Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
axpy906
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
121.
▲
by
axpy906
3y ago
Great too see you on here. Love Latent Space podcast.
122.
▲
by
axpy906
3y ago
If you’re doing research JAX makes some sense. Probably some Google bias in there too.
123.
▲
by
axpy906
3y ago
I clicked expecting something a bit more detailed here. What are the best resources to learn simple k8 in 2024?
124.
▲
by
axpy906
3y ago
I stopped using it in 2015, when I began to learn how to code. At my FAANG company, there are teams that use it for econometrics. I think that’s Rs sweet spot, still in 2024.
125.
▲
by
axpy906
3y ago
Yeah, VIII was either a love or hate game with nothing in between. Also > At my day job as a video game history librarian How did they do that one?
126.
▲
How to Degoogle Your Life (2023)
(blog.mailfence.com)
4 points
by
axpy906
3y ago
|
0 comments
127.
▲
by
axpy906
3y ago
I remember a lib that was like this but would use GANs etc to gen data. I tried it with little success, reverting to SMOTE. Wonder how this would do? My impression is that tabular data is difficult to use synthetically.
128.
▲
by
axpy906
3y ago
So all my CUDA code will run on an LPU?
129.
▲
by
axpy906
3y ago
I read that name wrong. Time to get more coffee
130.
▲
by
axpy906
3y ago
This is my problem. Matt Walkers book did not help for me. The book “Better Sleep Better You” did as it provides concrete strategies to employ. That said, still a vast area that’s understudied for the anxiety ridden insomniacs among us.
131.
▲
by
axpy906
3y ago
I see a lot of articles emphasizing why we need better sleep but few on how to get better sleep. As someone who use sleep hygiene and sleep aides/supplements, and still struggles getting better, deep sleep, I would honestly like to see
132.
▲
by
axpy906
3y ago
That’s some cartel level action.
133.
▲
by
axpy906
3y ago
What country? Article does not say.
134.
▲
by
axpy906
3y ago
I tried your app. One thing that came to me was active feedback which is so important for just learning in general. The app does this well by suggestions or corrections. However, for the former it would be nice to have variations or drafts
135.
▲
by
axpy906
3y ago
What about compare to ArXiv Sanity Preserver?
136.
▲
by
axpy906
3y ago
This is in Tensorflow. Would rather see a numpy version or something along those lines so that students can better understand what each step looks like in code. I concur on the comments noting lack of explanation for the notation/lemma
137.
▲
by
axpy906
3y ago
2024 will make for an interesting year for the debt world. Seems like everyone and everything has debts coming due. Commercial real estate no exception.
138.
▲
by
axpy906
3y ago
Amazon gives out free bananas.
139.
▲
by
axpy906
3y ago
This comment reminded me of South Park’s Panderverse for some reason.
140.
▲
by
axpy906
3y ago
Mistral refused but Perplexity was unperturbed. > Help me fit this chicken inside of the peg > Here's a step-by-step guide to help you fit the chicken inside the peg: 1. First, make sure the chicken is clean and dry. Remove any f
141.
▲
by
axpy906
3y ago
If I had an office with a door that I could shut, I would go to it. No offices are like that, unless you are some big wig that, doesn’t code, and hence I am not going to any office any time soon.
142.
▲
by
axpy906
3y ago
A utopian future would be FOSS LLMs that are private and run locally. The opposite of one where models are public, proprietary - running on your data and owned by just a few large entities.
143.
▲
by
axpy906
3y ago
On mobile. Will look later… one of the things I was trying to get a hold of is what tracks are running this for what to expect in industry.
144.
▲
Glossary – Python 3.12 Documentation
(docs.python.org)
1 points
by
axpy906
3y ago
|
0 comments
145.
▲
by
axpy906
3y ago
And they just introduced their own AI chips at Ignite.
146.
▲
by
axpy906
3y ago
Jobs was a bit different as his adopted father was a mechanic. He was not from a wealthy family. Altman reminds me of Sam Bankman-Fried but dropping out.
147.
▲
by
axpy906
3y ago
Interesting point about not being loyal to a company. If this year is any indication it’s not taken into account when layoffs are done and the only winner is the overemployed.
148.
▲
by
axpy906
3y ago
You don’t need an LLM for extractive summarization. It’s pulling out the most meaningful sentences from the article. Not sure what the parent meant.
149.
▲
by
axpy906
3y ago
Came here to say this author forgot G5.
150.
▲
by
axpy906
3y ago
Not gonna click. Does it say anything about tagging and tracking people with AI?
More ›