turmeric_root

5 ms·

1.
by turmeric_root 3y ago
> Isn't bragging the main thing people do on Social Media? start following more interesting people
2.
by turmeric_root 3y ago
whisper doesn't seem to support diarization (identifying when the speaker changes), which is needed for subtitle formatting.
3.
by turmeric_root 3y ago
c'mon just write a function that takes in text and tells you whether or not it's true, how hard could it be
4.
by turmeric_root 3y ago
i think it'd be cool if you could make art and still own a house
5.
by turmeric_root 3y ago
> When users join a sub it's because they like the content and tacitly approve of the moderation. I disagree, I see tons of people blaming admins for moderation actions and vice-versa.
6.
by turmeric_root 3y ago
clearly the arbitration panel ruled against Lindell because it's part of the deep state. /s
7.
by turmeric_root 3y ago
Windows reserves a certain percentage for VRAM for some reason. So I'd recommend Linux. Or find a way to disable the desktop/UI in Windows.
8.
by turmeric_root 3y ago
exactly! I gave a heartfelt letter to my shredder the other day and it simply destroyed it. issues like these are why AI alignment research is so critical.
9.
by turmeric_root 3y ago
Have you tried writing code on a smartphone?
10.
by turmeric_root 3y ago
this seems like it might be useful for arguing on HN
11.
by turmeric_root 3y ago
Worldcoin price skyrockets => cryptocurrency speculation returns => GPU shortage prevents further training of LLMs
12.
by turmeric_root 3y ago
Hell, what about Skynet?
13.
by turmeric_root 3y ago
The model weights were only shared by FB to people who applied for research access. Github repos containing links to the model weights have been taken down by FB.
14.
by turmeric_root 3y ago
I like using them for memeing
15.
by turmeric_root 3y ago
More VRAM => larger models. IME it is absolutely worth maxing out VRAM for the significant improvement in quality, especially with LLaMA (though even with a 4090, you won't be able to run the largest 65-billion parameter model even
16.
by turmeric_root 3y ago
they're just microdosing it's ok
17.
by turmeric_root 3y ago
A lot of the 'look what I made with AI' images that get shared around also don't include the creator's workflow. There's usually lots of trial-and-error, manual painting/inpainting, multiple models involved etc
18.
by turmeric_root 3y ago
ugh, that's so shitty. so many people in this space seem to be absurdly demanding and angry at devs, but one thing I've noticed is that every text AI project discord I've hung out in has this sleazy, obsessive 4chan /g&#
19.
by turmeric_root 3y ago
> the "number B" stands for "number of billions" of parameters... trained on? No, it's just the size of the network (i.e. number of learnable parameters). The 13/30/65B models were each trained on ~1.4
20.
by turmeric_root 3y ago
'accuracy' and 'truth' are legacy 0.1X concepts, move fast and break things
21.
by turmeric_root 3y ago
yeah when getting DL up and running on AMD requires using a datacentre card then it's no wonder CUDA is more popular. AMD is enabling ROCm for commercial GPUs now but it's still a pain to get it up and running, because of the iner
22.
by turmeric_root 3y ago
if the AI is trained on LW then I think we'll be safe, just use the word 'woke' and it'll lose its shit and get stuck in an endless loop of telling you why it's not actually racist
23.
by turmeric_root 3y ago
though unless you've disabled sampling it will be difficult to determine how prompts affect the output, these could just be due to RNG
24.
by turmeric_root 4y ago
Yep.
25.
by turmeric_root 4y ago
I disagree with the linked post, most people use 'REST' to refer to JSON-over-HTTP now.
26.
by turmeric_root 4y ago
Yeah I spent a week or two getting excited playing with ChatGPT but then I got bored. I also bought a Quest 2 a while ago and sold it after a few months, so I guess the novelty just wears off quickly for me.
27.
by turmeric_root 4y ago
It seems to be about as good as gpt3-davinci. I've had it generate React components and write crappy poetry about arbitrary topics. Though as expected, it's not very good at instructional prompts since it's not tuned for inst
28.
by turmeric_root 4y ago
So since making that comment I managed to get 65B running on 1 x A100 80GB using 8-bit quantization. Though I did need ~130GB of regular RAM on top of it.
29.
by turmeric_root 4y ago
> so does this mean you got it working on one GPU with an NVLink to a 2nd, or is it really running on all 4 A40s? it's sharded across all 4 GPUs (as per the readme here: https://github.com/facebookresearch/llama
30.
by turmeric_root 4y ago
the 7B model runs on a CUDA-compatible card with 16GB of VRAM (assuming your card has 16-bit float support). I only got the 30b model running on a 4 x Nvidia A40 setup though.

More ›