正在加载视频...

视频加载失败

返回首页

damn. spatial intelligence can be FAST asf! "D4RT can continuously understand what's moving while running 18x–300x faster than previous methods - processing a 1-minute video in roughly 5 seconds on a single TPU chip."

bilawalsidhu's profile picture

Bilawal Sidhu

105,018 subscribers

33,587 次观看 • 9 个月前 •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

A bear can reach 40 mph, faster than a sports car; running can be fatal. [📹 tom.maley]
0:11

A bear can reach 40 mph, faster than a sports car; running can be fatal. [📹 tom.maley]

Massimo

681,308 次观看 • 10 个月前

We’re continuously investing in our network so millions of homes and businesses can enjoy speeds faster than a gig.
0:15

We’re continuously investing in our network so millions of homes and businesses can enjoy speeds faster than a gig.

Comcast

1,159,076 次观看 • 3 年前

Grok-1 support in gpt-fast at faster(?) than anyone else has reported so far. 75 tok/s for a 300B+ parameter model on an 8xA100 node. If I understand correctly, ColossalAI reported 15 seconds to generate 100 tokens. gpt-fast takes 4.2 seconds to generate *400* tokens.
0:10

Grok-1 support in gpt-fast at faster(?) than anyone else has reported so far. 75 tok/s for a 300B+ parameter model on an 8xA100 node. If I understand correctly, ColossalAI reported 15 seconds to generate 100 tokens. gpt-fast takes 4.2 seconds to generate *400* tokens.

Horace He

54,193 次观看 • 2 年前

SAM + Optical Flow = FlowSAM FlowSAM can discover and segment moving objects in a video and outperforms all previous approaches by a considerable margin in both single and multi-object benchmarks 🔥
0:13

SAM + Optical Flow = FlowSAM FlowSAM can discover and segment moving objects in a video and outperforms all previous approaches by a considerable margin in both single and multi-object benchmarks 🔥

Dreaming Tulpa 🥓👑

157,793 次观看 • 2 年前

Dogecoin Got a 300X Speed Boost Fintech firm, Metallicus, launched an alpha version of DogecoinVM on Metal Blockchain mainnet. According to the company, the processing speed of transactions can be up to 300 times faster than before. Marshall Hayner, CEO of Metallicus, referred to it as a response to Elon Musk’s (Elon Musk) request to make dogecoin:native transactions faster. The DogecoinVM is a new layer, independent from Dogecoin (Dogecoin) Core, Transactions reportedly get settled in 0.15 to 0.25 seconds, although deposits are limited to 100 DOGE.
0:15

Dogecoin Got a 300X Speed Boost Fintech firm, Metallicus, launched an alpha version of DogecoinVM on Metal Blockchain mainnet. According to the company, the processing speed of transactions can be up to 300 times faster than before. Marshall Hayner, CEO of Metallicus, referred to it as a response to Elon Musk’s (Elon Musk) request to make dogecoin:native transactions faster. The DogecoinVM is a new layer, independent from Dogecoin (Dogecoin) Core, Transactions reportedly get settled in 0.15 to 0.25 seconds, although deposits are limited to 100 DOGE.

BSCN

39,176 次观看 • 25 天前

A mongoose can react to a cobra's strike in just 0.06 seconds, which is faster than the blink of an eye
0:10

A mongoose can react to a cobra's strike in just 0.06 seconds, which is faster than the blink of an eye

Wildlife Uncensored

46,734 次观看 • 6 个月前

A young father has smashed the world record for running a mile with a pushchair, finishing nearly nine seconds faster than the previous mark. Read more about how Stephen Biebelhausen ended up running with his daughter 👇
0:11

A young father has smashed the world record for running a mile with a pushchair, finishing nearly nine seconds faster than the previous mark. Read more about how Stephen Biebelhausen ended up running with his daughter 👇

The Telegraph

21,272 次观看 • 2 个月前

Simplex, now running at 200Hz in the commonware Library. Sequence blocks faster than a typical monitor can refresh. Finalize them across a global network in 300ms.
0:13

Simplex, now running at 200Hz in the commonware Library. Sequence blocks faster than a typical monitor can refresh. Finalize them across a global network in 300ms.

patrickogrady.xyz

14,373 次观看 • 2 个月前

A few seconds from prompt to video — that's faster than Twitch streams live. H3 Max on MiniMax Design, post-trained by fal. The speed-and-cost combo is what really stands out. You can iterate in seconds, not minutes.
0:15

A few seconds from prompt to video — that's faster than Twitch streams live. H3 Max on MiniMax Design, post-trained by fal. The speed-and-cost combo is what really stands out. You can iterate in seconds, not minutes.

KICHOCHEZI

311,115 次观看 • 2 个月前

UI selection seems much faster, you can lock in legit 2-3 seconds if you're fast with the selection
0:12

UI selection seems much faster, you can lock in legit 2-3 seconds if you're fast with the selection

Joseph Ugarte

80,223 次观看 • 1 个月前

(1/n) 🚀 With FastVideo, you can now generate a 5-second video in 5 seconds on a single H200 GPU! Introducing FastWan series, a family of fast video generation models trained via a new recipe we term as “sparse distillation”, to speed up video denoising time by 70X! 🖥️ Live demo: (Thanks to @gmicloud for the support!) 🔗 Blog: 🔓 We fully open-source our models, code, and data with Apache-2.0 licenses
0:13

(1/n) 🚀 With FastVideo, you can now generate a 5-second video in 5 seconds on a single H200 GPU! Introducing FastWan series, a family of fast video generation models trained via a new recipe we term as “sparse distillation”, to speed up video denoising time by 70X! 🖥️ Live demo: (Thanks to @gmicloud for the support!) 🔗 Blog: 🔓 We fully open-source our models, code, and data with Apache-2.0 licenses

Hao AI Lab

79,140 次观看 • 1 年前

Video quality will probably be garbled but I updated the lightbox on my site so that the maps can be viewed in fullscreen & zooms in on a single click rather than double. Should be a lot more intuitive & fit nicer on a 2nd monitor.
0:15

Video quality will probably be garbled but I updated the lightbox on my site so that the maps can be viewed in fullscreen & zooms in on a single click rather than double. Should be a lot more intuitive & fit nicer on a 2nd monitor.

REƎMR

95,145 次观看 • 11 个月前

NVIDIA open-sourced a 600M model that transcribes 40 languages in real-time at 80ms latency and it costs $0. that's faster than you can blink. across mandarin, arabic, hindi, portuguese, tagalog, whatever,from a SINGLE checkpoint. → 17x more concurrent streams than buffered ASR on the same H100. → punctuation + capitalization built-in. no post-processing. → runs on your own GPU. no API bill 100% Open Source.
0:17

NVIDIA open-sourced a 600M model that transcribes 40 languages in real-time at 80ms latency and it costs $0. that's faster than you can blink. across mandarin, arabic, hindi, portuguese, tagalog, whatever,from a SINGLE checkpoint. → 17x more concurrent streams than buffered ASR on the same H100. → punctuation + capitalization built-in. no post-processing. → runs on your own GPU. no API bill 100% Open Source.

Superman

105,212 次观看 • 3 个月前

BIOS Dev Update: Fast Chat, Thinking Traces & More What's New: • Fast Chat Mode: A faster pipeline for simple questions that switches to literature search only when needed. It runs in an agentic loop, searching as many times as it needs until it has enough context to give you a complete answer. • Thinking Traces on Screen: BIOS now shows its reasoning on the main loading screen while Deep Research is running, so you can follow along in real time.
0:13

BIOS Dev Update: Fast Chat, Thinking Traces & More What's New: • Fast Chat Mode: A faster pipeline for simple questions that switches to literature search only when needed. It runs in an agentic loop, searching as many times as it needs until it has enough context to give you a complete answer. • Thinking Traces on Screen: BIOS now shows its reasoning on the main loading screen while Deep Research is running, so you can follow along in real time.

BIOS

10,285 次观看 • 6 个月前

🔥Spatial intelligence needs fast, *interactive* 3D world generation 🎮 — introducing WonderWorld: generating 3D scenes interactively following your movement and content requests, and see them in <10 seconds! 🧵1/6 Web: arXiv:
0:10

🔥Spatial intelligence needs fast, *interactive* 3D world generation 🎮 — introducing WonderWorld: generating 3D scenes interactively following your movement and content requests, and see them in <10 seconds! 🧵1/6 Web: arXiv:

Hong-Xing (Koven) Yu

169,241 次观看 • 2 年前

Luffy in Gear 5 can literally turn from a goofy man into a monster in 5 SECONDS
0:16

Luffy in Gear 5 can literally turn from a goofy man into a monster in 5 SECONDS

𝐌𝐚𝐫𝐜𝐨

1,206,176 次观看 • 2 年前

You should never run from a charging elephant. Elephants are much faster than people and running can actually make it chase you harder.
0:13

You should never run from a charging elephant. Elephants are much faster than people and running can actually make it chase you harder.

Mike Sonko

278,315 次观看 • 1 年前