, efficient... Also sensible and trouble-less.
8.
by mdp2021 6d ago
Thank you! Can you please check how much RAM does it consume (and require)?
9.
by mdp2021 6d ago
Not all architectures are supported by llama.cpp . The GGUF format encodes the NN in a standardized way, but then you need code that can use that NN structure. I understand that llama.cpp could only output text, last time I checked (I do no
10.
by mdp2021 6d ago
> Diffusers, ComfyUI, vLLM-Omni, SGLang, and LightX2V I think that's all Python (not a direct executable). You could just do (see the "Quick Start") four `pip install` and have a dozen lines script to generate the image.
11.
by mdp2021 6d ago
> finally fixed their VAE Can you share the sources?
12.
by mdp2021 6d ago
I am not sure that llama.cpp also supports image generation models.
13.
by mdp2021 6d ago
How do you use this model locally, similarly to using `llama-server -m <model>`? (I mean: outside direct or substantial use of Python, and running the Neural Network in the most efficient way.)
14.
by mdp2021 6d ago
Do you also feel «forced and unnatural» "Il est médecin // il est un médecin // il est le médecin"? "Él es médico // él es un médico" // él es il médico"? Are you feeling thro
15.
by mdp2021 6d ago
The attempts to end anonimity and to end cash flow are ideological and pretty surely did not come from the USA.
16.
by mdp2021 6d ago
The fault in that comment is to be unsupported by arguments. Otherwise, there would be no divisiveness in fairly assessing measures in Economics.
17.
by mdp2021 6d ago
> fastest growing European country by many metrics Countries grow in secondary matters and the primary ones (civilization) remain unattended. And by the way, when something "grows fast", do not misunderstand the indicator...
18.
by mdp2021 6d ago
> Similarities between languages If they are substantial instead of superficial. > This comment just explains ... idiom in terms of the logic behind it. Which is common to many languages.
19.
Laser Your Way into Debug Mode on the RP2350(hackaday.com)
4 points by mdp2021 6d ago|0 comments
20.
by mdp2021 6d ago
It is not the set-pertinence that would get transformed into a plural.
21.
by mdp2021 6d ago
You should not look at superficial similarities between languages, but at logic. To the question: "What is he?", the answer can be "He is married, tall, doctor in medicine". On another point of view, "He is a doct
22.
by mdp2021 7d ago
> always wild If they do not have a developed prefrontal cortex acting as a filter from instinct to action, they will be nice or calm for a long time - until the whim occurs and they will act violently. Maybe just for a fraction of a s
23.
by mdp2021 7d ago
> Memory prices rose sevenfold over the last 12 months, as reported by PC Watch The whole consequences of this trend have not been explored and clarified in full. Part of the demand, possibly a numerous set, will not be able to access
24.
Lawsuit: Illegal Agreement of Anthropic, OpenAI, SpaceXAI, Google on AI Slowdown(independent.co.uk)
18 points by mdp2021 7d ago|1 comments
25.
by mdp2021 7d ago
And how much would it cost? In the current times performance/price seems to be increasingly important. (Not to mention several other factors that once did not even exist, or were irrelevant.)
26.
by mdp2021 7d ago
Probably internal tests? The software identifies the hardware, performs the benchmarks, and sends the heading and results to the Geekbench DB. Somebody who had the machine at hand must have run it.
27.
by mdp2021 7d ago
That of HN is a largely non-random public, not a fully selected public. Of course Paretian distributions are also valid here. (Edit: and it is evident in this very page.)
28.
by mdp2021 7d ago
The submitter reduced the complexity of the situation to a whim, and misrepresented it. Reality is what it is, some heavier some lighter: we discuss reality, its details ranked by weight.
29.
EFF Statement on California Governor's Executive Order on AI(eff.org)
13 points by mdp2021 7d ago|0 comments
30.
by mdp2021 7d ago
> model [to] predict future The attempt is already there: "predicting the future" is called "intelligence".

More ›