well this was result of a flux kontext face swap workflow

DDr. Odin and when it comes to the lora strength... what do you recomend.. assuming i did ...

Furkan Gözükara SECourses•8/25/25, 4:00 PM

i use 1

�🍭🎀 𝒜𝓋𝒶 𝐹𝓇𝒾𝑔𝑔 🎀🍭well this was result of a flux kontext face swap workflow

Timson•8/25/25, 9:07 PM

This is a feature of flux kontext, no way to avoid it completly, only post-process fix / gen over

Timson•8/25/25, 9:08 PM

I was never able to reploduce amazing results using flux kontext workflow. Everything it generates looks like flux schnell - old generation / midjourney / sdxl quality

Timson•8/25/25, 9:08 PM

It excels at not touching parts of the image it doesn;'t need to touch, but for parts that is does - it woudl go full schnell bs

Timson•8/25/25, 9:10 PM

Maybe it's better with FLUX.1 Kontext Pro version

Jimi•8/28/25, 4:55 PM

for 32gb vram should i use the 24gb one?

JJimi for 32gb vram should i use the 24gb one?

JonkoXL•8/28/25, 5:19 PM

what is this?

Jimi•8/28/25, 5:20 PM

for DreamBooth configration

JonkoXL•8/28/25, 5:24 PM

probably

JonkoXL•8/28/25, 5:24 PM

it seesm there is an explanation text file in there too, have you read it?

JJimi for 32gb vram should i use the 24gb one?

Furkan Gözükara SECourses•8/28/25, 6:00 PM

no use 48 gb config

Furkan Gözükara SECourses•8/28/25, 6:01 PM

it uses 28200 mb so fits into 32 gb

Jimi•8/28/25, 6:49 PM

THanks its working

JJimi THanks its working

Furkan Gözükara SECourses•8/28/25, 6:52 PM

you are welcome

Jimi•8/29/25, 6:40 AM

how continue training from that checkpoint ? just load the last save checkpoint or do i have to change any other option from gui?

Jimi•8/29/25, 6:51 AM

@Dr. Furkan Gözükara

JJimi how continue training from that checkpoint ? just load the last save checkpoint ...

Furkan Gözükara SECourses•8/29/25, 9:07 AM

yes just load last checkpoint instead of base model

Furkan Gözükara SECourses•8/29/25, 9:07 AM

and reduce number of epochs

Furkan Gözükara SECourses•8/29/25, 9:07 AM

like 100 completed then 100 left = 200

Jimi•8/29/25, 9:17 AM

Thanks you

lerois1•8/30/25, 2:34 AM

Hi, today flux krea is better than flux d to dreambooth training? qwen is gonna be better or worse?

Llerois1 Hi, today flux krea is better than flux d to dreambooth training? qwen is gonna ...

Furkan Gözükara SECourses•8/30/25, 8:41 AM

i compared krea

Furkan Gözükara SECourses•8/30/25, 8:41 AM

you can see here results

Furkan Gözükara SECourses•8/30/25, 8:41 AM

https://www.patreon.com/posts/kohya-flux-fine-112099700

Patreon

Kohya FLUX Fine Tuning (Full Checkpoints) Training Full Tutorial Fo...

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

Furkan Gözükara SECourses•8/30/25, 8:42 AM

read 13 August 2025 Update

ZiriussBlack•8/31/25, 7:28 PM

hi for qwen train with musubi i need change any config for 12gb Vram? i have 3060 rtx, any tutorial or tips?

ZZiriussBlack hi for qwen train with musubi i need change any config for 12gb Vram? i have 306...

Furkan Gözükara SECourses•8/31/25, 7:35 PM

yes . it wont work for now. wait me completing parameters

Furkan Gözükara SECourses•8/31/25, 7:36 PM

i mean it works but you cant randomly train with tht card

Jimi•9/1/25, 11:13 AM

what's the sdxl epoch for 50 image dreambooth training?

�

🍭🎀 𝒜𝓋𝒶 𝐹𝓇𝒾𝑔𝑔 🎀🍭OP•9/1/25, 11:28 AM

its not a fixed epoch, go up to 160 and compare i.e. save every 20 epoch

JonkoXL•9/1/25, 11:36 AM

JJimi what's the sdxl epoch for 50 image dreambooth training?

Furkan Gözükara SECourses•9/1/25, 11:40 AM

ye go up to 160 even 200

Furkan Gözükara SECourses•9/1/25, 11:40 AM

and compare

Jimi•9/1/25, 11:41 AM

Wow that's 1 epoch take like 1 hour 37 min locally

Jimi•9/1/25, 11:41 AM

So it will take how many days

Jimi•9/1/25, 11:42 AM

I used your 24gb config

Jimi•9/1/25, 11:42 AM

With 40 repeat

JJimi With 40 repeat

Furkan Gözükara SECourses•9/1/25, 11:56 AM

40 repeat how many epoch?

Furkan Gözükara SECourses•9/1/25, 11:56 AM

40 repeat means you need to do 5 epoch

Furkan Gözükara SECourses•9/1/25, 11:56 AM

= 200 epochs

Jimi•9/1/25, 11:56 AM

Ya I understand now

Jimi•9/1/25, 11:57 AM

Thanks

JonkoXL•9/1/25, 12:02 PM

how many images are you using

JonkoXL•9/1/25, 12:02 PM

if you keep epoch + repeats the same it will take a longer time with more images

�

🍭🎀 𝒜𝓋𝒶 𝐹𝓇𝒾𝑔𝑔 🎀🍭OP•9/1/25, 1:05 PM

Error on newest image processor app, ran on massedcompute. image resize gave this error:
SUCCESSFULLY SAVED: /home/Ubuntu/Downloads/crop/1024x1024/1637004-candid shot taken with iPhone 15 of ohwx-modelcolomb_krea_1-000120-80.png
Resizing images: 0%| | 0/19 [00:00<?, ?image/s]Traceback (most recent call last):
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/gradio/queueing.py", line 667, in process_events
response = await route_utils.call_process_api(
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/gradio/route_utils.py", line 349, in call_process_api
output = await app.get_blocks().process_api(
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/gradio/blocks.py", line 2274, in process_api
result = await self.call_function(
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/gradio/blocks.py", line 1793, in call_function
prediction = await utils.async_iteration(iterator)
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/gradio/utils.py", line 760, in async_iteration
return await anext(iterator)
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/gradio/utils.py", line 751, in anext
return await anyio.to_thread.run_sync(
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/anyio/to_thread.py", line 56, in run_sync
return await get_async_backend().run_sync_in_worker_thread(
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/anyio/_backends/_asyncio.py", line 2476, in run_sync_in_worker_thread
return await future
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/anyio/_backends/_asyncio.py", line 967, in run
result = context.run(func, *args)
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/gradio/utils.py", line 734, in run_sync_iterator_async
return next(iterator)
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/venv/lib/python3.10/site-packages/gradio/utils.py", line 898, in gen_wrapper
response = next(iterator)
File "/home/Ubuntu/Downloads/Ulti_Img_Process_V22/app.py", line 909, in resize_images
if processing_state.stop_flag is not None and processing_state.stop_flag.value:
File "/home/Ubuntu/.pyenv/versions/3.10.16/lib/python3.10/multiprocessing/managers.py", line 1137, in get
return self._callmethod('get')
File "/home/Ubuntu/.pyenv/versions/3.10.16/lib/python3.10/multiprocessing/managers.py", line 817, in _callmethod
conn.send((self._id, methodname, args, kwds))
File "/home/Ubuntu/.pyenv/versions/3.10.16/lib/python3.10/multiprocessing/connection.py", line 206, in send
self._send_bytes(_ForkingPickler.dumps(obj))
File "/home/Ubuntu/.pyenv/versions/3.10.16/lib/python3.10/multiprocessing/connection.py", line 411, in _send_bytes
self._send(header + buf)
File "/home/Ubuntu/.pyenv/versions/3.10.16/lib/python3.10/multiprocessing/connection.py", line 368, in _send
n = write(self._handle, buf)
BrokenPipeError: [Errno 32] Broken pipe

�

🍭🎀 𝒜𝓋𝒶 𝐹𝓇𝒾𝑔𝑔 🎀🍭OP•9/1/25, 1:16 PM

interesting, it did complete the resize process, even if it gave the error

�🍭🎀 𝒜𝓋𝒶 𝐹𝓇𝒾𝑔𝑔 🎀🍭Error on newest image processor app, ran on massedcompute. image resize gave thi...

Furkan Gözükara SECourses•9/1/25, 8:23 PM

i feel like machine error

Furkan Gözükara SECourses•9/1/25, 8:23 PM

did you try again on another machine?

Furkan Gözükara SECourses•9/1/25, 8:24 PM

ah i see multi threading issue

Furkan Gözükara SECourses•9/1/25, 8:24 PM

how many thrads did you set?

well this was result of a flux kontext face swap workflow

Similar Threads

Similar Threads

Similar Threads