CTO @ app.mago.studio Beta is open! warpfusion, ArcaneGAN, face2comics. All tweets are sarcastic unless stated otherwise.
- Tweets 2,601
- Following 1,412
- Followers 7,252
- Likes 13,774
😻do you still have the init video? :D
WarpFusion was pretty sick. This was by far one of the most insane Google notebooks I ever touched and @devdef was such a rockstar for all the hand holding he did 🤘✨🐙🪸
2
4
413
Damn, 5090 renders 5s 1344x768 3 step in one minute 🤣
3-Step Minimax H3 generation on Mac!
Phosphene just dropped a way to generate H3 videos in just 3 steps. Just tried it it works!
You can either just generate the lowest resolution and then automatically upscale with 1-click, or just keep the low-res.
1
8
746
is it just me or vae upscale is the best one?
Update dropped: MiniMax‑H3 Latent Upscaler
The full FP32-precision version is now available.
huggingface.co/Asirus/Minima…
3
468
hm, why is @Kijaidesign 3 step lora much smaller?
Minimax H3😃TaoMate LORA for Comfy
3Step LORA👇
huggingface.co/Robert1212sta…
Use this with TaoMate-H3 👇huggingface.co/TaoLiveAIGC/T…
1
477
so it's confirmed: gta6 is built on DiscoDiffusion Warp :D
en.gamegpu.com/news/igry/sbo…
A technical glitch in neural rendering was documented in Grand Theft Auto V. While the game was paused, the internal engine stopped sending fresh frames to the presentation buffer.
Lacking a valid new frame, the model entered a recursive loop by constantly enhancing its own previous output. The blurred background was repeatedly scanned, which quickly distorted the character's face into chaotic visual artifacts.
This anomaly stems from a swapchain hook error that ignored the pause state. Generative models require precise engine integration and must shut down during static menu screens to prevent uncontrolled artifact generation.
#DLSS #Gaming #Graphics #TechNews #PCGaming #GameDev #Bugs
1
34
4,237
Is this intentionally misleading? The speedup from the VDN is 3x, the rest comes from going from 50 to 8 steps and from 1 to 8 GPUS.
Open-source video generation is now faster than playback without compromising quality.
Introducing Video Delta Net (VDN): hybrid attention for live text-to-video with near-lossless quality.
VDN accelerates Minimax-H3 by 75 - 90 x, generating 14 seconds of 768p video in 11 seconds on 8× NVIDIA B200 GPUs.
Checkpoints + training/inference code + Technical Blog ⬇️
(1/6)
2
8
453
for this to work you need to provide drop-in instructions on how to install and train and run your servos. the phenomenon of @pollenrobotics microduck is because it's plug and play.
Microduck is a really fun robot. We highly recommend giving our brushless digital servo a try. 🥰🥰unitree.com/DigitalServo
Community note
Microduck is not using Unitree's Servo.
x.com/matth_lapeyre/…
282
