curl-up

20 ms·

61.
by curl-up 3y ago
Very interested to see where this goes. Good luck!
62.
by curl-up 3y ago
True across Balkans. There are some fruits (e.g. apricots) which have a highly variable yield from year to year, so eating a 5+ year old jam (from a year with exceptionally high yield) is not strange.
63.
by curl-up 3y ago
My point is that it is silly to complain about people being nerdy about a specific topic, just because there are others with more expertise. Especially in cooking, the concept of "inventors" and "authorities" for such ba
64.
by curl-up 3y ago
Next time, instead of traveling, just ask someone who's already been there for a summary.
65.
by curl-up 3y ago
There is no system prompt in Llama, just as there isn't one pre-set for GPT 3.5/4 either. There is a system prompt set for ChatGPT (just as you can set your own for GPT 3.5/4 if you're accessing the model directly), but
66.
by curl-up 3y ago
TOS of OpenAI API (which tools like this use) do not allow for model training on the data either. You might be confusing their API with ChatGPT, which has a different policy.
67.
by curl-up 3y ago
Yellow/beige/brown spheres for me as well. Other commenter had spheres also. I never even thought that this might be something others experienced. Sadly, trying to find more information on this only delivers "dream interpreta
68.
by curl-up 4y ago
Congrats on the launch! If it's not a secret, where do you get the job postings from?
69.
by curl-up 4y ago
Where in the API docs is this described?
70.
by curl-up 4y ago
Thanks! In the "modern setup", I guess this maps to queries being in PowerBI (e.g. DAX queries) or Tableau etc. Maybe someone working with these tools on a large scale can confirm this.
71.
Ask HN: How do you organize your analytics/BI SQL code?
3 points by curl-up 4y ago|3 comments
72.
by curl-up 4y ago
Great work! In your medium post, in the Post-processing section (step 1) you wrote: "If so, we re-run the evaluation prompt and force content_contains_answer to be false." I assume you do this forcing by adding 'false' t
73.
Ask HN: How do you organize your analytics/BI SQL code?
3 points by curl-up 4y ago|1 comments
74.
by curl-up 4y ago
Sadly, no, that doesn't seem to be the case. They haven't pulled the plug, just made it harder to break it seems (probably through fine tuning), but you can still do it. And whem you do, it's clear all it does is "halluc
75.
by curl-up 4y ago
> one thing is perfectly clear: art is human. Genuine question: what would change your mind on this? Or is "human-made" part of your definition of the word "art", and as such there is nothing that would change your mi
76.
by curl-up 4y ago
I would have said 3-4 years as well, and I would have been certain beyond any doubt that it was more than 2 years. Crazy.
77.
by curl-up 4y ago
Yeah, agree - building and verifying apps built on this models is very hard as they are hard to test extensively. But so are most other ML-based apps. The error in that example was on my side, not on the model. I mean, to me this seems simi
78.
by curl-up 4y ago
Ha, do I feel stupid now! Just checked, and seems like I misconfigured my system (the part that builds the Codex prompt from the schema info) - Invoice and InvoiceLine tabls got confused. After fixing it now, it works as expected (using IL.
79.
by curl-up 4y ago
Please point me to an "old" model that was able to do something similar to my example, and was general enough to do that not only for any custom schema, but for basically all often used languages as well (both natural languages a
80.
by curl-up 4y ago
Can you elaborate? To me (but I know very little about it) it seems like part of the incremental progress in the internet availability. What am I missing?
81.
by curl-up 4y ago
Check out the SQL example I posted below. If you're interested, I'd be happy to post more. To me, this is not about "is the machine accurate enough already". Maybe it isn't and it needs to mature. But the door has b
82.
by curl-up 4y ago
This is what amazes me, since it seems like such big news, and people in the field are just not aware of it. Just for reference of what I am talking about, here is a piece of code that was generated, without any cherry picking at all (you j
83.
by curl-up 4y ago
What I am talking about has nothing to do with Intellisense or your workflow. What I am saying is that, if someone in 2019 told you that there is a "thing" that is able to take a very complex sentence and qith high accuracy (and a
84.
by curl-up 4y ago
How come general developer audiences aren't more acquainted with GPT-3 (and Codex in particular) capabilities? People in the twitter thread all seem completely mind blown over an app that basically just passes your code to an existing
85.
The strange appeal of garden lawns(bbc.com)
3 points by curl-up 4y ago|0 comments
86.
by curl-up 4y ago
Mirror neurons seem to be the primary mechanism of learning from other members of the group. So in a sense, every species able to learn from others (by replicating the behaviour) would automatically also be able to identify the "self&q
87.
by curl-up 4y ago
I've always wondered, but could never find good sources on this: couldn't MSR results simply be produced by mirror neurons firing? I.e., touching a marked part of the body does not mean "Animal in the mirror is me", it c
88.
by curl-up 5y ago
Very happy to have been able to help, and thanks for broadcasting my post :) Your theory actually sounded very plausible to me, which is why it peaked my interest so much. I hope we're able to "crowdsource" and inspect future
89.
by curl-up 5y ago
Check my other, more detailed comment - farm's address is rarely in a place where farm is actually located. Farm land is just an asset of a company, which is registered elsewhere. If this is a small, family owned business, it makes sen
90.
by curl-up 5y ago
OP, why do you think a farm business (which gmaps are showing the location of) should be in the same place where farm land is? These simply seem to be the addresses where a "farm business" is registered, as can be confirmed on a c

More ›