Martineski

joined 2 years ago
MODERATOR OF
[–] Martineski 6 points 10 months ago

Damn, they got me good.

[–] Martineski 5 points 10 months ago
[–] Martineski 3 points 10 months ago (2 children)

Do you know how feasible is lowering the price permanently?

[–] Martineski 5 points 10 months ago

I wish I could watch the movie already. 😭

[–] Martineski 8 points 10 months ago (1 children)

What we do to this planet is disturbing. (I know I don't bring in anything to this conversation but felt like saying it.)

[–] Martineski 5 points 10 months ago (15 children)
[–] Martineski 2 points 10 months ago

Meanwhile my phone lowers the volume by itself an I'm thinking how to fix it. I assume it's because of dust or something. Annoying ah with this random popping out and lowering volume when I want sound.

[–] Martineski 4 points 10 months ago (2 children)

Gals with axes? 🤔

[–] Martineski 1 points 10 months ago (1 children)

Is there something about the meme that reminds you of it? Never played the game so am curious. haha

[–] Martineski 1 points 10 months ago

Sure, you don't need to ask. :)

[–] Martineski 1 points 10 months ago (1 children)

Nope, a neural network:

https://youtu.be/0Xn8xGV_w9w

https://arxiv.org/abs/2408.14837 "Diffusion Models Are Real-Time Game Engines"

https://gamengen.github.io/

We present GameNGen, the first game engine powered entirely by a neural model that enables real-time interaction with a complex environment over long trajectories at high quality. GameNGen can interactively simulate the classic game DOOM at over 20 frames per second on a single TPU. Next frame prediction achieves a PSNR of 29.4, comparable to lossy JPEG compression. Human raters are only slightly better than random chance at distinguishing short clips of the game from clips of the simulation. GameNGen is trained in two phases: (1) an RL-agent learns to play the game and the training sessions are recorded, and (2) a diffusion model is trained to produce the next frame, conditioned on the sequence of past frames and actions. Conditioning augmentations enable stable auto-regressive generation over long trajectories.

[–] Martineski 2 points 10 months ago (3 children)

Recently someone even managed to make a proof of concept doom running on a neural network.

 
 
769
Isekai (lemmy.dbzer0.com)
 
948
Digimon (lemmy.dbzer0.com)
 
271
SAO Players (lemmy.dbzer0.com)
 
96
hmmm (lemmy.dbzer0.com)
 
68
hmmm (lemmy.dbzer0.com)
 
221
hmmm (lemmy.dbzer0.com)
 
271
title 2 (lemmy.dbzer0.com)
submitted 2 years ago by Martineski to c/adhd
 
216
hmmm (lemmy.dbzer0.com)
 
161
hmmm (lemmy.dbzer0.com)
 
110
Dr. Stone (lemmy.dbzer0.com)
 
view more: ‹ prev next ›