Search Setup for Free

Software Engineering Courses (SECourses)•15mo ago

Furkan Gözükara SECourses

way lower than I expect

way lower than I expect

Software Engineering Courses (SECourses)Join

Hello everyone. I am Dr. Furkan Gözükara. PhD Computer Engineer. SECourses is a dedicated YouTube channel for the following topics : Tech, AI, News, Science, Robotics, Singularity, ComfyUI, SwarmUI, ML, Artificial Intelligence, Humanoid Robots, Wan 2.2, FLUX, Krea, Qwen Image, VLMs, Stable Diffusion

11,665Members

View on Discord

Resources

Recent Announcements

Similar Threads

Was this page helpful?

Product

Pricing Docs Communities

Resources

About Blog Changelog Contributors

Legal

Terms Privacy Cookies EULA

Twitter GitHub Discord

© 2026 Hedgehog Software, LLC. All rights reserved.

© 2026 Hedgehog Software, LLC

Twitter GitHub Discord

More

Communities Docs About Terms Privacy

Recent Announcements

Furkan Gözükara SECourses

@everyone Qwen Image 2511 is truly a massive upgrade compared to 2509. Full step by step how to use tutorial published and i have compared as well https://youtu.be/YfuQuOk2sB0 If you can upvote, leave a comment I appreciate that: https://www.reddit.com/r/comfyui/comments/1pv9snm/qwen_image_edit_2511_is_a_massive_upgrade/

Furkan Gözükara SECourses · 2d ago

Furkan Gözükara SECourses

@everyone After doing massive research our Wan 2.2 LoRA training tutorial is ready and published thankfully https://youtu.be/ocEkhAsPOs4 If you can upvote, leave a comment I appreciate that: https://www.reddit.com/r/comfyui/comments/1pscwvr/wan_22_complete_training_tutorial_text_to_image/

Furkan Gözükara SECourses · 5d ago

Furkan Gözükara SECourses

@everyone After doing massive research our Z Image Turbo LoRA training tutorial is ready and published thankfully https://youtu.be/ezD6QO14kRc If you can upvote, leave a comment I appreciate that: https://www.reddit.com/r/comfyui/comments/1pedr0v/zimage_turbo_lora_training_with_ostris_ai_toolkit/

Furkan Gözükara SECourses · 4w ago

Similar Threads

training nothing more fancy than things

SEC(Software Engineering Courses (SECourses) / qwen-flux-wan-such

hello everyone I need your help!!! I am

SEC(Software Engineering Courses (SECourses) / dreambooth-training

I need some help. I trained a flux LoRA

SEC(Software Engineering Courses (SECourses) / dreambooth-training

Interesting! I will experiment with

SEC(Software Engineering Courses (SECourses) / dreambooth-training

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 11:33 AM

maybe it was that pod

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 11:34 AM

i didnt test h100 speed

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 11:34 AM

you should test effective step speed

Furkan Gözükara SECourses

FFurkan Gözükara SECourses i didnt test h100 speed

ar.1

A

ar.1•10/20/24, 11:34 AM

ill test it later today and report back

ar.1

Aar.1 ye I just testing with something for now, but my biggest question is how many st...

Vinch

V

Vinch•10/20/24, 11:35 AM

I do fine-tuning with dreambooth batch-7, I'm currently testing with 2e-5 and 1e-5 (4e-5 was indeed too high, but better too high than too low), with 4000 steps, I also activated buckets for multiple aspect ratio's (where each bucket must have at least 7 images). The first 10% of steps are warmup and the LR decreases with a cosine. It's looking like 2e-5 is working the best for my major concept.

Vinch

VVinch I do fine-tuning with dreambooth batch-7, I'm currently testing with 2e-5 and 1e...

ar.1

A

ar.1•10/20/24, 11:39 AM

how many images total ?

ar.1

Aar.1 how many images total ?

Vinch

V

Vinch•10/20/24, 11:41 AM

I'm currently doing stage 1 that has 556 images. The images in stage 1 has high concept and low details, most of these images are AI generated to really get the concept. For stage 2, I have 2551 images, these images have less of the concept but are high in detail and quality, which are mostly production grade photographs and marketing material for movies and such.

Vinch

V

Vinch•10/20/24, 11:41 AM

I don't know what the config will be for stage 2.

Vinch

VVinch I do fine-tuning with dreambooth batch-7, I'm currently testing with 2e-5 and 1e...

ar.1

A

ar.1•10/20/24, 11:43 AM

how are you ensuring each bucket has at least 7 ?

Vinch

V

Vinch•10/20/24, 11:43 AM

I made a script. You want?

ar.1

A

ar.1•10/20/24, 11:43 AM

yeah sure if you dont mind sharing that would be great

Vinch

V

Vinch•10/20/24, 11:44 AM

The instructions are in the top of the file

fit_in_buckets.py12.22KB

Vinch

VVinch The instructions are in the top of the file

ar.1

A

ar.1•10/20/24, 11:46 AM

amazing thank you bro, ill use it on my next test

Vinch

VVinch I made a script. You want?

ar.1

A

ar.1•10/20/24, 11:47 AM

how did you come up with the idea to have min number of images per bucket ? did you notice it improves anything ?

ar.1

Aar.1 how did you come up with the idea to have min number of images per bucket ? did ...

Vinch

V

Vinch•10/20/24, 11:55 AM

It is because if you work with batches and you have buckets turned on. A batch will only contain images from a single bucket. So if you have buckets with 1 image. That means that it has to duplicate that single image 7 times to fill a batch, Making it basically ineffective.

Vinch

VVinch It is because if you work with **batches** and you have __buckets__ turned on. A...

ar.1

A

ar.1•10/20/24, 11:55 AM

makes sense

Furkan Gözükara SECourses

FFurkan Gözükara SECourses way lower than I expect

ar.1

A

ar.1•10/20/24, 1:44 PM

what speed where you getting ?

ar.1

A

ar.1•10/20/24, 1:44 PM

Ive got H100 going now, its caching latents

ar.1

Aar.1 what speed where you getting ?

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 1:55 PM

let me check video part

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 1:55 PM

i am still editing :d

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 1:55 PM

21.21 second

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 1:55 PM

batch size 7

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 1:56 PM

by the way if dataset cant be divided into 7 exactly, you wont get very accurate speed

Furkan Gözükara SECourses

FFurkan Gözükara SECourses 21.21 second

ar.1

A

ar.1•10/20/24, 2:03 PM

so H100, batch size 12 I getting 16s/it

ar.1

A

ar.1•10/20/24, 2:04 PM

65GB GPU memory usage

ar.1

Aar.1 so H100, batch size 12 I getting 16s/it

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:08 PM

wow nice

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:08 PM

almost 1 second / it per image

ar.1

A

ar.1•10/20/24, 2:09 PM

I can probably fit 14 batch size

Furkan Gözükara SECourses

FFurkan Gözükara SECourses almost 1 second / it per image

ar.1

A

ar.1•10/20/24, 2:12 PM

so with H100 at $3 per hour, 10k images with 15 epochs (batch size 12) is gonna cost around $170

ar.1

Aar.1 so with H100 at $3 per hour, 10k images with 15 epochs (batch size 12) is gonna ...

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:29 PM

looks like way expensive compared to A6000

Furkan Gözükara SECourses

FFurkan Gözükara SECourses looks like way expensive compared to A6000

ar.1

A

ar.1•10/20/24, 2:30 PM

yeah thats what im thinking

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:30 PM

10k images 15 epochs would take 53$

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:30 PM

0.31 cent per hour

ar.1

A

ar.1•10/20/24, 2:30 PM

much better

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:31 PM

yep

Furkan Gözükara SECourses

FFurkan Gözükara SECourses yep

ar.1

A

ar.1•10/20/24, 2:32 PM

we cant do multi gpu with A6000 dreambooth finetuning right?

ar.1

A

ar.1•10/20/24, 2:33 PM

I wish there was option to add more SSD in massed compute

ar.1

A

ar.1•10/20/24, 2:34 PM

256GB on the A6000 isnt that great

ar.1

Aar.1 we cant do multi gpu with A6000 dreambooth finetuning right?

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:34 PM

yes you need 80 gb gpus

ar.1

Aar.1 I wish there was option to add more SSD in massed compute

Vinch

V

Vinch•10/20/24, 2:34 PM

I keep having to upload checkpoints and delete them to make room for the next checkpoint

Vinch

VVinch I keep having to upload checkpoints and delete them to make room for the next ch...

ar.1

A

ar.1•10/20/24, 2:34 PM

yeah its a pain haha

ar.1

A

ar.1•10/20/24, 2:34 PM

Has anyone tried A100

ar.1

A

ar.1•10/20/24, 2:35 PM

I wonder its speed against the H100 as its almost half the price

Vinch

VVinch I keep having to upload checkpoints and delete them to make room for the next ch...

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:39 PM

you can delete existing apps and models

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:39 PM

shown in new coming video

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:39 PM

also yes i agree 256 GB disk space is very limiting for flux fine tuning :/

Furkan Gözükara SECourses

FFurkan Gözükara SECourses shown in new coming video

ar.1

A

ar.1•10/20/24, 2:40 PM

very eager to watch it, whats the ETA ?

dsienra

D

dsienra•10/20/24, 2:40 PM

@Dr. Furkan Gözükara this is cool!!!

ar.1

Aar.1 very eager to watch it, whats the ETA ?

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:40 PM

still editing

dsienra

Ddsienra @Dr. Furkan Gözükara this is cool!!!

Furkan Gözükara SECourses

F

Furkan Gözükara SECoursesOP•10/20/24, 2:40 PM

this is next hopefully after i published video to test