> The prompt comprehension is incredible! #auraflow
> "a cat that is half orange tabby and half black, split down the middle. Holding a martini glass with a ball of yarn in it. He has a monocle on his left eye, and a blue top hat, art nouveau style "
Plus an image that somewhat resembles that prompt. The cat has a human-like hand with a chopped off thumb and 6 fingers in total, differently colored eyes, a branch in front of its face, the ball of yarn is somehow floating in mid-air.]
These are somewhat valid issues. But given the currently available open models, this is a massive improvement. The human-like hand and changing the styles on the sides of the head isn't even bad - those are valid artistic choices you'd see on similar illustrations - they're just badly executed here.
That's a marketing opportunity being missed, especially given how crowded the space is now. The HN crowd is more likely to run it themselves when presented with signing up just to test out a single generation.
Uh, thanks for noticing it! We generally turn it off for popular models so people can see the underlying inference speed and the results but we forgot about it for this one, it should now be auth-less with a stricter rate limit just like other popular models in the gallery.
I wanted to use your service for a project but you can only sign in through github, I emailed your support about this and never got an answer, in the end I ended up installing SD Turbo locally. I think that a github only auth is losing you potential customers like myself.
I just got rate-limited on my first generation. The message is "You have exceeded the request limit per minute". This was after showing me cli output suggesting that my image was being generated.
I guess my zero attempts per minute was too much. You really shouldn't post your product on HN if you aren't prepared for it to work. Reputations are hard to earn, and you're losing people's interest by directing them to a broken product.
We believe chaining different models can lead to impressive user experiences and as an AI product owner you can really differentiate yourself from others if you use several models in creative ways.
In the example there is python code to do the model inference as well as the javascript code to build the application. I believe this would be a great reference implementation for people trying to build their own AI apps.
