Medijski sadržaj
- Tweetovi
- Tweetovi i odgovori
- Medijski sadržaj, trenutna stranica.
-
At a loss of ~0.35, starting to get some interesting samples, eg https://gist.github.com/shawwn/69c0e95fa363e75c83cbaab204d21a1f … / https://www.gwern.net/docs/ai/music/2020-01-28-gpt2-midi-abc-111803542.mp3 … (improves substantially ~0m55s, 2m15s mark, impressively maintains themes throughout). Colab: https://colab.research.google.com/drive/1OJB0_mYMn5o4KTOdJFT4QNUMHkaeWcpY … Current training log:pic.twitter.com/UEX2Jm8ts3
Prikaži ovu nit -
Another example of fashion being multi-level signaling games: from "Signaling Status with Luxury Goods: The Role of Brand Prominence" Han et al 2020 http://tarjomefa.com/wp-content/uploads/2016/04/4578-English.pdf#page=3 …pic.twitter.com/iG5TTHwHw6
-
Imagine how much better these behavioral genetics Twitter debates with like Eric Turkheimer would be with these memes!pic.twitter.com/OzYdLoAmMq
-
-
@slatestarcodex The concentration-of-misery/everything-is-correlated in a graph: "Clustering of health, crime and social-welfare inequality in 4 million citizens from two nations", Richmond-Rakerd et al 2020: https://www.gwern.net/docs/sociology/2020-richmondrakerd.pdf …pic.twitter.com/kYmuk5evTW
-
That's a mystery. We need someone like
@lucidrains to train a Reformer on our ABC/MIDI dataset to compare directly. :) In any case, here's a screenshot of our 30k model training today:pic.twitter.com/gbGiGHviim
-
Some people say GPT-2 doesn't scale. They don't understand: the only thing that matters is—BRUTE STRENGTH! They want Reformer (they're probably beta orbiters). But TPUs have—BRUTE STRENGTH! TPUs run GPT-2-117M with 30,000 context windows for MIDI generation—with BRUTE STRENGTH!pic.twitter.com/lMTsqvJLUT
Prikaži ovu nit -
This Waifu Does Not Exist, v3 (https://www.thiswaifudoesnotexist.net/index.html ): 100k StyleGAN 2 anime portrait samples. Model: https://twitter.com/gwern/status/1216188531472850946 … "The blob artifacts, they're gone... it's so beautiful…(•̩̩̩̩_•̩̩̩̩)"pic.twitter.com/cPiG5Wxnw9
Prikaži ovu nit -
StyleGAN 2, anime portraits, 4 GPUs by
@SkyLi0n, >2 weeks, iteration #24,664: - Model: https://mega.nz/#!PeIi2ayb!xoRtjTXyXuvgDxSsSMn-cOh-Zux9493zqdxwVMaAzp4 … - Notebook: https://colab.research.google.com/drive/1Pv8OIFlonha4KeYyY2oEFaK4mG-alaWF …@Artbreeder@halcy@theshawwn@_Ryobot@ak92501 @roadrunning01@sunkworld@Gansodeva@__meimiya__@Buntworthypic.twitter.com/qTfiSRcCNFOvo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
-
One unexpected possibility: web browsers have support for furigana etc as 'ruby' (https://en.wikipedia.org/wiki/Ruby_character …), so you can do interlinear glossing easily. Doesn't work well here, though, as it basically doubles vertical space required, even if you use rubrication for emphasis.pic.twitter.com/eWhzKF0jlX
Prikaži ovu nit -
Another year, another traffic update: https://www.gwern.net/About#july-2019january-2020 … http://Gwern.net stats: 1,361,195 pageviews by 671,774 uniques in 2019; lifetime: 8,004,312 page-views by 3,814,522. 366 pages, 3.7m words (27MB), 13.3k edits, 33k files, 995 images, 9.9k docs, 18.3GB total.pic.twitter.com/rxSWbL5jrW
Prikaži ovu nit -
Does he tweet too much, or does he *re*tweet too much? People seem to always forget that you can disable seeing retweets from accounts.pic.twitter.com/M6Ws0U2qrM
-
-
-
Researchers in every paper's discussion section vs every press release/media articlepic.twitter.com/XYmZUrzTVF
-
And I 𝘵𝘩𝘪𝘯𝘬 we just hit a loss of 1, too. Amazing, considering that I was impressed by results from a loss of 1.6 from 117M on our combined poetry dataset. So I guess a swarm of ~100 preemptible TPUs can finetune GPT-2-1.5b completely in ~3-4 days with the right LR.pic.twitter.com/tRO0uJAo4r
-
This is the ideal 100-TPU swarm; u may not liek it but this is what peak performance looks like.pic.twitter.com/p78zZmwbza
-
What a difference switching from SGD or Adafactor optimizers to Adam makes for training GPT-2-1.5b-poetry! We just broke a loss of 2. (Still an initial spike, but many ways to fix that.)pic.twitter.com/df8lhXEZr5
-
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.