SeASnAkE

@hjessmith Wow, fantastic work! 🎉 Is this an updated iteration of animated drawing? Really hoping there are open source plans in the works - this looks absolutely amazing and the community would love to contribute! Can't wait to see what's next! 🚀

Excited to share that our demo, Animating MetaGen Stickers, has been accepted to the demo track of #CVPR2025! Come say hello and play around with the demo yourself if you're attending CVPR! We'll be set up June 13th in the afternoon session.

@felixsanzm @StabilityAI now flux comes!

When you write for 1 whole month a huge article on how to optimize Stable Diffusion XL and within minutes of publishing it, @StabilityAI announces Stable Diffusion 3

@KBlueleaf @Datou 大佬 主板 cpu 啥的能分享下吗🙏

@Datou 3090 nvlink只能雙卡接 說實在話沒啥用 而且DDP也不需要nvlink 不會增強多少

Time to train few more epoch

@op7418 所以stable cascade支持的分辨率都有哪些啊 ?

昨天 Stable Cascade 创建图片有伪影的问题解决了，主要原因是我随便打得图像比例和分辨率不是Stable Cascade支持的分辨率。 更换成 2048*1152 就没问题了，没有伪影的图片质量相当不错，下面是昨天同样提示词生成的四张图，2K 分辨率直出。

@ZHOZHO672070 @Kijaidesign 能发个workflow吗

在 ComfyUI 中初步测试了一下 X-Adapter ，效果真不错！！！1.5和XL生态打通的感觉太好了，速度也很快，cool！！！ 陆续的测试内容我都会放在这个帖子里 目前使用的是 @Kijaidesign 做的插件（开始没找到这个插件，这里要感谢摆渡哥的消息）：github.com/kijai/ComfyUI-… 希望 ComfyUI 未来能直接支持

@ZHOZHO672070 为什么英文版被删除了

卧槽！Stable Cascade 新模型架构啊！一会详细看一下，效率和质量都提升了，并且还兼容Controlnet、Lora，未来会公开训练和微调代码 官方博文：ja.stability.ai/blog/stable-ca…

@shengxj1 好耶好耶

开个系列，锐评一下热门书籍

@naiivememe timing

The ComfyUI workflow image is the implementation of all 4 methods above

StableDiffusion Upscaler method 1. AI res upscale model （ a41 extra tab） 2. generate in small res first， then Upscale latent in latent space （a41 high res） 3. Use AI resolution upscale model first，then split the image and inpaint every tile 4. kohya deep shrink

开始搞实习报告，nnd

md 我还以为 x 能修改推文，原来也要会员，以为和即刻会员差不多价，果然，是我太天真了，哈哈哈。来一点小小的美元震撼。

赛博朋克：刷正在关注的 AIGC 刷到焦虑，小手往右一划，来到推荐黄推，刷到贤者，小手再往左滑回来，继续卷

@DrJimFan @shipin_xiazai

Overfitting in embodied agents😅

@HongKong_Doll @shipin_xiazai

要被撐大啦 24小時發💋 DM in 24 hours, RT & Like now 💋

New short course on Fine-tuning LLMs! Many developers are moving beyond only prompting, to also fine-tuning LLMs - that is, taking a pre-trained model and training it further on your own data, which can deliver superior results inexpensively. In this course, @realSharonZhou, CEO of Lamini (disclosure: I’m a minor shareholder) shows you how to recognize when fine-tuning can be help, and how to train an open-source LLM on your own data. I hope you enjoy the course! deeplearning.ai/short-courses/…

@auditore_k 大佬 在数据集收集上有什么建议吗

Today is Texas the Omertosa! Tried a more refined background #AI #stablediffusion #arknights #Texas #アークナイツフェス #明日方舟