XPost: comp.os.linux.advocacy
On Thu, 18 Jul 2024 21:27:43 -0400, CrudeSausage <
crude@sausa.ge> wrote in <tijmO.52531$
xYQc.5429@fx12.iad>:
On 2024-07-18 8:17 p.m., vallor wrote:
On Thu, 18 Jul 2024 20:10:41 -0400, DFS <nospam@dfs.com> wrote in
<v7cau1$2l49c$1@dont-email.me>:
On 7/18/2024 7:56 PM, vallor wrote:
On Thu, 18 Jul 2024 19:29:14 -0400, DFS <nospam@dfs.com> wrote in
<6699a547$0$1733$882e4bbb@reader.netnews.com>:
What's the exact 'text input' you submitted that returned that
image?
"close-up of black-haired sicilian woman wearing a summer hat in a
field of flowers at noontime"
Cool.
Do you get a different pic every time you submit that description?
Yes, unless you feed it the same seed. (By default, it uses a random
seed.)
And here is "angels are the heros":
https://imgur.com/LbVJf2j
Get Fooocus and try it!
I tried it 5x at https://fooocus.one/playground
Each time I got:
"Error This application is currently busy. Please try again."
I hadn't considered that anyone would try to run it as a hosted app.
Run it locally on your own GPU:
https://github.com/lllyasviel/Fooocus
I installed it on Windows and while it makes nice images, I prefer the results I get with Microsoft Designer through Edge. Of course, the
latter allows you to write in plain English. I imagine that what I'm
putting in the prompts is not correct.
Not sure I understand. My prompt is in "plain English", but it is a description:
"close-up of black-haired sicilian woman wearing a summer hat in a
field of flowers at noontime"
BTW, under Foocus/outputs are subdirectories (by date) with
your generated images, as well as a file "log.html" that
will show you the images with the prompts that generated them.
[ Note addition of "comp.ai.shells" to newsgroups. ]
--
-v System76 Thelio Mega v1.1 x86_64 NVIDIA RTX 3090 Ti
OS: Linux 6.9.9 Release: Mint 21.3 Mem: 258G
"I tried to get a life once, but they were out of stock."
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)