# Google’s Parti AI: Magical Results! 💫 ## Метаданные - **Канал:** Two Minute Papers - **YouTube:** https://www.youtube.com/watch?v=XgdgSHweBUI - **Дата:** 30.07.2022 - **Длительность:** 8:17 - **Просмотры:** 225,680 ## Описание ❤️ Check out Lambda here and sign up for their GPU Cloud: https://lambdalabs.com/papers 📝 The paper "Google Parti: Pathways Autoregressive Text-to-Image model" is available here: https://parti.research.google/ 4 of my favorite prompts from the video (add these to benchmarks if you feel like it): - surprised scholars looking at a magical parchment emitting magic dust high detail digital art disney style - scholar delighted by a very long disintegrating magical parchment with sparks and smoke coming out of it fantasy digital art disney style - stern looking fox in a labcoat, casting a magic spell, digital art - shiny cybertronic robot frog with leds studio lighting high detail digital art 📝 The fluid control works are available here: https://users.cg.tuwien.ac.at/zsolnai/gfx/real_time_fluid_control_eg/ https://users.cg.tuwien.ac.at/zsolnai/gfx/fluid_control_msc_thesis/ ❤️ Watch these videos in early access on our Patreon page or join us here on YouTube: - https://www.patreon.com/TwoMinutePapers - https://www.youtube.com/channel/UCbfYPyITQ-7l4upoX8nvctg/join 🙏 We would like to thank our generous Patreon supporters who make Two Minute Papers possible: Aleksandr Mashrabov, Alex Balfanz, Alex Haro, Andrew Melnychuk, Benji Rabhan, Bryan Learn, B Shang, Christian Ahlin, Eric Martel, Geronimo Moralez, Gordon Child, Ivo Galic, Jace O'Brien, Jack Lukic, John Le, Jonas, Jonathan, Kenneth Davis, Klaus Busse, Kyle Davis, Lorin Atzberger, Lukas Biewald, Matthew Allen Fisher, Michael Albrecht, Michael Tedder, Nevin Spoljaric, Nikhil Velpanur, Owen Campbell-Moore, Owen Skarpness, Rajarshi Nigam, Ramsey Elbasheer, Steef, Taras Bobrovytsky, Ted Johnson, Thomas Krcmar, Timothy Sum Hon Mun, Torsten Reil, Tybie Fitzhugh, Ueli Gallizzi. If you wish to appear here or pick up other perks, click here: https://www.patreon.com/TwoMinutePapers 0:00 Google Parti 0:27 OpenAI's DALL-E 2 0:52 The problem 1:29 Google Imagen 2:09 Finally, Google Parti appears 2:22 1. Napoleon Cat returns 3:02 2. Water crocodile 3:14 3. Creativity in a machine 3:40 Why does this exist? 4:33 How does this help? 4:53 Let's test a huge prompt! 5:25 Watch it learn! 6:20 A new benchmark 6:36 More results 7:05 The age of AI generated images is here Thumbnail background design: Felícia Zsolnai-Fehér - http://felicia.hu Károly Zsolnai-Fehér's links: Instagram: https://www.instagram.com/twominutepapers/ Twitter: https://twitter.com/twominutepapers Web: https://cg.tuwien.ac.at/~zsolnai/ #parti #dalle ## Содержание ### [0:00](https://www.youtube.com/watch?v=XgdgSHweBUI) Google Parti dear fellow scholars this is two minute papers with dr carol jonaife i cannot tell you how excited i am by this paper wow today you will see more incredible images generated by google's newest ai called party yes this is not from open ai but from google so what is going on here just a few months ago openai's image generator ai called ### [0:27](https://www.youtube.com/watch?v=XgdgSHweBUI&t=27s) OpenAI's DALL-E 2 dolly 2 took the world by storm with that you could name almost anything cat napoleon a teddy bear on a skateboard on times square a basketball player dunking as an explosion of a nebula and it was able to create an appropriate image for it is also good enough to be used in our own thumbnails however there was one interesting thing ### [0:52](https://www.youtube.com/watch?v=XgdgSHweBUI&t=52s) The problem about it what do you think the prompt for this must have been not easy right well it was a sign that says deep learning oh yes this was one of the failure cases please remember this now we always say that in research do not look at where we are always will be two more papers down the line that is the first law of papers however we didn't even make it to two more papers down the line what's more we barely made two months down the line and ### [1:29](https://www.youtube.com/watch?v=XgdgSHweBUI&t=89s) Google Imagen scientists at google came up with an amazing follow-up paper it was called imogen imogen was absolutely incredible as it could now finally synthesize text properly it could also understand when we say a couple of glasses on a table and one this little linguistic battle against open ai's dolly too and all this just two months after it that is absolutely amazing but hold on to your papers because you won't believe this one i certainly didn't when i saw it first just about one month after ### [2:09](https://www.youtube.com/watch?v=XgdgSHweBUI&t=129s) Finally, Google Parti appears imogen here is an even newer paper on ai image generation called party that is fantastic welcome to our world little ai matt why does this exist ### [2:22](https://www.youtube.com/watch?v=XgdgSHweBUI&t=142s) 1. Napoleon Cat returns well this is why i'll explain in a moment but first let's have a look at what it can do through three of my favorite examples and then we'll discuss why it exists one let's start with the banger and recreate the legendary napoleon cat with the new method this is dolly 2 solution and let's see the new one i cannot believe it this is at least as good as dolly two's legendary solution and i have to say maybe even a touch better what a time to be alive two a crocodile made of water ### [3:02](https://www.youtube.com/watch?v=XgdgSHweBUI&t=182s) 2. Water crocodile as someone who has spent some time researching controlling fluid and smoke simulations this one is highly appreciated ### [3:14](https://www.youtube.com/watch?v=XgdgSHweBUI&t=194s) 3. Creativity in a machine appreciated three a detailed athenian vase with egyptian hieroglyphics and more i love how party was able to bring all of these concepts together into one coherent solution this may be subjective but if someone told me that a person made this i would say that person is quite creative but creativity in a machine how cool is that ### [3:40](https://www.youtube.com/watch?v=XgdgSHweBUI&t=220s) Why does this exist? now remember this image i said that this is why party exists so what is going on here well look the two previous techniques used a diffusion based model this means that when we ask it something it starts out from noise and over time it learns to organize these pixels to form a beautiful image that matches our description better now look aha this party technique is not a diffusion diffusion-based model it is an autoregressive model what does that mean it means that it uses no diffusion it does not create a piece of noise and refine it into an image no sir instead it thinks of an image as a collection of little puzzle pieces ### [4:33](https://www.youtube.com/watch?v=XgdgSHweBUI&t=273s) How does this help? why well this hopefully helps with two other shortcomings of dolly 2 one is generating a specific number of objects that did not work too well before and it can also deal with super long prompts much longer than previous ones you got me excited now you know what let's test ### [4:53](https://www.youtube.com/watch?v=XgdgSHweBUI&t=293s) Let's test a huge prompt! that right now together let's add this prompt oh my goodness now that is a long prompt who can paint this almost nobody in fact it is the description of van gogh's starry night without saying that we are looking for starry night i am itching to see this and oh wow all of them are lovely now we have two more really spectacular things about this paper ### [5:25](https://www.youtube.com/watch?v=XgdgSHweBUI&t=325s) Watch it learn! one we can witness how it learns to draw these beautiful images as we increase the model size which roughly tells us how capable the ai is we can take a smaller model and ask it to create a kangaroo with sunglasses and a sign that says hello friends and we get this well that is a start it can't really write yet and the details are lacking but when we use the same architecture with the difference that we increase the model size to be about 50 times bigger we get this oh my goodness it not only learned to write but the quality of the output is also leaps and bounds better what do you think what results would another 50-fold increase result in that must be something truly incredible ### [6:20](https://www.youtube.com/watch?v=XgdgSHweBUI&t=380s) A new benchmark let me know in the comments below alongside the paper scientists at google also released a bunch of prompts as a benchmark for testing future image generator ais and yes there are some good ones in there but if i may make some ### [6:36](https://www.youtube.com/watch?v=XgdgSHweBUI&t=396s) More results recommendations i would love to see some of these prompts of mine in such a benchmark for instance the fox scientists the scholars and the cyber frog are very well received by you fellow scholars and it would be super cool to be able to compare how new more elaborate ai models are able to deal with these i also put a text version of these in the video description if someone is interested ### [7:05](https://www.youtube.com/watch?v=XgdgSHweBUI&t=425s) The age of AI generated images is here so it's official the age of beautiful ai generated images is now here does this get your mind going what do you think let me know in the comments below if you're looking for inexpensive cloud gpus for ai lambda now offers the best prices in the world for gpu cloud compute no commitments or negotiation required just sign up and launch an instance and hold on to your papers because with lambda gpu cloud you can get on-demand a100 instances for 1. 10 per hour versus 4. 10 per hour with aws that's 73 percent savings did i mention they also offer persistent storage so join researchers at organizations like apple mit and caltech in using lambda cloud instances workstations or servers make sure to go to lambdalabs. com papers to sign up for one of their amazing gpu instances today thanks for watching and for your generous support and i'll see you next time --- *Источник: https://ekstraktznaniy.ru/video/13500*