Tweetovi
- Tweetovi, trenutna stranica.
- Tweetovi i odgovori
- Medijski sadržaj
Blokirali ste korisnika/cu @sherjilozair
Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @sherjilozair
-
Sherjil Ozair proslijedio/la je Tweet
.
@NVIDIA is the only place crazy enough to conceive of rendering 4K 60+ Hz graphics at home by running a neural network for every frame. DLSS: Trained in@PyTorch, running on the Tensor Cores in RTX GPUs, redefining graphics with AI. And we’re just getting started!https://twitter.com/wiredp/status/1222943514742415360 …Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
My take on this is that there is no difference between ideas and execution. Execution is but a series of good ideas. The first idea can be had in the armchair. But the rest require experimentation and analysis.https://twitter.com/andrey_kurenkov/status/1222930287111233536 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
I haven't found a single person who has used jax and said they don't like it. I've been actively priming people to criticize it, but noone does. Instead they tell me how good it feels getting off of Tensorflow. Looking forward to jaxxing myself soon.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
Videogames are "just algorithms" in the same way that novels are "just ink arranged on pieces of dead, processed wood."
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
New blog post with
@carlesgelada -- "A Sober Look at Bayesian Neural Networks": https://jacobbuckman.com/2020-01-17-a-sober-look-at-bayesian-neural-networks/ … Without a good prior, Bayesian uncertainties are meaningless. We argue that BNN priors are likely quite poor, and concretely characterize one specific failure mode.Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
We are redefining real-time graphics with deep learning. DL gives us both better image quality and 1.4-2X higher frame rate. Usually you get either speed or quality, not both. So glad to see this technology shipping in games - check out this review!https://www.youtube.com/watch?v=gpzFX4P1Jow …
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
btw did you know that deep learning (praise be upon it) is perfect, has unlimited capabilities, and can solve all problems under the sun? like why are people even doing ml research now? i don't get it.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
The funny thing about AIDungeon is due to all the GPT-2 pre-training on swaths of random web text, you can create an adventure game about literally anything. Here's one about installing OpenCV from source:pic.twitter.com/zJvtjBwcbq
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
This remains my major gripe with papers that focus on solving the simulator. Solving MontezumasRevenge by any means necessary is not the goal, it's a *proxy* for real-world tasks. When we start leveraging simulator tricks, it stops becoming a proxy problem.
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
Today,
@theintercept published an exposé by Rodrigo Ochigame (a PhD candidate at MIT) on the invention of "ethical AI" by big tech. https://theintercept.com/2019/12/20/mit-ethical-ai-artificial-intelligence/ … Here's why I think it's important to read this piece and share it widely. 1/Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
I often hear people say that X is broken. Congress is broken. Academia is broken. Physics is broken. Deep learning is broken. People like to stand in crowds and point fingers at the thing that is broken. I prefer the quiet folks who roll up their sleeves and get to work, fixing.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
The solution, as emphasized by Yoshua and
@jefrankle, is to get rid of artificial trophies such as best paper awards or even reviews. Instead, test of time should be the true sign of merit.Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Highly recommend watching the
@MLRetrospective panel: https://slideslive.com/38921895/retrospectives-a-venue-for-selfreflection-in-ml-research-2 …. Amazing performance by@jefrankle who held his own amongst giants of the field. We cannot deny that the ladder exists. Asking students to hurt their career is not the way forward.Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
Unsupervised pre-training now outperforms supervised learning on ImageNet for any data regime (see figure) and also for transfer learning to Pascal VOC object detection https://arxiv.org/abs/1905.09272v2 …pic.twitter.com/cciL5Db73x
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
We are presenting our
#NeurIPS2019 work "Unsupervised State Representation Learning in Atari" on Thursday 10:45 AM (Hall B+C #70). Come around if are interested in representation learning for RL! https://arxiv.org/abs/1906.08226 w/@evanracah@sherjilozair Yoshua@Cote_Marc@devon_hjelmpic.twitter.com/3kPal8pHd4
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
"Every block of stone has a statue inside it and it is the task of the sculptor to discover it." -- Michelangelo
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
A Chinese Communist Party member has leaked a trove of documentation on how China has organized mass detentions of Muslim minorities. Ramzy: "I’ve reported on China since 2003, and can’t recall a leak like this". This is a big deal, please check it out!https://twitter.com/austinramzy/status/1195688077731061760 …
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
Hikvision, SenseTime, Face++, etc, companies on the recent U.S. blacklist have been regular sponsors of top computer vision conferences, such as CVPR, ECCV, ICCV. They use those venues to attract talent for developing surveillance technologies. (ie: see http://cvpr2018.thecvf.com/sponsors )https://twitter.com/hardmaru/status/1181377780212088833 …
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
Extrapolating out these results we find that by 2022 the entire world's compute resources will be used on a single OpenAI project. We may even be able to get a robot hand to solve a rubik's cube without dropping 20% of the time. Isn't the future amazing!
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Sherjil Ozair proslijedio/la je Tweet
ICLR papers with perfect scores (all 8s, total 11 papers): 1. https://openreview.net/forum?id=BygzbyHFvB … "FreeLB: Enhanced Adversarial Training for Language Understanding" 2. https://openreview.net/forum?id=BJlrF24twB … "BackPACK: Packing more into Backprop"
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
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.