Hello everyone, Rodney from Kleebz Tech ,
and I'm excited to announce a new update for Fooocus version 2.2.0. If you're running Fooocus
normally, it should update itself automatically when you start it. Today, I'll highlight some of
the key new features in this update. I'm not going to cover every little tiny update because some
of them I'm not even exactly sure what they do, and I don't think they pertain to how I'm
using Fooocus. One of the first things that you're going to notice when you launch the new
Fooocus update is, one thing I've noticed is, under the setting tab in the advanced section,
the quality, speed, and extreme speed buttons have been rearranged, which actually makes more
sense now on how these are set up. You're also going to notice that when you scroll down,
you now have the option to change the output format. The default has always been PNG, but
now you can choose JPEG or WEBP, and those also can be changed in your default config file
as well. So, if you're looking to change those, you can actually add this line to your config
file, and that will set the default every time that you load up Fooocus. Another new thing
that a lot of people I know have run into as an issue is if you had your output folder in a
separate directory then your root Fooocus folder, you couldn't click on the history log. You would
actually have to go into the folder and open it up that way. But now, since the latest update, if
that's even outside your anywhere on your system, you should now be able to click on that, and it
will load up your log file. Like here, mine is kept in a totally different location, and now it
loads up fine. Now, the next obvious change would be under the models tab. If you look in here,
you're going to now see there are checkboxes, so you can enable and disable each LoRA. So that's
actually a new feature as well. And since we're on the subject, the LoRAs actually has a couple of
other changes that pertain to that. You can now have the ability to load checkpoints and LoRAs
from multiple locations. So, you can have them on different drives. So, if you don't have enough
room on just one drive, you can actually put them in a different drive. You would change your config
file something like this. I did a little bit experimenting and found that this is the format of
how you would do it. You'd put the square brackets around it, and you can add multiple locations, and
it would be the same thing for the LoRAs as well. You can do the same thing with those. Another
thing is if you look on mine, mine is actually now showing six instead of the normal five, and
that's another new feature that can be added. Max default LoRA number, you can add to your config
file and change it to whatever number you want. If you want 20 in there, you can change it to 20.
And while we're still on the subject of LoRAs, the other thing you can do is before, when you,
the weight for the LoRAs were either -2 to +2, you couldn't change it anywhere higher or lower. I
don't normally, but apparently, some people do, so they've added the ability now that you can change
your default LoRA's minimum and maximum weight. If you add these to your configuration, it will allow
you to adjust those numbers to whatever you want in there. So, for example, on mine, I have mine
set. I was experimenting, and now I can increase it up to +4 or down to -4. Now, another new minor
thing that's been added, I don't, I use my own wild cards for the most thing, most part, but they
have added a new Wild Card file. So there's now an animal one, 100 most popular animals, on in this
list. So you do have that now as an option in the wild cards. Since we're on the subject of wild
cards, I know a lot of people in the past, I've seen when they talk about wild cards, they like to
be able to have them do them in order instead of being entirely random. Well, they've actually
added now support for arrays. Now, the arrays are a little bit different. They're not separate
files. They're actually added to your prompt. So, for this example here, I will go ahead and what
you want to do for the arrays is you use the double square bracket to start and then to close
it, and then each one is separated by a comma in between. And you can use multiple arrays as well.
That's not an issue. I haven't played around with nested arrays or anything yet, but I imagine
there's probably a good chance you can do that, uh, if you want to have arrays within arrays. So,
what this will do is when we go ahead and generate now, it's going to generate, well, let's go ahead
and I'll show you exactly how it will do it. Okay, so now, as you can see here, it went through,
it did a red, blue, green, and yellow dress in the order that I specified. Another interesting
aspect of this is if we go down into here, we will notice that, as usual, the seed number increments.
So we have here it ends, you know, you can see the 179, the 180. This, by default, will generate each
image, and that's how normally Fooocus works. Once it uses the first image for the seed, whatever
seed it uses for the first image, it then uses the next one in order sequentially to generate
the next image. That's how it works by default, and normally, you wouldn't have much of a reason
to change that because if you don't change anything else and you use the same seed, you're
always going to get the same images. Let's say you want to try different things but on the same
seed; well, this isn't going to help you if you do it that way. But there is another option, another
feature that they've added. If you go into the advanced section under the developer debug mode
and go down to the bottom, down by debug tools, you're going to find some new options here, and
I'll cover these as we go. But you'll see here, well, first, I'm actually going to cover this
while we're at it because this is actually another option. Basically, you can disable
the intermediate results. What that means is, when it's generating, it won't show you after
it produces each image the resulting image. So if you're generating eight images, you won't see
the final, you won't see each image as it's done generating; it'll just show them all at the end. I
have no need for that myself, but then down here, you're going to see 'disable seed increment.' So
what this feature does is now that I've checked that off, it will use the same seed number for
each image. Now, this could be under random, or it could be under, you could go with the
set seed, that's up to you, but at this point, when I hit generate, it's going to, because I have
it checked on random, it's going to pick a random seed number, but for each of those images it goes
through, it's going to keep using that same seed, but then it will just change each array item. So,
I'll still get the differences, and I can see what difference I get from each image but using the
same seed. So, and this can be used for many other things. I'm using the color of the dress
for one example, but you could use it for just trying different other different prompts. Okay,
so we have our four images generated now. If we go into the log and pull that up, and we look in
here, we'll notice this seed number is the same as this seed, and this one, and this one. If you want
to experiment, testing things out, this is great because now you can actually have it generate
a bunch of images with the same seed and use an array to change slightly different parameters,
and you can have it cycle through all of those. That's actually a pretty useful feature, and I'm
actually pretty excited to be able to use that. In all honesty, that's actually one of the big ones
that I'm really happy to see added. Now, another thing that appears to be a change as well, and
I'm not too familiar with this because I don't run mine using for the CPU, but my understanding is
if you're using a CPU for generation, you can now specify the number of threads that are used using
the command line argument. I haven't tested this; I don't know much about it, but my understanding
when you use that always CPU flag, you can add a number after it, and that will set the number of
threads that are used. Like I said, I don't know much about that because I haven't played around
with that myself yet. And last but not least, and I'm not going to be covering, like I said,
every single change, I'll put those up on screen, probably the other big change that a lot of people
have wanted is along with the new image formats that you can save at, if you go into your advanced
section under the developer debug mode, we can go down here, and below the 'disable seed increment,'
we also have the ability to now save metadata to images. Fooocus, by default, for privacy reason,
doesn't do that; it only adds it to the log, so you can click on the 'copy to clipboard,'
but just like that, you can now check this off, then you can choose which format you want. I'm
not going to dive into which ones do which; you can now have it save the metadata to the images,
and you can also see you can change that in your config file, so you don't have to go in there and
change that every time you launch Fooocus. You can go in and change the default to true, and then
whichever scheme that you want from that point, so that can be done in the config files as well. Now
when we generate an image, all that information is going to be saved in the image itself, so we can
bring that back in later. So for this example, I'm just going to cut this down to one. I'll
generate a new image here. For privacy reasons, if you don't want this embedded into all your
images, it's probably a good idea to not have that as being default checked off, but that's up
to you. Do keep that in mind when you do that; it's going to bring all that and add that to the
image. So, and that was one of the reasons they didn't add that feature originally to Fooocus,
but now you do have the option to enable it if you want. Okay, so we have our image now. Now I
can still do it the old way, I can click on 'copy on clipboard' and go into here and hit paste, and
then I can load the parameters. That still works, and this actually works pretty much the same way,
but instead of going into the log to do that, that's now saved with your image. So all I need to
do, and this comes down to the another new feature they've added, we check off the 'input image'
checkbox, and you're going to see a new tab in here called 'metadata.' So you can drag that
image that you've created into that top box there, and then after a couple of seconds, it's going
to show you all the metadata that's attached to that image, and at that point, you can either
hit the 'copy' button, or if you want to just apply it directly to your prompt, we can hit
'apply metadata' down at the bottom. Actually, let me go ahead and actually demonstrate. We're
going to change a couple of these settings just to show you that it will actually do what we
want. So, okay, so we have our image down here, we have the metadata, so as you saw, I don't have
anything in the prompt, and those settings aren't set, so now I'll hit 'apply metadata.' Let's
go back up; you'll see that everything has been changed back to the way it should be.
It's got the correct seed number in here, and I can go ahead and regenerate that image again
just by dragging and dropping that in. So that's all the changes I'm going to be covering in this
video directly. This is a list of the rest, all the changes that were made, if you want to look
those over. You can also go to the GitHub page. I do notice there's some stuff about Docker and some
other stuff in here as well. So that's pretty much it for this video on the latest changes, and if
you found it useful, please do consider clicking that like button because it definitely does
help, and until the next video, have a great day.