TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

DALL-E 2 open source implementation

520 点作者 jonas_kgomo大约 3 年前

14 条评论

christoph大约 3 年前
One thing I’ve not seen mentioned/asked about DALL-E 2, which I’m really, really intrigued about is what hardware its typically running on and the average time it takes for image generation? Maybe I’ve somehow missed it, but I’m really intrigued by what their hardware set up is and how long it on average takes to generate the eight images from a users text input? I’m guessing it’s really quick, but I honestly have no idea… How does this system scale? Typical cost per image generated?
评论 #31229228 未加载
评论 #31231691 未加载
评论 #31231136 未加载
评论 #31231594 未加载
lagrange77大约 3 年前
This technology always reminds me of the TNG episode Schisms, where the holodeck renders an alien site based on evolving text descriptions.<p><a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=lPLiaYD1lUo" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=lPLiaYD1lUo</a>
评论 #31234142 未加载
评论 #31231630 未加载
评论 #31231892 未加载
评论 #31230166 未加载
manquer大约 3 年前
The overlap between the hardware requirements for model training and web3 mining has me thinking could there be a coin where the proof of work could be training a model ?
评论 #31232160 未加载
评论 #31232621 未加载
评论 #31251738 未加载
dkga大约 3 年前
There&#x27;s been some discussion below on OpenAI not being open. Contrary to most arguments, I have to share that I like their current approach. We can all easily imagine the sort of negative things that could come out of the misuse of their models, and it is their responsibility to ensure it drips into public domain rather than just putting it out there. It also allows them to more carefully consider any implications that they might have missed during development. And even when you do have access to them, at least in my experience they make sure to properly caveat some uses even during runtime (eg, in their playground for the DaVinci model). So while I&#x27;m not associated with OpenAI in any way, I would like to go out and say I broadly support their &quot;closedness&quot; for now and hope we can all be in a place in the future well release of high-performance models does not need to be so carefully considered because people would be more reasonable overall.
评论 #31229656 未加载
评论 #31231995 未加载
评论 #31230595 未加载
djleni大约 3 年前
Software developer that knows nothing about ML: What would it take to mess around with this?<p>From a skim of the README it seems like install the package, and then find some training dataset, and then ??? to use it.
评论 #31229420 未加载
评论 #31230820 未加载
评论 #31229493 未加载
评论 #31229473 未加载
评论 #31287396 未加载
victor9000大约 3 年前
Is there a reason why the pretrained models are not included as well? Or perhaps this is coming at some point in the future? There are many users who simply want to play around with the functionality, who lack the time and hardware required to replicate the training steps.
评论 #31230773 未加载
评论 #31231059 未加载
评论 #31231608 未加载
natly大约 3 年前
Lucidrains is an incredible guy. Nice guy in his discord too.
评论 #31229219 未加载
评论 #31229049 未加载
评论 #31228845 未加载
评论 #31228881 未加载
评论 #31232099 未加载
arsome大约 3 年前
Is it just me or is OpenAI the exact opposite of open?<p>They may as well just call it ClosedAI at this point, seems like every interesting thing they do is kept under wraps.
评论 #31229168 未加载
评论 #31229332 未加载
评论 #31229357 未加载
评论 #31231654 未加载
评论 #31229253 未加载
评论 #31230091 未加载
评论 #31239085 未加载
评论 #31229245 未加载
评论 #31229539 未加载
评论 #31234375 未加载
评论 #31229507 未加载
londons_explore大约 3 年前
This needs distributed training...<p>Years ago I made a shared tensor library[1] which should allow people to do training in a distributed fashion around the world. Even with relatively slow internet connections, training should still make good use of all the compute available because the whole lot runs asynchronously with highly compressed and approximate updates to shared weights.<p>The end result is that every bit of computation added has <i>some</i> benefits.<p>Obviously for a real large scale effort, anti-cheat and anti-spam mechanisms would be needed to ensure nodes aren&#x27;t deliberately sending bad data to hurt the group effort.<p>[1]: <a href="https:&#x2F;&#x2F;github.com&#x2F;Hello1024&#x2F;shared-tensor" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;Hello1024&#x2F;shared-tensor</a>
a1371大约 3 年前
Sorry I&#x27;m being too lazy to run the code, but are there any examples of its output? Like, how comparable is it to the openAi&#x27;s
评论 #31229464 未加载
mensetmanusman大约 3 年前
We are at the beginning of a post verification society.<p>Any image or video of anything will soon be faked by commoners. This will be a wild ride!
评论 #31230094 未加载
评论 #31230033 未加载
评论 #31230115 未加载
imwillofficial大约 3 年前
I’m not an AI person, when can I use my MacBook to make crazy cool images I’ve seen on various twitter threads via DALL-E 2?
评论 #31249701 未加载
评论 #31247956 未加载
dannyw大约 3 年前
I remember lucidrains from his site, Epicmafia. Good times! Made so many friends there, many of whom I still talk to today.
DitheringIdiot大约 3 年前
This is a wonderful thing. But does it have any positive applications outside of entertainment?<p>I can think of lots of malicious things it can be used for.
评论 #31228991 未加载
评论 #31229034 未加载
评论 #31228866 未加载
评论 #31229676 未加载
评论 #31230425 未加载