Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
mkaic
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
28 ms
·
361.
▲
by
mkaic
4y ago
Ah, a fellow Honeycrisp devotee. Always good to meet another member of the Fellowship of Delicious Apples. Have you tried the newest engineered Honeycrisp successor, Cosmic Crisp? Imagine a Honecrisp, but twice the size, twice the price, an
362.
▲
A terrible guide to the terrible terminology of U.S. Health Insurance [video]
(youtube.com)
2 points
by
mkaic
4y ago
|
0 comments
363.
▲
by
mkaic
4y ago
from the HN guidelines: What to Submit On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one'
364.
▲
by
mkaic
4y ago
Hooray, can't wait to buy a 3090 once the prices fall even further than they already have and completely ignore these new, overpriced cards that draw half a kilowatt! I'm more excited about the Hopper server-class cards — not beca
365.
▲
Psychopath Renderer – Notes on Color: 01
(psychopath.io)
1 points
by
mkaic
4y ago
|
0 comments
366.
▲
by
mkaic
4y ago
I recently got into Blender addon development and have been really pleasantly surprised with how extensive its Python API actually is — it's not perfect yet, but it's certainly very very useful! My personal wishlist for the future
367.
▲
Reverse-Engineering Blackmagic Camera Color Spaces
(psychopath.io)
1 points
by
mkaic
4y ago
|
0 comments
368.
▲
by
mkaic
4y ago
I'm no expert, but this paper[0] has piqued my interest recently. From what I can kind of piece together, it discusses treating the latent space as a manifold from differential geometry, and creates a Riemannian metric for measuring di
369.
▲
Building a Better LK Hash for Owen Scrambling in Psychopath
(psychopath.io)
1 points
by
mkaic
4y ago
|
0 comments
370.
▲
by
mkaic
4y ago
> Building something from scratch is rarely a good idea. In the context of this article, I agree. Professional products should probably not try to build everything from scratch. But I do think it's important to acknowledge that bu
371.
▲
by
mkaic
4y ago
Pyramid Scheme Pro Max Plus Ultra 2022 v2.0: Electric Boogaloo
372.
▲
by
mkaic
4y ago
I'm currently working on tools for better latent space exploration in generative models. The latent space of Stable Diffusion for example, is incredibly rich, and the traditional txt2img and img2img pipelines for accessing it are only
373.
▲
by
mkaic
4y ago
"In my client's defense, your honor, they were merely following essential protocol and delivering precious cargo . I now present exhibit A, the Double-Double — add mustard, add chopped chilies, add raw chopped onions — with a fry
374.
▲
by
mkaic
4y ago
GPUs are already flooding the market. 3090s are up for ~900 bucks on ebay at the moment which is 25% below their MSRP, and I strongly suspect it's miners dumping their stock.
375.
▲
by
mkaic
4y ago
> What Went Wrong? It's in the title of the article. It's a "Blockchain Game". That's what went wrong. It was a flawed concept from its inception. Blockchains are pretty cool from a math/cryptography standpo
376.
▲
by
mkaic
4y ago
Just filed a complaint! I didn't know this form existed, as I've never genuinely wanted to file an antitrust complaint before, but there's a first time for everything and I despise Adobe so it gets the honor of being in my
377.
▲
by
mkaic
4y ago
As someone who's working on a project right now that's full of whimsical/cryptic names (project is called Sandman, modules within the project include Mystic, Hypnos, Lethe, Morpheus, Dreamer, and Lucid), this was validating t
378.
▲
by
mkaic
4y ago
I expect short video generation to be at the level of DALLE 2 in 1 year — e.g. able to get the gist of a prompt, but with lots of artifacts, requiring a lot of compute, and frequently ignoring large parts of the prompt.
379.
▲
by
mkaic
4y ago
This 1000%. I'm currently working on my own slightly more complex/customizable interface for Stable Diffusion because I firmly believe that the value lies not in simple prepackaged text-to-image, but in the wealth of information,
380.
▲
by
mkaic
4y ago
This is genius and if I ever get into this field I am 100% stealing it.
381.
▲
by
mkaic
4y ago
This is a really big announcement. Kudos to Meta. I use PyTorch daily at work and am very happy with both its current state and where it appears to be headed. PyTorch makes deep learning accessible, understandable, and portable, and I love
382.
▲
by
mkaic
4y ago
90%. I'm not qualified in the slightest to make that prediction but I'm making it anyways!
383.
▲
by
mkaic
4y ago
This is so fun! Spent 10 minutes just jamming out. I think the scale and time quantizations are what make it truly fun to use, because it's almost hard to make anything sound bad! Love it.
384.
▲
by
mkaic
4y ago
I don't personally see much difference between how I trained myself to be a portrait artist vs. how diffusion models do. In order to learn to draw stylized portraits, I looped over: 1. Find a photo of a person as reference 2. Create po
385.
▲
by
mkaic
4y ago
My best guess is that I just did something wrong lol. Autocast seems to convert most of the model to fp16 anyways and that works great, so I'll just keep using that!
386.
▲
by
mkaic
4y ago
All art is derivative and there's no such thing as originality. Every human artist draws inspiration from their visual and emotional experiences, copyrighted or otherwise, how is this different? If I watch Star Wars and then make a spa
387.
▲
by
mkaic
4y ago
hmm -- I tried converting to float16 just using a naive model.half() call and saw some quality degradation in my images compared to just autocasting parts of the model to float16 while leaving others at float32. Curious if anyone else has h
388.
▲
by
mkaic
4y ago
Yes, that's actually the biggest reason this is such a cool announcement! You just need to download the model checkpoints from HuggingFace[0] and follow the instructions on their Github repo[1] and and you should be good to go. You ba
389.
▲
by
mkaic
4y ago
oh heck, you're right! I edited the original comment, thanks.
390.
▲
by
mkaic
4y ago
(OP here) I agree. I am an artist (not by trade but by lifelong obsession) in several different mediums, but also an AI engineer--so I feel a weird mixture of emotions tbh. I'm thrilled and excited but also terrified lol.
More ›