LogoThread Easy
  • 탐색
  • 스레드 작성
LogoThread Easy

트위터 스레드의 올인원 파트너

© 2025 Thread Easy All Rights Reserved.

탐색

Newest first — browse tweet threads

Keep on to blur preview images; turn off to show them clearly

Read 👉 https://t.co/sByKoyBt2E

Read 👉 https://t.co/sByKoyBt2E

Maker of a tech note-taking app https://t.co/xcprYEEMJQ. My indie dev journey 👉https://t.co/JbdJ58XbTN 🇯🇵@craftzdog Dad👨‍👩‍👧 UTC+9

avatar for Takuya 🐾 devaslife
Takuya 🐾 devaslife
Fri Oct 31 08:36:51
cyberleads customer launched an outbound campaign a few days ago and already has promising calls 🖤

cyberleads customer launched an outbound campaign a few days ago and already has promising calls 🖤

Unemployed. Made my first $1M by helping agencies grow with https://t.co/NRrprPVGSl. Failed with 19 products, got lucky with 1. Taking a break and writing all about it.

avatar for Alex West 🚀
Alex West 🚀
Fri Oct 31 08:35:59
I used to be a heavy drinker, but I quit and haven’t kept booze at home since I started working out (alcohol slows down metabolism)

At events, I just get lime + tonic and pretend I’m drunk, people can’t tell when they’re drunk anyway, since alcohol impairs cognitive function

I used to be a heavy drinker, but I quit and haven’t kept booze at home since I started working out (alcohol slows down metabolism) At events, I just get lime + tonic and pretend I’m drunk, people can’t tell when they’re drunk anyway, since alcohol impairs cognitive function

“The woman scooping Silicon Valley” — BBC・hacker turned builder + blogger・ex: Threads, Instagram, startups, etc

avatar for Jane Manchun Wong
Jane Manchun Wong
Fri Oct 31 08:35:28
github:https://t.co/0JOYF5Ypli
HF:https://t.co/4uTyyR30XD

github:https://t.co/0JOYF5Ypli HF:https://t.co/4uTyyR30XD

致力于让每个想拥抱AI的人都能找到适合自己的AI产品,助力企业定制AIGC应用

avatar for AIGCLINK
AIGCLINK
Fri Oct 31 08:34:00
英伟达版的“Banana”,英伟达开源了一款物理级图像编辑模型:ChronoEdit-14B,静态图+一句话,4秒生成一张符合物理的图像

ChronoEdit具备“时间观念”和“物理常识”,它能理解动作,比如说“推倒”、“拿起”这种动作,可以根据它学到的物理常识,画出这个动作发生后,世界应该变成什么样子

8 步扩散完成一次图像编辑,在H100上约4秒一张图

从效果看,预测动作发生结果的能力比较强,改变姿势后,人物、服装、背景风格的一致性也保持的比较好,光影、反射效果处理的也比较符合逻辑

可以用在游戏制作、电影特效、机器人训练等场景上

#ChronoEdit #AI图像编辑

英伟达版的“Banana”,英伟达开源了一款物理级图像编辑模型:ChronoEdit-14B,静态图+一句话,4秒生成一张符合物理的图像 ChronoEdit具备“时间观念”和“物理常识”,它能理解动作,比如说“推倒”、“拿起”这种动作,可以根据它学到的物理常识,画出这个动作发生后,世界应该变成什么样子 8 步扩散完成一次图像编辑,在H100上约4秒一张图 从效果看,预测动作发生结果的能力比较强,改变姿势后,人物、服装、背景风格的一致性也保持的比较好,光影、反射效果处理的也比较符合逻辑 可以用在游戏制作、电影特效、机器人训练等场景上 #ChronoEdit #AI图像编辑

github:https://t.co/0JOYF5Ypli HF:https://t.co/4uTyyR30XD

avatar for AIGCLINK
AIGCLINK
Fri Oct 31 08:33:59
RT @weijunext: 分享一个很多人不知道的Google调试经验。

如果你确认配置正确,但就是无法登录成功,那么不用怀疑就是无法连上Google的服务。

解决办法很简单,梯子开启tun模式/虚拟网卡模式/增强模式就可以了。 https://t.co/p3pXEN0X…

RT @weijunext: 分享一个很多人不知道的Google调试经验。 如果你确认配置正确,但就是无法登录成功,那么不用怀疑就是无法连上Google的服务。 解决办法很简单,梯子开启tun模式/虚拟网卡模式/增强模式就可以了。 https://t.co/p3pXEN0X…

⛵️全栈工程师,出海开发者,Next.js 手艺人 | Next.js SaaS 模板:https://t.co/YzCDmGttme | Nextjs 中文文档:https://t.co/Ek37EFh3tI | 开源:https://t.co/vMeadXy9OD

avatar for WeNext(nexty.dev)
WeNext(nexty.dev)
Fri Oct 31 08:33:07
  • Previous
  • 1
  • More pages
  • 1608
  • 1609
  • 1610
  • More pages
  • 2127
  • Next