ทวีตที่ปักหมุด
Yuuki Ai
2.7K posts

Yuuki Ai
@YuukiAi
Prompt Engineer & Digital Artist Specializing in generative AI workflows for art, video, and music synthesis. Passionate about technology and storytelling
United States เข้าร่วม Ekim 2023
342 กำลังติดตาม696 ผู้ติดตาม

@keshiAIart This was fun to experience. Keshi always makes the best content 🥰
English

@ByTheBroken1 Yeah it's very similar to a1111 and forge so you'll get along with it well if you're already used to those ui
English

@YuukiAi Im unfamiliar with Stability Matrix. is that a new Stable Diffusion UI?
English

@ByTheBroken1 Not by me. I used it until forge and then just a week ago finally moved from forge to stability matrix
English

@YuukiAi Will I be judged harshly if I admit to still using Auto 1111?
English

@SoftRenderAi I have a 4070 and yeah, it hates flux 😆 but it also complains if I crank the resolution and use 3 LoRA so 🤷♀️
English

@YuukiAi Spent countless ours experimenting with automatic1111 sd1.4 then 1.5, then flux came out and the machine could not keep the pace 🤣.
English

@YuukiAi The issue with that is data control and licensing.
Anything you do on a VM or on other people's hardware can be scraped.
I have the money just my motivation has been really dead with a lot of the external factors and running out of steam to finish things I started.
English

@hyenosaurus Grok is really good for video. I like to use my own stuff from forge on grok video.
English

@YuukiAi StableDiffusion, then Copilot, then Grok, now ChatGPT Plus. Grok is still great though.
English

@YuukiAi I started on a bootleg version of NovelAI that predated SD, when the first model escaped to the wild. Good times
English

@YuukiAi I used NightCafe first, then picked up Forge. I use Forge NEO now but I need a better card if i wanna do video.
English

@DarkInfinityArt Don't tell anyone, but I still don't know how to properly train models. I know there's more than just having images to use but I never learned it myself since so many LoRA are just simply out there.
English

@YuukiAi I started with SD1.5 with Auto1111, but didn't get into LoRA training and Model finetuning until SDXL and comfyui.
English

@Holezinya1 I was messing with midjourney just today because of grok rate limits! I can't believe how good mj can look and yet still seems to struggle with hands. Or is that just me?
English













