Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

how different layers of residual vector quantized audio build upon each other (max hierarchical level is ~83 codes per second in the neural codec)

kalomaze

24,811 subscribers

19,809 views • 11 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Claude Fable 5 is insane again. The first time, I only used it to build a 3D game clone. I just used Fable 5 inside Rork to build a new tower defense game in about 1 hour. From the initial idea → level design → heroes & monsters → in-game economy → game art → sound effects → code Each level will have different difficulty, different map upgrades, different monsters, and different bosses. Each tower weapon has unique skills, counters other weapons in different ways, and the heroes are controllable with skills that can support each other... All in one flow. This is getting crazy me again. 🙃

Claude Fable 5 is insane again. The first time, I only used it to build a 3D game clone. I just used Fable 5 inside Rork to build a new tower defense game in about 1 hour. From the initial idea → level design → heroes & monsters → in-game economy → game art → sound effects → code Each level will have different difficulty, different map upgrades, different monsters, and different bosses. Each tower weapon has unique skills, counters other weapons in different ways, and the heroes are controllable with skills that can support each other... All in one flow. This is getting crazy me again. 🙃

Alex Nguyen

100,797 views • 3 months ago

I'm working on visualizations of how I'm using my LayerProcGen framework in my game. Layers building upon each other is how I can generate highly contextual features (like paths based on path-finding) that seamlessly cross chunk boundaries. #ProcGen

I'm working on visualizations of how I'm using my LayerProcGen framework in my game. Layers building upon each other is how I can generate highly contextual features (like paths based on path-finding) that seamlessly cross chunk boundaries. #ProcGen

Rune Skovbo Johansen →🦋🐘

10,080 views • 2 years ago

Ever wondered what neural networks are and how they work? Systems like ChatGPT use neural networks to work as well as they do. Neural networks are composed of "layers" of neurons, layers with different functions; connections between layers called "weights"; and mathematical functions called "activation functions". If you’re interested in learning about these systems, check the comments. Ultimately, the neural network structure of the model serves to visually demonstrate that it is, in fact, a complex mathematical equation. When companies release the model's weights, they are releasing a key component needed to run the model's complete equation. Without the weights, the equation is incomplete. For the math-minded: the weights of a model are the learned numbers (they are variables during training) that are then used as constants in the mathematical functions that make up the model. Neural networks are ultimately just one big, hyper-complex mathematical function, and when a model is trained, it learns the constants associated with the high-dimensional input.

Ever wondered what neural networks are and how they work? Systems like ChatGPT use neural networks to work as well as they do. Neural networks are composed of "layers" of neurons, layers with different functions; connections between layers called "weights"; and mathematical functions called "activation functions". If you’re interested in learning about these systems, check the comments. Ultimately, the neural network structure of the model serves to visually demonstrate that it is, in fact, a complex mathematical equation. When companies release the model's weights, they are releasing a key component needed to run the model's complete equation. Without the weights, the equation is incomplete. For the math-minded: the weights of a model are the learned numbers (they are variables during training) that are then used as constants in the mathematical functions that make up the model. Neural networks are ultimately just one big, hyper-complex mathematical function, and when a model is trained, it learns the constants associated with the high-dimensional input.

Harper Carroll

39,391 views • 11 months ago

this is probably one of the most perfect scenes in gilmore girls because of how subtly good it is. the face acting in each second of the scene, the dancing, the way they seem kind’ve tense towards the beginning but become more comfortable w each other

this is probably one of the most perfect scenes in gilmore girls because of how subtly good it is. the face acting in each second of the scene, the dancing, the way they seem kind’ve tense towards the beginning but become more comfortable w each other

madi ౨ৎ

95,130 views • 3 years ago

High speed frame by frame video of popping water balloons in 0g. It only takes 1-2 frames for the balloon skin to contract leaving the body of water to residual forces without the effects of gravity. Original movie shot at 420 frames per second and each frame sequentially shown for 1/2 second.

High speed frame by frame video of popping water balloons in 0g. It only takes 1-2 frames for the balloon skin to contract leaving the body of water to residual forces without the effects of gravity. Original movie shot at 420 frames per second and each frame sequentially shown for 1/2 second.

Don Pettit

56,449 views • 2 months ago

excited to see how the ColoPC build is going! can't wait to see what@/oddertechs does in the second part of the build!

Sensitive content

excited to see how the ColoPC build is going! can't wait to see what@/oddertechs does in the second part of the build!

Colodraws

118,447 views • 2 years ago

Incisive is still the most effective way to build up in #FC26. Here’s how Levy and other pros are setting it up to get the MAX out of this playstyle ⬇️

Incisive is still the most effective way to build up in #FC26. Here’s how Levy and other pros are setting it up to get the MAX out of this playstyle ⬇️

Venny

45,436 views • 1 year ago

If anyone is wondering how they did this, they used twitter's compression algorithm and uploaded to photo too big for the maximum demensions. There are 3 different maximum demensions. One when opend, the second which is displayed in the timeline, the 3rd when you longpress. They layered 3 photos that when compressed to each set of demensions displays somthing else. Cool little trick here is a video showing how the 3 different layers change as the compression changes.

If anyone is wondering how they did this, they used twitter's compression algorithm and uploaded to photo too big for the maximum demensions. There are 3 different maximum demensions. One when opend, the second which is displayed in the timeline, the 3rd when you longpress. They layered 3 photos that when compressed to each set of demensions displays somthing else. Cool little trick here is a video showing how the 3 different layers change as the compression changes.

J.P. Picard

127,612 views • 2 months ago

The era of 1-bit LLMs is here — now with WebGPU acceleration! 🤯 It's incredible to think that a quantized 1.7B model (just 290MB in size) can run at ~100 tokens per second entirely in your browser. Try the demo yourself 👇

The era of 1-bit LLMs is here — now with WebGPU acceleration! 🤯 It's incredible to think that a quantized 1.7B model (just 290MB in size) can run at ~100 tokens per second entirely in your browser. Try the demo yourself 👇

Xenova

105,543 views • 5 months ago

“And I think this group is just different. I think the first thing is they care for each other. And the second thing is they care for basketball.” 🤝 🎥 Watch the Season Finale of Open Gym : link in bio Presented by Bell

“And I think this group is just different. I think the first thing is they care for each other. And the second thing is they care for basketball.” 🤝 🎥 Watch the Season Finale of Open Gym : link in bio Presented by Bell

Toronto Raptors

17,109 views • 4 months ago

At second 7 she has another girl's face in both hands and they are laughing into each other 😳 At second 19 she is nose to nose with the guy in the red cap. 12 seconds apart. Same necklace. Different PERSON.

Sensitive content

At second 7 she has another girl's face in both hands and they are laughing into each other 😳 At second 19 she is nose to nose with the guy in the red cap. 12 seconds apart. Same necklace. Different PERSON.

abovee 1%

820,730 views • 5 days ago

video trying to demonstrate a workflow™ system of layers being bound to tools. The idea is that during drawing you only need to pick the tool you want without having to interact with layers or other panels. If i remembered where each tool is you'd see how seamless and fast it is

video trying to demonstrate a workflow™ system of layers being bound to tools. The idea is that during drawing you only need to pick the tool you want without having to interact with layers or other panels. If i remembered where each tool is you'd see how seamless and fast it is

redi

12,749 views • 4 months ago

Here's a real-time audio-reactive (sound on 🔊) fire sim running in Unreal 5.6. Each of the small emitters is linked to a different frequency of the audio spectrum. I've been experimenting with audio-driven Niagara FX and will be sharing more soon.

Here's a real-time audio-reactive (sound on 🔊) fire sim running in Unreal 5.6. Each of the small emitters is linked to a different frequency of the audio spectrum. I've been experimenting with audio-driven Niagara FX and will be sharing more soon.

Jesse Pitela

26,780 views • 1 year ago

10.4 million points per second, 256 channels, color captured with depth in the same photon event Here is a Rev8 OS1 Max walkthrough of the Cathedral of Seville, taken in a single pass.

10.4 million points per second, 256 channels, color captured with depth in the same photon event Here is a Rev8 OS1 Max walkthrough of the Cathedral of Seville, taken in a single pass.

Ouster

13,947 views • 1 month ago

❗️Reporter: How are the U.S. actions in Venezuela different from Putin’s actions in Russia? Orbán: Each comparison is more than dangerous. Each case is different by legal character and by political context. We have to take each of them as they are.

❗️Reporter: How are the U.S. actions in Venezuela different from Putin’s actions in Russia? Orbán: Each comparison is more than dangerous. Each case is different by legal character and by political context. We have to take each of them as they are.

🪖MilitaryNewsUA🇺🇦

31,141 views • 9 months ago

A little bit of analysis In these 4 saves by #Millwall Max Crocombe Max Crocombe you can see what kind of goalkeeper he is Max has a double reflex per second. If the ball changes direction, he additionally changes his intervention With certainty the best goalie in the Championship.

A little bit of analysis In these 4 saves by #Millwall Max Crocombe Max Crocombe you can see what kind of goalkeeper he is Max has a double reflex per second. If the ball changes direction, he additionally changes his intervention With certainty the best goalie in the Championship.

Miodrag

27,917 views • 10 months ago