Why is it not recognising celebrity/famous names when trying to input them into the prompt box?..I did type in “Taylor Swift” yesterday (simply because she’s so well known) & after waiting an age got something resembling her,but just tried “River Phoenix” and got firstly a woman then an actual phoenix…is this because the whole thing has started from scratch?
Quote from the notes you can find by clicking the little link that appears when you are generating an image:
This new image model is still training. It’s currently worse than the previous one at some things. It’ll steadily improve in quality and gain new knowledge over the next few months.
MONTHS
Training takes a long time and a lot of the GPU power on the hardware this model runs on (part of the speed issue). It is getting better as more training data completes, but please be patient.
“Months” though!!!..how is this an improvement?..seems like we’ve gone so far backwards.
To use a loose analogy, that’s like getting a new (much better) car and calling it worse than the previous one because the fuel tank hasn’t been filled with gas yet.
Training is the bread and butter that makes an image generator work. It is the reference data that it pulls from to create images. Without it, there is no “fuel” for the engine to run, so the results will be sub-optimal.
And, again, training is an extremely GPU intensive process. The fact the dev is able to run the training data and still support image gen at a usable level is pure wizardry.
Ok,but what is the reason for this update if it’s putting us all back who knows how many months…all seems a bit daft to say the least.
Perhaps my analogy wasn’t very good?
Flux is a dramatically better image gen model. Most consider it better than SDXL, yet still lightweight enough (with a bunch of optimization) to run on the limited resources the dev can afford. When the training is complete, this model will be far more capable then the old one ever was, but the training needs to finish for it to get better.
I’m sorry you are having trouble with this, but a large number of people have complained about the quality of the old model, yet there is no way to fix those issues without upgrading the model. Now that the old model is being replaced, all many people can do is complain, rather than be thankful that the dev is working hard to give us a better model.
Fair enough - I hate to complain but for me the site was perfect as it was - if in 6 months it’s as good as it was then great - but it’s a long time to wait though to get back to where we were!
I would appreciate furry characters such as Leomon from digimon, Amicus from Adastra, LinHu from Nekojish
its probably because of all the celebrities attacking and suing ai for deepfake recreating them (and Artists complaining about theft of their work) that happened after 1.5 causing them to change that aspect of Flux. total guess i didnt look up.
at least twilight sparkle and mlp escaped it
