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.

Awesome. NVIDIA dropped PiD - fast high-res latent decoding via pixel diffusion! - replace VAE - 4/8x upsampling - 2k decoding in <1s on RTX 5090 - works with FLUX.1/SD3/Z - rapid generation previews sharper details, much lower hardware lag compared to standard methods.

Wildminder

12,292 subscribers

23,806 views • 5 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

5 Comments

Ramon Guthrie5 months ago

🙄 @ComfyUI when?

MADmotionAI5 months ago

I wonder if this has a chance of making it into the LTX 2.3 workflow, their RTX Video Super Resolution works pretty well.

Sen Fang5 months ago

NVIDIA's approach is similar to one we disclosed two months ago. However, they use DiT to decode in <1 second, while we use CNN to decode in 0.02 seconds. They didn't notice our work: Project Page:

Brian Berneker5 months ago

When can we get it for our workflows?

Matt Hillebrand5 months ago

Very interesting! Can't wait to try this with my FLUX.2-dev LoRAs.

Related Videos

Fast Byte Latent Transformer is accepted to ICML 2026! ⚡🥪 Byte-level LMs promise to free us from subword tokenizers, but decoding one byte at a time is super slow. We make BLT generation more efficient with BLT-D: text diffusion for parallel byte decoding. 1/

Fast Byte Latent Transformer is accepted to ICML 2026! ⚡🥪 Byte-level LMs promise to free us from subword tokenizers, but decoding one byte at a time is super slow. We make BLT generation more efficient with BLT-D: text diffusion for parallel byte decoding. 1/

Julie Kallini✨COLM26

107,444 views • 5 months ago

FLUX.2 Klein in pure pixel space! No VAE. AsymFlow - hyper-realistic images by working directly in pixel space rather than using compressed latent representations. - sharper textures, superior visual fidelity - 40% faster - low-rank noise parameterization to solve high-dimensional bottlenecks ComfyUI support incoming

FLUX.2 Klein in pure pixel space! No VAE. AsymFlow - hyper-realistic images by working directly in pixel space rather than using compressed latent representations. - sharper textures, superior visual fidelity - 40% faster - low-rank noise parameterization to solve high-dimensional bottlenecks ComfyUI support incoming

Wildminder

27,210 views • 5 months ago

Nvidia's PID (Pixel diffusion Decoder) It decodes latent representations into high-resolution images, replacing the decode–then–super-resolve cascade while achieving lower latency and higher visual quality page: HF repo:

Nvidia's PID (Pixel diffusion Decoder) It decodes latent representations into high-resolution images, replacing the decode–then–super-resolve cascade while achieving lower latency and higher visual quality page: HF repo:

Stable Diffusion Tutorials

37,906 views • 5 months ago

9 Key AI Concepts Explained in 7 minutes - Tokenization - Text Decoding - Prompt Engineering - Multi Step AI Agents - RAGs - RLHF - VAE - Diffusion Models - LoRA

9 Key AI Concepts Explained in 7 minutes - Tokenization - Text Decoding - Prompt Engineering - Multi Step AI Agents - RAGs - RLHF - VAE - Diffusion Models - LoRA

Bytebytego

100,072 views • 8 months ago

Use FLUX with under 12GB VRAM Wow can confirm this workflow for FLUX on ComfyUI actually works as advertised! The default workflow was much slower. This workflow uses much less memory and much faster. Here's an unedited recording of 1X speed generation on NVIDIA RTX 4090.

Use FLUX with under 12GB VRAM Wow can confirm this workflow for FLUX on ComfyUI actually works as advertised! The default workflow was much slower. This workflow uses much less memory and much faster. Here's an unedited recording of 1X speed generation on NVIDIA RTX 4090.

cocktail peanut

35,476 views • 2 years ago

The latent-vs-pixel debate misses the point. GPT Image 2 shows what users notice: pixel-level fidelity. Latent models show what scales: compact semantic structure. We connect them by replacing VAE/RAE decoders with a Pixel Diffusion Decoder. Code and Model available: 🧵(1/N)

The latent-vs-pixel debate misses the point. GPT Image 2 shows what users notice: pixel-level fidelity. Latent models show what scales: compact semantic structure. We connect them by replacing VAE/RAE decoders with a Pixel Diffusion Decoder. Code and Model available: 🧵(1/N)

Xuanchi Ren

669,706 views • 5 months ago

MotionStreamer announced on Hugging Face Streaming Motion Generation via Diffusion-based Autoregressive Model in Causal Latent Space

MotionStreamer announced on Hugging Face Streaming Motion Generation via Diffusion-based Autoregressive Model in Causal Latent Space

AK

43,422 views • 1 year ago

(1/5) 5 seconds of video. 1.8s seconds of generation. One NVIDIA GeForce RTX 5090 on FastVideo. 🤯🚀 - FastWan-QAD, a new family of video generation models - Trained with FastVideo's Quantization-Aware Distillation (QAD) recipe. - Powered by FastVideo, we push a single NVIDIA GeForce RTX 5090 to its absolute limit: generating a 5-second 480P video in 1.8s end-to-end! 📜 Blog: 💻 Code: 💽 Model:

(1/5) 5 seconds of video. 1.8s seconds of generation. One NVIDIA GeForce RTX 5090 on FastVideo. 🤯🚀 - FastWan-QAD, a new family of video generation models - Trained with FastVideo's Quantization-Aware Distillation (QAD) recipe. - Powered by FastVideo, we push a single NVIDIA GeForce RTX 5090 to its absolute limit: generating a 5-second 480P video in 1.8s end-to-end! 📜 Blog: 💻 Code: 💽 Model:

Hao AI Lab

13,411 views • 4 months ago

Speed and quality can finally coexist in diffusion-based language generation. Introducing DiDi-Instruct, a Discrete Diffusion Divergence Instruct method that distills a pre-trained discrete diffusion language model (dLLM) into a few-step student for ultra-fast generation. Built on integral KL-divergence minimization, DiDi-Instruct achieves up to 64× faster decoding, surpasses both its teacher and GPT-2, and cuts training time by 20×. Ultra-Fast Language Generation via Discrete Diffusion Divergence Instruct Paper: Code: Project: Our report: 📬 #PapersAccepted by Jiqizhixin

Speed and quality can finally coexist in diffusion-based language generation. Introducing DiDi-Instruct, a Discrete Diffusion Divergence Instruct method that distills a pre-trained discrete diffusion language model (dLLM) into a few-step student for ultra-fast generation. Built on integral KL-divergence minimization, DiDi-Instruct achieves up to 64× faster decoding, surpasses both its teacher and GPT-2, and cuts training time by 20×. Ultra-Fast Language Generation via Discrete Diffusion Divergence Instruct Paper: Code: Project: Our report: 📬 #PapersAccepted by Jiqizhixin

机器之心 JIQIZHIXIN

18,196 views • 11 months ago

Qwen 32B (4-bit) generates at >40 toks/sec on an M4 Max with assisted decoding and Qwen 0.5B as the draft model. Coming soon to mlx-lm. Compare regular decoding (left) to assisted decoding (right):

Qwen 32B (4-bit) generates at >40 toks/sec on an M4 Max with assisted decoding and Qwen 0.5B as the draft model. Coming soon to mlx-lm. Compare regular decoding (left) to assisted decoding (right):

Awni Hannun

50,375 views • 1 year ago

A new paradigm for robot learning that unifies latent world modeling and action generation. Astribot has introduced Lumo-2, a latent world-action model for robot control. - Validated on a real robot through 20+ intricate chores, extending beyond simple benchmark testing. - A 3-stage recipe that progressively aligns actions with world dynamics, then vision, then language - Action quality depends on the geometry of the latent space, not just reconstruction accuracy - BAR (block-wise decoding) speeds up inference by 2.71x compared to standard autoregressive decoding - Shows cross-embodiment transfer without specialized transfer learning Results: - Beats π0.5 and Fast-WAM in temporal reasoning, physical understanding, long-horizon, and dexterous manipulation - Lumo-2's VLM consistently outperforms Qwen-3.5 4B on embodied benchmarks

A new paradigm for robot learning that unifies latent world modeling and action generation. Astribot has introduced Lumo-2, a latent world-action model for robot control. - Validated on a real robot through 20+ intricate chores, extending beyond simple benchmark testing. - A 3-stage recipe that progressively aligns actions with world dynamics, then vision, then language - Action quality depends on the geometry of the latent space, not just reconstruction accuracy - BAR (block-wise decoding) speeds up inference by 2.71x compared to standard autoregressive decoding - Shows cross-embodiment transfer without specialized transfer learning Results: - Beats π0.5 and Fast-WAM in temporal reasoning, physical understanding, long-horizon, and dexterous manipulation - Lumo-2's VLM consistently outperforms Qwen-3.5 4B on embodied benchmarks

The Humanoid Hub

20,623 views • 3 months ago

Dealing with data in Dart? 📉🎯 dart:convert makes it easy to handle common conversions like decoding UTF8 bytes, encoding/decoding JSON, and much more.

Dealing with data in Dart? 📉🎯 dart:convert makes it easy to handle common conversions like decoding UTF8 bytes, encoding/decoding JSON, and much more.

Flutter

11,570 views • 1 year ago

MinerU-Diffusion A 2.5B diffusion-based OCR model that replaces slow autoregressive decoding with parallel block-wise diffusion, achieving up to 3.2x faster inference while improving robustness on complex documents with tables, formulas, and layouts.

MinerU-Diffusion A 2.5B diffusion-based OCR model that replaces slow autoregressive decoding with parallel block-wise diffusion, achieving up to 3.2x faster inference while improving robustness on complex documents with tables, formulas, and layouts.

DailyPapers

15,304 views • 7 months ago

Fameye - Gen Z | Decoding

Fameye - Gen Z | Decoding

code micky

48,731 views • 18 days ago

20-year-old hardware decoding 44.1/16 with the whole interface running. low diff.

20-year-old hardware decoding 44.1/16 with the whole interface running. low diff.

Obsolete Sony

210,851 views • 2 months ago

LTX-2.5 is native in ComfyUI, day zero with the weights Model highlights: → Diffusion Fidelity Rendering - compute allocated by scene complexity, not spread evenly → Diffusion Video Decoder - replaces VAE decoding, sharper faces and legible text → Native multi-shot - one run, multiple connected shots, consistent across the cuts → Custom Gemma 4 12B text encoder - long prompts survive intact, no dropped clauses → Prompt enhancer + auto duration - two fewer parameters to tune The change LTX leads with is Diffusion Fidelity Rendering, and it's what lifts detail across the whole stack rather than one stage of it. Learn more and download the workflows with the link below 👇

LTX-2.5 is native in ComfyUI, day zero with the weights Model highlights: → Diffusion Fidelity Rendering - compute allocated by scene complexity, not spread evenly → Diffusion Video Decoder - replaces VAE decoding, sharper faces and legible text → Native multi-shot - one run, multiple connected shots, consistent across the cuts → Custom Gemma 4 12B text encoder - long prompts survive intact, no dropped clauses → Prompt enhancer + auto duration - two fewer parameters to tune The change LTX leads with is Diffusion Fidelity Rendering, and it's what lifts detail across the whole stack rather than one stage of it. Learn more and download the workflows with the link below 👇

ComfyUI

2,405,487 views • 2 months ago

try out the Gradio Demo for AudioLDM: Text-to-Audio Generation with Latent Diffusion Models on Hugging Face demo:

try out the Gradio Demo for AudioLDM: Text-to-Audio Generation with Latent Diffusion Models on Hugging Face demo:

AK

82,138 views • 3 years ago