LLAMA 3 vs Stable Diffusion 3 vs DALL-E 3 - Prompts and Images
VloĆŸit
- Äas pĆidĂĄn 21. 04. 2024
- LLAMA 3 can produce prompts and render images, SD3 Is now available via Comfy and why this might be better than DALL-E 3
Join đ this channel to get access to exclusive content and perks:
bit.ly/XOVjoin
As an Amazon Associate Pixovert earns from qualifying purchases, as well as from associations and partnerships with other retailers, advertisers and sponsors.
Beginner's Guide to ComfyUI and Stable Diffusion
CZcams - bit.ly/BeginCOMFYUISDXL
Udemy - bit.ly/SDGenStart - VÄda a technologie
Hi, I am a hobbyist photographer playing with AI since the beginning of this year. Still looking after joining photography and AI into something more than "pop like visuals"... Moving between bit abstract graphics and analog photo style in conjunction with Loras trained on me and my wife. Lot of fun, not yet got the idea for any serious project. Recently fell in love with stable cascade due to its ability to do what I really like, double exposures, different analog techniques (even wet plate). Now I'm going to train stable cascade with my own works style. For now SDXL looks really artificial... :) will give SD3 a shot for sure.
Cascade is nearly as good as SD3, but its still at primary stage, I wonder what it will be like if we get to Cascade 3đ. I've struggled with double exposures in cascade, but it is really powerful.
Can't wait to see SD3 community models + controlnets, will be crazy good.
Also, to me right now, Ideogram is the best amongst all AI in terms of art+prompt adhesion to me. (midjourney still the best in only art, but ideogram better than dalle 3 and sd3 to me)
The SD3 images look so overbaked right now cant wait until the weights are released.
The model is still being developed, the images you see here were mainly random prompts - once we can prompt professionally the results will be 3x better
why didn't you used midjourney? It may have done better than chatgpt
ChatGPT has a lot more functionality than midjourney or even stable assistant. I would still need to replicate that functionality using other GPTs. Groq and Perplexity AIs are pretty good at some of these functions, and Meta's Llama seems promising too.