Loading video...

Video Failed to Load

Go Home

We present PeRFlow (Piecewise Rectified Flow): Fast sampling. Efficient training. Transferrable to various SD-based models/methods. Fully Open-Sourced. ➡️Website: ➡️Github: Screenshot for fast text-to-multiview with PeRFlow:

XingchaoL's profile picture

Xingchao Liu

4,628 subscribers

28,800 views • 2 years ago •via X (Twitter)

EducationScience & Technology
LIVE

Anya Rossi• Live Now

Private livecam show

2 Comments

Suraj Patil's profile picture
Suraj Patil2 years ago

Really cool work! Would be nice to have the demos on hf spaces, happy to help!

Xingchao Liu's profile picture
Xingchao Liu2 years ago

Contributions are welcomed!

Related Videos

Apple just released and open-sourced FastVLM! FastVLM is a lightning-fast vision-language model that combines rapid image and text understanding with efficient on-device performance. 100% Open Source
0:21

Apple just released and open-sourced FastVLM! FastVLM is a lightning-fast vision-language model that combines rapid image and text understanding with efficient on-device performance. 100% Open Source

Sumanth

43,685 views • 9 months ago

Introducing Adjoint Sampling, a new learning algorithm that trains generative models based on scalar rewards. Based on theoretical foundations developed by FAIR, Adjoint Sampling leads to a highly scalable practical algorithm, and can become the foundation for further research into highly scalable sampling methods. Read our research paper on Adjoint Sampling and download the model, code, and benchmark ➡️
0:28

Introducing Adjoint Sampling, a new learning algorithm that trains generative models based on scalar rewards. Based on theoretical foundations developed by FAIR, Adjoint Sampling leads to a highly scalable practical algorithm, and can become the foundation for further research into highly scalable sampling methods. Read our research paper on Adjoint Sampling and download the model, code, and benchmark ➡️

AI at Meta

36,987 views • 1 year ago

Stability AI just dropped Stable Audio Open Small on Hugging Face Fast Text-to-Audio Generation with Adversarial Post-Training
0:43

Stability AI just dropped Stable Audio Open Small on Hugging Face Fast Text-to-Audio Generation with Adversarial Post-Training

AK

55,100 views • 1 year ago

🎉 Diffusion-style annealing + sampling-based MPC can surpass RL, and seamlessly adapt to task parameters, all 𝘁𝗿𝗮𝗶𝗻𝗶𝗻𝗴-𝗳𝗿𝗲𝗲! We open sourced DIAL-MPC, the first training-free method for whole-body torque control using full-order dynamics 🧵
0:31

🎉 Diffusion-style annealing + sampling-based MPC can surpass RL, and seamlessly adapt to task parameters, all 𝘁𝗿𝗮𝗶𝗻𝗶𝗻𝗴-𝗳𝗿𝗲𝗲! We open sourced DIAL-MPC, the first training-free method for whole-body torque control using full-order dynamics 🧵

Haoru Xue ✈️ CVPR

172,397 views • 1 year ago

Text-to-image generative models, meet robotics! We present ROSIE: Scaling RObot Learning with Semantically Imagined Experience, where we augment real robotics data with semantically imagined scenarios for downstream manipulation learning. Website: 🧵👇
0:26

Text-to-image generative models, meet robotics! We present ROSIE: Scaling RObot Learning with Semantically Imagined Experience, where we augment real robotics data with semantically imagined scenarios for downstream manipulation learning. Website: 🧵👇

Fei Xia

196,378 views • 3 years ago

It’s velocity based training… Not fast-based training Vitruve
1:01

It’s velocity based training… Not fast-based training Vitruve

Cody Hughes

19,639 views • 11 months ago

prototyping a new onboarding flow concept ⦿ all text-based interface ⦿ fully keyboard navigable ⦿ traverse multi-step flows with ctrl+b ⦿ getting started to connect provider ⦿ highlighting free models implementation is in progress
0:46

prototyping a new onboarding flow concept ⦿ all text-based interface ⦿ fully keyboard navigable ⦿ traverse multi-step flows with ctrl+b ⦿ getting started to connect provider ⦿ highlighting free models implementation is in progress

David Hill

18,842 views • 6 months ago

./RedirH4x : python-based vulnerability scanner designed to identify open redirect flaws in website applications. Github :
0:49

./RedirH4x : python-based vulnerability scanner designed to identify open redirect flaws in website applications. Github :

Gospel

15,117 views • 1 year ago

Bengaluru, we just made EV delivery hyper fast! With #HyperDelivery, it’s simple: walk in ➡️ pick your S1 ➡️ ride out. Fast, smooth, and oh-so-effortless. 🛵
0:40

Bengaluru, we just made EV delivery hyper fast! With #HyperDelivery, it’s simple: walk in ➡️ pick your S1 ➡️ ride out. Fast, smooth, and oh-so-effortless. 🛵

Ola Electric

40,884 views • 1 year ago

Introducing MiniCPM 4.1-8B: First Open-Source Reasoning LLM with Trainable Sparse Attention ✅ Strong Reasoning Capability: Surpasses similar-sized models on 15 tasks! ✅ Fast Generation: 3x decoding speedup for reasoning ✅ Efficient Architecture: Trainable sparse attention, frequency-ranked speculative decoding Download Models: Huggingface: Github: Technical Report: #AI #MiniCPM #LLM #OpenBMB #ArtificialIntelligence #MachineLearning
0:54

Introducing MiniCPM 4.1-8B: First Open-Source Reasoning LLM with Trainable Sparse Attention ✅ Strong Reasoning Capability: Surpasses similar-sized models on 15 tasks! ✅ Fast Generation: 3x decoding speedup for reasoning ✅ Efficient Architecture: Trainable sparse attention, frequency-ranked speculative decoding Download Models: Huggingface: Github: Technical Report: #AI #MiniCPM #LLM #OpenBMB #ArtificialIntelligence #MachineLearning

OpenBMB

19,070 views • 9 months ago

Designing an Encoder for Fast Personalization of Text-to-Image Models TL;DR: use an encoder to personalize a text-to-image model to new concepts with a single image and 5-15 tuning steps abs: project page:
0:43

Designing an Encoder for Fast Personalization of Text-to-Image Models TL;DR: use an encoder to personalize a text-to-image model to new concepts with a single image and 5-15 tuning steps abs: project page:

AK

165,151 views • 3 years ago

🚨In our NeurIPS paper, we bring encoder-decoders back.. for diffusion language models! ⚡️Encoder-decoders make diffusion sampling fast: a small (fast) decoder denoises tokens progressively and a large (slower) encoder represents clean context.
0:31

🚨In our NeurIPS paper, we bring encoder-decoders back.. for diffusion language models! ⚡️Encoder-decoders make diffusion sampling fast: a small (fast) decoder denoises tokens progressively and a large (slower) encoder represents clean context.

Marianne Arriola

31,659 views • 7 months ago

I wrote an interactive article explaining the geometric intuition behind Rectified Flows. I visually explain why flow-models tend to learn curved trajectories, why this is bad for sampling latency, and a relatively simple technique for mitigating it. Check it out! Link 👇
0:25

I wrote an interactive article explaining the geometric intuition behind Rectified Flows. I visually explain why flow-models tend to learn curved trajectories, why this is bad for sampling latency, and a relatively simple technique for mitigating it. Check it out! Link 👇

Alec Helbling

158,628 views • 4 months ago

Bridge your tokens with ease! 🌉 Moonbeam ↔ Moonriver Network transfers are fast, secure, and fully transparent. ✅ Now LIVE ➡️
2:08

Bridge your tokens with ease! 🌉 Moonbeam ↔ Moonriver Network transfers are fast, secure, and fully transparent. ✅ Now LIVE ➡️

Moonbeam

14,167 views • 7 months ago

Project Talon is here. This next-gen autonomous aircraft is made to adapt fast. ➡️ Modular by design ➡️ Mission-ready ➡️ Built for the challenges ahead
0:23

Project Talon is here. This next-gen autonomous aircraft is made to adapt fast. ➡️ Modular by design ➡️ Mission-ready ➡️ Built for the challenges ahead

Northrop Grumman

63,195 views • 6 months ago

btw we shipped our own blazing-fast block explorer for tempo and open sourced it built on cloudflare workers and index supply, costs few dollars / month to maintain given an RPC
0:30

btw we shipped our own blazing-fast block explorer for tempo and open sourced it built on cloudflare workers and index supply, costs few dollars / month to maintain given an RPC

Georgios Konstantopoulos

90,218 views • 5 months ago

🎙️Do you know you now have all the building blocks for full speech-to-speech? - Voxtral Realtime: High-quality, real-time speech-to-text. - Mistral Small 4: Fast, efficient, general-purpose agentic model. - Voxtral TTS: Realistic customizable text-to-speech with streaming output.
0:50

🎙️Do you know you now have all the building blocks for full speech-to-speech? - Voxtral Realtime: High-quality, real-time speech-to-text. - Mistral Small 4: Fast, efficient, general-purpose agentic model. - Voxtral TTS: Realistic customizable text-to-speech with streaming output.

Mistral AI for Developers

27,647 views • 2 months ago

Presto! Distilling Steps and Layers for Accelerating Music Generation Despite advances in diffusion-based text-to-music (TTM) methods, efficient, high-quality generation remains a challenge. We introduce Presto!, an approach to inference acceleration for score-based diffusion transformers via reducing both sampling steps and cost per step. To reduce steps, we develop a new score-based distribution matching distillation (DMD) method for the EDM-family of diffusion models, the first GAN-based distillation method for TTM. To reduce the cost per step, we develop a simple, but powerful improvement to a recent layer distillation method that improves learning via better preserving hidden state variance. Finally, we combine our step and layer distillation methods together for a dual-faceted approach. We evaluate our step and layer distillation methods independently and show each yield best-in-class performance. Our combined distillation method can generate high-quality outputs with improved diversity, accelerating our base model by 10-18x (230/435ms latency for 32 second mono/stereo 44.1kHz, 15x faster than comparable SOTA) -- the fastest high-quality TTM to our knowledge.
1:20

Presto! Distilling Steps and Layers for Accelerating Music Generation Despite advances in diffusion-based text-to-music (TTM) methods, efficient, high-quality generation remains a challenge. We introduce Presto!, an approach to inference acceleration for score-based diffusion transformers via reducing both sampling steps and cost per step. To reduce steps, we develop a new score-based distribution matching distillation (DMD) method for the EDM-family of diffusion models, the first GAN-based distillation method for TTM. To reduce the cost per step, we develop a simple, but powerful improvement to a recent layer distillation method that improves learning via better preserving hidden state variance. Finally, we combine our step and layer distillation methods together for a dual-faceted approach. We evaluate our step and layer distillation methods independently and show each yield best-in-class performance. Our combined distillation method can generate high-quality outputs with improved diversity, accelerating our base model by 10-18x (230/435ms latency for 32 second mono/stereo 44.1kHz, 15x faster than comparable SOTA) -- the fastest high-quality TTM to our knowledge.

AK

30,430 views • 1 year ago

WHY Salah scores MORE with Ekitike ➡️ Ideal foil for Slot system ➡️ Trivela pass from LIF = tap in by Mo ➡️ Can drift deep & rotate with Mo ➡️ Becomes 4321 with Salah CF IP ➡️ Encouages LFC to dominate territory ➡️ Fast combo with Wirtz edge of box A THREAD 🧵 #LiverpoolFC
1:34

WHY Salah scores MORE with Ekitike ➡️ Ideal foil for Slot system ➡️ Trivela pass from LIF = tap in by Mo ➡️ Can drift deep & rotate with Mo ➡️ Becomes 4321 with Salah CF IP ➡️ Encouages LFC to dominate territory ➡️ Fast combo with Wirtz edge of box A THREAD 🧵 #LiverpoolFC

Pythagoras In Boots ⚽️

43,944 views • 10 months ago