Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
amorroxic
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
1.
▲
by
amorroxic
4mo ago
Curious on what backs this assertion. As a counterpoint we’ve been running 200+ models in production for more than 5 years - language models, embedding, classifiers, low tens to hundred M params. Traffic in the order of 1-2M requests/d
2.
▲
by
amorroxic
10mo ago
Thanks a lot for this. Also one question in case anyone could shed a bit of light: my understanding is that setting temperature=0, top_p=1 would cause deterministic output (identical output given identical input). For sure it won’t prevent
3.
▲
by
amorroxic
1y ago
A real-time video processor in hardware: - static (mp4) or live (hdmi) input, hdmi output - a collection of effects (various distortions, color glitches, demoscene-like) applied over incoming video stream, maximum of 4 effects stackable on
4.
▲
by
amorroxic
3y ago
Maybe this could be useful? https://www.playonlinux.com/en/ (able to run Photoshop myself)
5.
▲
by
amorroxic
3y ago
first $100: pdf photography courses with weekly chapters first $1000+: built a shop selling silk scarves (100-odd) I collected myself from Laos/Cambodia/Thailand to balance trip budget (25yo back then, mad travelling). Sold all sc
6.
▲
by
amorroxic
4y ago
the null concept being referred to as “the billion dollars mistake” - until discovering optionals and gradually opening up to FP, category theory, etc
7.
▲
by
amorroxic
5y ago
This. Same scenario here, fondly remember 1080ti, that was the workhorse for us alongside cloud training (startup with ~100 nlp prod models, ner, siamese, count models, etc). ULMFiT and language transfer was the moment when upgrades felt ne
8.
▲
by
amorroxic
5y ago
and unfortunately social makes it possible for one to yell “Fire” in all crowded places at the same time.
9.
▲
by
amorroxic
5y ago
it’s still there unfortunately, no close button either.
10.
▲
by
amorroxic
5y ago
Signed it too :) Can’t remember a particular reason though in our case (former cofounder) we were discovering the actor model, immutability, streams,.. tbh remember a great deal of fun programming during those days, insane schedules and wha
11.
▲
by
amorroxic
5y ago
Three strikes and you’re banned forever, better choose the moment well.
12.
▲
by
amorroxic
6y ago
Thank you for the perspective here. A wish there was a way to bookmark comments on HN..
13.
▲
by
amorroxic
6y ago
Agree, this being a cyber-fraud company I believe one could keep a team busy for a long time focusing on building secure systems rather than operating them to the point where no-one gets to see what's actually being stored within. The
14.
▲
by
amorroxic
6y ago
In the US maybe, gig economy is global though.
15.
▲
by
amorroxic
7y ago
Man, don't beat yourself up. Americans are great people and not that bad of a society now. (someone in a WE society). Hang in there, we'll bicker after :)
16.
▲
Insights on OS X platform support for Tensorflow/GPU?
1 points
by
amorroxic
7y ago
|
0 comments
17.
▲
by
amorroxic
7y ago
Genuinely interested in good numbers to compare then, any chance to elaborate?
18.
▲
by
amorroxic
7y ago
You could have a look at ArangoDB ( https://www.arangodb.com ), a multi-model database - it may cover your use cases. Hopefully linking is ok in comments, will remove if its breaks community rules - no affiliation etc..
19.
▲
by
amorroxic
8y ago
What's the point of a random comment on the internet when there are far more efficient ways to spend one's time?
20.
▲
by
amorroxic
8y ago
thanks for this perspective.
21.
▲
by
amorroxic
8y ago
Same, Focus is great. Also a very warm mention for what I think became one of my favourite features while browsing - one single tab.
22.
▲
by
amorroxic
8y ago
Involved in development with pretty insane schedules, 14-18+ hrs/day pretty often and weird as it sounds enjoying my activity immensely. Marijuana keeps me focused at times without a single Alt+Tab for most of the day. Far from ideal o
23.
▲
by
amorroxic
8y ago
Sentence similarity were my explorations with WMD too, reached a setup in Keras with a siamese configuration, Wasserstein + KL loss (have a known vocabulary and feeding both word vector sequences as well as their LDA distributions as input)
24.
▲
by
amorroxic
8y ago
So look. 3 hours after you wrote this (now, that is) you probably forgot your comment (whether a random HN project becomes a business or not). Here's my take for an answer - it doesn't matter :) It probably touched base for everyo
25.
▲
by
amorroxic
8y ago
I like it, really! I believe this to be a good use-case for decentralization as opposed to anonimity - if paired with a good verification mechanism (so leaving behind the crypto-anonimity mantra) this would also guarantee integrity of timel
26.
▲
by
amorroxic
8y ago
> Or Google and Facebook will be held to a higher standard of compliance. Than whom else? VK? Baidu? ICQ?
27.
▲
by
amorroxic
8y ago
This would be a valid point or topic of debate should the subject be an actual law (the 70 years copyright law for instance?) What we're dealing with here is loopholes and nothing more under various incarnations - take patents in the U
28.
▲
by
amorroxic
8y ago
European here and I feel this is a bit shortsighted. Not really sure the sentiment is to target US companies (if for nothing else ours are quite open/interconnected economies and there's lots of fondness for US services - check bo
29.
▲
by
amorroxic
9y ago
On the same page. Have gone through endless flavors of Linux, Win and OSX by simply copying the emails folder from system to system for close to 20 years now - and never lost one email. To say I appreciate this piece of software would be an
30.
▲
by
amorroxic
9y ago
Props to you guys, outstanding product really.
More ›