Hey Bob, just thought I'd share my tip. I've found whilst creating models if I use high res head shot photos (typically shot from a dslr in raw/dng) and then cropping the pic in Photoshop to around 2000 x 2000px resolution for each jpg shot to upload (around 20-25 pics - your sweet spot tip ;) I get some incredible results compared to uploading FB or other online pics. Very accurate. I don't know whether this matters or not to the training algorithm but I guess more detail is better than less! I would really love to see the ability to use multiple models in one image tho, something OpenArt struggles with. That would be truly game-changing. And better image/folder organisation like in Midjourney. OpenArt is pretty awesome tho!
Hey Bob, I don't think there's currently a way to use two models in one photo, since your in communication with openart, could you ask them to implement a feature like that?
@@gumvue.studio these are Lora files built on top of the flux dev model so you’re bound by the licensing terms of that. But from a technical point of view, these Lora’s will run on any system thatsupports Lora files. So there’s nothing stopping you from sharing them.
@@jazzampa well, that’s simply not true. It’s a great site. I absolutely stand by it. Maybe you could be more specific with a comment like that. What part is crap to you?
That's a classless ad hominem attack. You're basically saying it's crap because that's your opinion. Either bring the evidence or be quiet. Otherwise it reveals you as shallow and not worth listening to.
Hey Bob, just thought I'd share my tip. I've found whilst creating models if I use high res head shot photos (typically shot from a dslr in raw/dng) and then cropping the pic in Photoshop to around 2000 x 2000px resolution for each jpg shot to upload (around 20-25 pics - your sweet spot tip ;) I get some incredible results compared to uploading FB or other online pics. Very accurate. I don't know whether this matters or not to the training algorithm but I guess more detail is better than less! I would really love to see the ability to use multiple models in one image tho, something OpenArt struggles with. That would be truly game-changing. And better image/folder organisation like in Midjourney. OpenArt is pretty awesome tho!
my online presence quadrupled x10 after uploading my first openart AI images to my social media platforms!
They truly do rock
Your job is just, so so much helpful, thanks again.
I'm digging the new flux realism model with Kling 1.6
I like your vibe, and always bringing cool Ai, can you please talk about a preferably a free way to enhance prompts for different Ai? Thanks ☺️
great video !!!
Kindly do a video on the one that work on mobile phones....... I mean the one similar to xpression camera
Hey Bob, I don't think there's currently a way to use two models in one photo, since your in communication with openart, could you ask them to implement a feature like that?
@@WhitneyAiVocals that is in the works and should be available first couple of months of next year.
could you please teach us how to make a consistent graphical style using a lora for design?
@@Xandercorp can you give me an example? It absolutely supports styles, but I don’t know what you mean by “graphical design“ necessarily.
With which train model OpenArt train models? LoRA/LoKr/DoRA?
9:58 cheers!
I bet what's next is movies!
thanks Bob, question , can I train models and sale them ? or share?
@@gumvue.studio these are Lora files built on top of the flux dev model so you’re bound by the licensing terms of that. But from a technical point of view, these Lora’s will run on any system thatsupports Lora files. So there’s nothing stopping you from sharing them.
@@BobDoyleMedia I dont knwo the terms for Dev Licence, but I guess if I charge small amount and do for other people .. that will be totaly fine?
Great.. , is it free.? Good free tier???
i sign up free with free credit and it wont let me train my lora models says i have to pay
Im telling you. There's something with the audio :D
@@Xandercorp it happens to me when I use my Bluetooth headset. But other than that, it looks normal to me.
@@BobDoyleMedia kool, take it easy
Not sure i need watch another AI channel at this point.
first
Guys, don't do it. The site is totally bull crap
@@jazzampa well, that’s simply not true. It’s a great site. I absolutely stand by it. Maybe you could be more specific with a comment like that. What part is crap to you?
That's a classless ad hominem attack. You're basically saying it's crap because that's your opinion. Either bring the evidence or be quiet. Otherwise it reveals you as shallow and not worth listening to.