Viewing profile — thewataccount
thewataccount
HN member- Joined
- Thu, Sep 01, 2022, 6:40 PM UTC
- HN karma
- 1,110
- Public activity
- 365 items
- HN profile
- View on Hacker News ↗
About thewataccount
No profile information was provided.
Recent public activity
-
comment
Comment #37710156
It's not the name registrar it's the dns server. You just use a wildcard dns record, and then the webserver uses the host header to determine which subdomain was used (for webserve…
-
comment
Comment #37709156
To my knowledge planes don't use the gps for very much, and AFAIK they use the barometric altimeter for maintaining their flight level. A controllable, plane that can maintain alti…
-
comment
Comment #37708217
My understanding is that would be covered under medical insurance?
-
comment
Comment #37679631
From my understanding many of the best players immediately look down to tell what "generation streetview car" they're using, and seem to know what continents/times they're from.
-
comment
Comment #37647349
From my understanding they're basically an extension of FIDO? You can just use a yubikey unless I'm mistaken?
-
comment
Comment #37617789
This is simplified a bit - It's just a "machine" that maps [set of inputs] -> [set of probabilities of the next output] First you define a list of tokens - lets say 24 letters beca…
-
comment
Comment #37590278
I've always wondered what percentage of "users" are "lurks" who never login, occasionally check a profile or click on a tweet on another website. One of the previous uses of twitte…
-
comment
Comment #37588473
Yeah I'll be interested to see how much you have to change the prompts to get similar styles
-
comment
Comment #37588381
> Because there's no way to control the seed, a direct comparison (using a before/after slider, for example) probably wouldn't make sense. Even if it was the same seed, from my und…
-
comment
Comment #37561673
I'm not sure what you mean - the last security update fixed massive 0-day which was an arbitrary code execution caused by the image decoder. CVE-2023-41064 AFAIK every app that use…
-
comment
Comment #37561548
I've been struggling with figuring out a good dataset for fine-tuning. Most of the ones that exist were purpose made for finetuning/training a model already. Does anyone have any t…
-
comment
Comment #37559927
FWIW that's 5 years of feature updates. They still do security fixes for iOS 15 which is the iphone 6S in 2015 so 8 years. The last security update for iOS 12 was 7 months ago too.…
-
comment
Comment #37559090
I haven't checked an email but using the completions api I do get a response using model "gpt-3.5-turbo-instruct". Interestingly the completions api is deprecated though? Are they …
-
comment
Comment #37527301
Eh I had some issues with microsoft's "guidance", eventually got it working with an older version. They haven't updated it in a month too. lmql might be a decent alternative. Any f…
-
comment
Comment #37527006
I've already had some success with Llava although existing image classifiers are still going to outperform it.
-
comment
Comment #37526927
From my experience "single prompt classification" isn't as simple as "type in sentence and it works" in practice. But you can use some methods to massively improve it's consistency…
-
comment
Comment #37512214
Yes, I wasn't sure if parent was suggesting they switch to django, or simply "using what facebook learned" - I didn't want to imply that there's no way Mastodon devs couldn't use C…
-
comment
Comment #37512120
Just to clarify - are you wanting the LLM itself to identify what a "anomalous latency" would be based on the data itself? If so then I don't think this will help you at all until …
-
comment
Comment #37510801
(I use django in production myself, nothing against it) It's probably worth noting Meta made their own JIT'd version of CPython just for Instagram/Threads. Maybe mastodon devs can …
-
comment
Comment #37501813
Openai - GPT2 2019 - https://openai.com/research/gpt-2-1-5b-release Google - T5 - Feb 2020 - https://blog.research.google/2020/02/exploring-transfer-lear... Both of these were and …
-
comment
Comment #37488572
For sure, and yeah it wouldn't be terrible you're right. You'd just need the api servers + a load balancer. My thing is that dynamically doing that is still a lot compared to just …
-
comment
Comment #37487879
> That's an exercise left to the reader for now, and is where your value/moat lies. Hopefully more on-demand services enter the space. Currently where I am we don't have the resour…
-
comment
Comment #37487543
What's the best way to use LLama2-70b without existing infrastructure for orchestrating it?
-
comment
Comment #37486800
You (currently) need a GPU to run any of the useful models. I haven't really seen a business use-case that runs it on the user's computer, but given the hardware requirements it wo…
-
comment
Comment #37484710
I happy for them that they were hired! I will say though, from my experience (with other software) what usually happens is the original plugins/software get's abandoned or neglecte…