Software Engineering Courses (SECourses)•10mo ago

The new style of photos do have a slightly more iphone feel to them, but still a bit too blurry back

The new style of photos do have a slightly more iphone feel to them, but still a bit too blurry background, not fantastic results

1208-jqxz20style20a20man20sitting20at20a20medium20si-modeliphonestyle1-000080-1869224826.png

1208-jqxz20style20a20man20sitting20at20a20medium20si-modeliphonestyle1-000080-1705101624.png

1209-jqxz20style20a20man20sitting20at20a20medium20si-modeliphonestyle1-000080-1882239269.png

Jjulius i finished fine tuning on the iphone style now

Furkan Gözükara SECourses•3/8/25, 12:11 PM

looks like worked decent

Jjulius The new style of photos do have a slightly more iphone feel to them, but still a...

Furkan Gözükara SECourses•3/8/25, 12:11 PM

increase your dataset size

juliusOP•3/8/25, 12:11 PM

How many do you think would be ideal? (Currently 170)

Jjulius How many do you think would be ideal? (Currently 170)

Furkan Gözükara SECourses•3/8/25, 12:12 PM

i would say try to increase 340 and see how it goes

Furkan Gözükara SECourses•3/8/25, 12:12 PM

as you improve dataset next training should be better

FFurkan Gözükara SECourses i would say try to increase 340 and see how it goes

juliusOP•3/8/25, 12:13 PM

okay will try! Took a long time just to gather the 170 though manually picking out from peoples instagram, should perhaps find a faster approach

Furkan Gözükara SECourses•3/8/25, 12:14 PM

image to audio no prompt really good

Jjulius okay will try! Took a long time just to gather the 170 though manually picking o...

Furkan Gözükara SECourses•3/8/25, 12:15 PM

yes it is a tedious task

Furkan Gözükara SECourses•3/8/25, 12:15 PM

aim maximum quality

juliusOP•3/8/25, 12:15 PM

Would you include people? Or avoid them? Currently 80% of photos had people. I only need the model to generate portraits photos in the end

FFurkan Gözükara SECourses image to audio no prompt really good

juliusOP•3/8/25, 12:16 PM

wtf thats so sick

Jjulius Would you include people? Or avoid them? Currently 80% of photos had people. I o...

juliusOP•3/8/25, 12:17 PM

And would you include girls, even though I only want to use the model to fine-tune again on photos of guys?

Jjulius Would you include people? Or avoid them? Currently 80% of photos had people. I o...

Furkan Gözükara SECourses•3/8/25, 12:26 PM

you can include

Furkan Gözükara SECourses•3/8/25, 12:26 PM

but dont include same person

Jjulius wtf thats so sick

Furkan Gözükara SECourses•3/8/25, 12:26 PM

yep

Jjulius And would you include girls, even though I only want to use the model to fine-tu...

Furkan Gözükara SECourses•3/8/25, 12:27 PM

not necessarily needed to add girls

Jjulius i finished fine tuning on the iphone style now

masood•3/8/25, 12:47 PM

how do you guys do that? first you take a base model and finetune it on iphone style images and then you take that same model and finetune it on specific person. Right? but wont the model forget the style it learnt from that iphone photos ?

FFurkan Gözükara SECourses image to audio no prompt really good

masood•3/8/25, 12:47 PM

what model is that?

masood•3/8/25, 12:48 PM

and can it do video to audio?

JJ2_Nedo now thats news! if you get the installer done i will asap test this!

Furkan Gözükara SECourses•3/8/25, 12:48 PM

you can test few more things i will fix in few hours

Furkan Gözükara SECourses•3/8/25, 12:48 PM

https://www.patreon.com/posts/117990364

Patreon

MMAudio 1-Click Installers for Windows, RunPod and Massed Compute -...

Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon

Mmasood what model is that?

Furkan Gözükara SECourses•3/8/25, 12:48 PM

https://www.patreon.com/posts/117990364

Patreon

MMAudio 1-Click Installers for Windows, RunPod and Massed Compute -...

Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon

Furkan Gözükara SECourses•3/8/25, 1:40 PM

adding step speed

blackhat•3/8/25, 1:47 PM

for training the data set for the person lora , what have you focused on in the image discription (prompts) you provided for training person lora on flux ?
where could i find this , what should i focus to make person lora work well with emitions and the persons accuracy -> realism @Furkan Gözükara SECourses

Bblackhat for training the data set for the person lora , what have you focused on in the ...

Furkan Gözükara SECourses•3/8/25, 1:48 PM

just ohwx man

Furkan Gözükara SECourses•3/8/25, 1:48 PM

it works best

Furkan Gözükara SECourses•3/8/25, 1:48 PM

also we have more proper example dataset

Furkan Gözükara SECourses•3/8/25, 1:48 PM

https://www.patreon.com/posts/114972274

Patreon

Perfect Quality Example Model Training Images Dataset - Can Be Used...

Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon

blackhat•3/8/25, 2:02 PM

@Dr. Furkan Gözükara doesn't captioning the dataset of the person flow make any good sense for ohwx man person lora trainig on the data set , i am not gettting it somewhere in the filex and the blog you shared

Bblackhat @Dr. Furkan Gözükara doesn't captioning the dataset of the person flow make any ...

Furkan Gözükara SECourses•3/8/25, 2:04 PM

captioning reduces likeliness

Furkan Gözükara SECourses•3/8/25, 2:04 PM

your dataset is not good in that case :d

Furkan Gözükara SECourses•3/8/25, 2:04 PM

captioni,ng only helps when your dataset is not consistent

blackhat•3/8/25, 2:06 PM

consistent in what sense , like if we have a data set of a person with all the things that you mentioned in the blog about the ohmx man training with diff shots and diff clothes at diff places also considered consistent or not @Dr. Furkan Gözükara

Bblackhat consistent in what sense , like if we have a data set of a person with all the t...

Furkan Gözükara SECourses•3/8/25, 2:10 PM

it is consistent as long as it is same person

Furkan Gözükara SECourses•3/8/25, 2:10 PM

and no other persons exists in dataset

FFurkan Gözükara SECourses it is consistent as long as it is same person

blackhat•3/8/25, 2:11 PM

what about he is having beard and diff hairstyle is that inconsitent , if yes then what to take care in captioning so we are not over eng things in the captioning for the lora traning

blackhat•3/8/25, 2:17 PM

@Dr. Furkan Gözükara

Bblackhat what about he is having beard and diff hairstyle is that inconsitent , if yes th...

Furkan Gözükara SECourses•3/8/25, 2:17 PM

i think still not needed

Furkan Gözükara SECourses•3/8/25, 2:17 PM

mmaudio almost ready

Furkan Gözükara SECourses•3/8/25, 2:17 PM

FFurkan Gözükara SECourses nice. if you have a 1 for all negative prompt we can update our default

The90sPope•3/8/25, 2:21 PM

This is my latest NEGATIVE Promt. It is meant for jus a regular realistic video. Wan tends to shift video styles depending on your prompt, lets say you want aliesn in your video, all of a sudden it starts making 3D render or animation instead of realistc video. This prompt will lock you in realistic style and produce less shitty videos by a good margin than |no negative prompt at all. Dont waste your time. Take it ant tweak to your needs:

Overexposure, static, blurred details, subtitles, paintings, pictures, still, overall gray, worst quality, low quality, JPEG compression residue, ugly, mutilated, redundant fingers, poorly painted hands, poorly painted faces, deformed, disfigured, deformed limbs, fused fingers, cluttered background, three legs, a lot of people in the background, upside down, inaccurate, distorted, motion smear, motion blur, motion stutter, video noise, video artifacts, warped, motionless, 3D render, animation, anime, 3D game, cartoon, comic book, manga, stylized, abstract