Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
tartakovsky
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
9 ms
·
31.
▲
by
tartakovsky
3y ago
This reminds me of this article for which symbol fine tuning helps overcome: https://arxiv.org/abs/2303.03846 Basically they show that larger models have an easier time giving up their semantic priors, so that, in the
32.
▲
by
tartakovsky
3y ago
Is this different from https://langchainx.web.app/ ? How?
33.
▲
by
tartakovsky
3y ago
Made me laugh because it's true, funny. Well done!
34.
▲
by
tartakovsky
3y ago
I’m just saying, it was more a hopeful nod in the direction of wishful thinking. Not a joke, just a bit of having my head in the sand.
35.
▲
by
tartakovsky
3y ago
Wow! I don't know how accuracy translates, I do see charts that look strong but unless I'm missing something, this is incredible. Would be curious and also terrified to see an endpoint so I can play around with it. I thought we we
36.
▲
by
tartakovsky
3y ago
Just when OP wrote: 'Griptape can be thought of as "Airflow for LLMs," providing an alternative to the agent-based LangChain approach.'
37.
▲
by
tartakovsky
3y ago
This looks cool. I am sort of unclear about this contrasting statement to Langchain when it seems you are building on too of / or using Langchain to achieve some functionality. Is that incorrect? Also, I looked at your repo to contribu
38.
▲
Ask HN: Is there a photo curation tool for Dropbox?
2 points
by
tartakovsky
4y ago
|
1 comments
39.
▲
by
tartakovsky
4y ago
Amazing.
40.
▲
by
tartakovsky
4y ago
Ok so can someone provide 10 steps to launching your own voice assistant?
41.
▲
Ask HN: Why can't I find an article on the propaganda of “Soviet” Inventions?
8 points
by
tartakovsky
4y ago
|
4 comments
42.
▲
by
tartakovsky
4y ago
Hi Hayk, I see that the inference code and the final model are open source. I am not expecting it, but is the training code and the dataset you used for fine-tuning, and process to generate the dataset open source?
43.
▲
by
tartakovsky
4y ago
What does “exclusive” mean? “Exclusive AI products…”
44.
▲
by
tartakovsky
4y ago
Thanks, yes I see it now. It stores the notebook, which is a json, as a value in the localStorage object, with the filename as a key. Interesting, yes, not quite easy to configure (though I haven't read through the documentation or co
45.
▲
by
tartakovsky
4y ago
Files seem to persist even after quitting chrome, refreshing or closing the tab. I'm not an expert, but assume it's cookie storage.
46.
▲
by
tartakovsky
4y ago
This GitHub Actions + GitHub Pages feature looks like the use case I would be looking to exploit for educational purposes. From this: https://github.com/jupyterlite/demo you can easily get to https://jupyter
47.
▲
by
tartakovsky
4y ago
No wonder it doesn’t load.
48.
▲
by
tartakovsky
4y ago
Facebook for actual one on one friendships sounds like overkill. Facebook for a small business to have a website sounds fine. And keeping in touch with large swaths of people who live far from you can be tricky, I guess you can use Facebook
49.
▲
by
tartakovsky
4y ago
I agree, it is sad and HN has a tradition of not condoning this kind of comment, :shrug:. I tried to call it out, too.
50.
▲
by
tartakovsky
4y ago
@BurningFrog, your comment itself smacks of bigotry, ironically. It seems to denigrate rather than enlighten and provide a solution. Also nowhere in the parent comment does it state that the HR employee themselves was not a POC (person of c
51.
▲
by
tartakovsky
4y ago
I agree with this comment. I think as we condition from profession to team to company culture - assuming there is anything causal with WFH - some may be more productive, while others less so.
52.
▲
by
tartakovsky
4y ago
Just because profits are higher after the occurrence of working from home, doesn’t mean it is because of it. This is a common logical and statistical fallacy to attribute causality from observation.
53.
▲
by
tartakovsky
4y ago
Hi @seanhunter, any chance you have a source for this?
54.
▲
by
tartakovsky
4y ago
Zero diagrams, but maybe they wouldn’t be helpful to clarify the concept? Guess it depends on the types of learners, I’m not sure.
55.
▲
by
tartakovsky
4y ago
Can we learn permutation matrices? Something like this but applied to learning order that maximizes some objective. Off the cuff it sounds like an NP problem but seems that people have tried before with some success. https://arxi
56.
▲
by
tartakovsky
4y ago
I would also be interested in understanding whether there is a proper pricing model for such things. Wordle comes to mind. Or a friend that has an IPad app that took 2 years to build that is something novel but not released. Some projects a
57.
▲
by
tartakovsky
4y ago
The goal here seems to be twofold. On the one hand, to put the onus on the one receiving messages from the troll or heckler to either stay silent and not give them a voice, or respond and allow the world to see the negative comment. On the
58.
▲
by
tartakovsky
4y ago
Really interesting concept. I think the previews on the thumbnails shouldn't be the beginning of the message. My reason is that it oftentimes contains longwinded introductions and I believe there could be more compelling text. Maybe th
59.
▲
by
tartakovsky
4y ago
What does “good” mean here? This is just one general way to arrive at an answer that doesn’t require much thinking. Who knows if it’s the best even if it is the one that requires the least thought. The past is not always a reliable moral co
60.
▲
by
tartakovsky
5y ago
Makes me think of … Asperger’s.
More ›