make3

136 ms·

931.
by make3 7y ago
Most of the field is moving to PyTorch because of what the author is describing. PyTorch is super nice by all accounts.
932.
by make3 7y ago
very few linguists are prescriptivists. Grammatical rules have always (only) emerged from use, and language is expected to evolve through the decades.
933.
by make3 7y ago
stylised way of asking how long it took
934.
by make3 7y ago
it arguably moved the other way (to infrastructure not being a part of the game at all), as MOBAs have taken over RTS
935.
by make3 7y ago
it does make it clear to me on MacOs ? I haven't noticed that at all
936.
Brow.sh: Modem Text Browser That Supports CSS, JavaScript and Even WebGL(brow.sh)
18 points by make3 7y ago|13 comments
937.
by make3 7y ago
but it really is so much more eco friendly
938.
by make3 7y ago
every single article on the matter has says that it would be a step towards potentially explaining dark matter and dark energy, which would help towards a unified theory
939.
by make3 7y ago
this is extremely exciting. makes me hope that we will see a unified theory of physics in our lifetime
940.
by make3 7y ago
that's an issue with electing presidents and powerful people that are over 70 years old lol
941.
by make3 7y ago
you mean for literally every piece of video taken everywhere..
942.
by make3 7y ago
I know you're joking, but serial line afaik has lower latency than the USB stuff being used today
943.
by make3 7y ago
Google (GCS) has huge gpu server farms everywhere too. that's not the factor here
944.
by make3 7y ago
the thing with multiplayer games is that they are often latency dependant like first person shooters or fighting games, or mobas. Portable and cheaper single player games may be a better application in practice, because you're more ra
945.
by make3 7y ago
no? the second link showed that it doesn't have to
946.
by make3 7y ago
wow they even added shading, to do fake lighting effects
947.
by make3 7y ago
not a single statistic or data point. this is really garbage content
948.
by make3 7y ago
still you could have it mailed lol
949.
by make3 7y ago
so you basically spent around one full week of your time just to not pay 50$ of extra checked-in luggage?
950.
by make3 7y ago
seriously with so many employees, nothing else would have sense imho
951.
by make3 7y ago
no, just usb-c
952.
by make3 7y ago
The first random state should be picked with start probability, which are equal to the frequency of a word beginning a sentence, which is exactly equal to picking the first word of a random sentence
953.
by make3 7y ago
In his dataset, all sentences start at the beginning of the line and end at it's end. This is strictly a toy, no one serious would use this for anything anyways.
954.
by make3 7y ago
the part about rare cards not being stronger than common cards is really not true anymore sadly, with competitively payable planeswalkers like Oko, for example, often being 50$+ each, and pretty much every card in competitive decks except b
955.
by make3 7y ago
it's one of the most pay to win games there is. there is a top to each format, in that the max performance can't be surpassed by investing more, but that top is currently around 700$ for the most popular format, standard, and doub
956.
by make3 7y ago
"The promised super smart products that were end-to-end \"AI\" never really materialized" What do you have in mind? IBM and Watson definitely hyped their shit beyond anything reasonable. Open AI is also responsible of th
957.
Bill Gates: You'd all be on Windows Mobile today, not Android, but I screwed up(zdnet.com)
2 points by make3 7y ago|0 comments
958.
by make3 7y ago
I understand what you mean but please understand that this code is targeted at people which would at least have some background knowledge, like having read the seminal Transformer paper, "Attention Is All You Need", https:/&
959.
by make3 7y ago
the machines are always trained with the same dataset for each task. the biggest difference right now is small technical modifications on models that are also pre trained on gigantic unlabelled datasets. this doesn't feel like we'
960.
by make3 7y ago
A part missing in your analysis is that time is involved in the decisions, and that time is continuous, and that time is very important

More ›