Software Engineering Courses (SECourses)•16mo ago

btw this link in the kaggle notebook is not working https://huggingface.co/Kijai/flux-fp8/resolve/ma

btw this link in the kaggle notebook is not working https://huggingface.co/Kijai/flux-fp8/resolve/main/flux1-dev-fp8.safetensors so I'm downloading "flux1-dev-fp8-e4m3fn.safetensors" and " flux1-dev-fp8-e5m2.safetensors" separately from here https://huggingface.co/Kijai/flux-fp8/tree/main

Mmagical btw this link in the kaggle notebook is not working https://huggingface.co/Kijai...

Neuromasters•9/8/24, 9:13 AM

I tested it, 0 difference and moved to fp8

Neuromasters•9/8/24, 9:13 AM

Errors fail due to memory on the regular version of Dev

Neuromasters•9/8/24, 9:14 AM

I'm just training the lora on he

NNeuromasters I tested it, 0 difference and moved to fp8

magicalOP•9/8/24, 9:14 AM

oh you mean for lora training?

Mmagical oh you mean for lora training?

Neuromasters•9/8/24, 9:15 AM

Yes, on instructions @Dr. Furkan Gözükara

Dreamer@TDFilmStudio•9/8/24, 9:27 AM

I have trained 3 custom Flux loras till now and everytime, I am getting scan lines in images. I use Rank 3 config (slow) with Kohya. Anyone else facing this issue? Not sure what I am doing wrong. I updated to latest Kohya version yesterday, but problem persists. Keeping Lora weight 0.8-1

DDreamer@TDFilmStudio I have trained 3 custom Flux loras till now and everytime, I am getting scan lin...

druhl•9/8/24, 9:29 AM

You mean those vertical lines? Can it be an issue with something you're using to process the initial dataset?

druhl•9/8/24, 9:29 AM

Even the bottom edge of the generation doesn't look clean, which makes me wonder.

Dreamer@TDFilmStudio•9/8/24, 9:30 AM

I have used 2 different datasets for custom Lora training till now. Each of 20 images

Dreamer@TDFilmStudio•9/8/24, 9:30 AM

Both are producing such lines

Dreamer@TDFilmStudio•9/8/24, 9:31 AM

Even when I use different versions of same Lora saved at various epochs, getting scan lines
Not sure if this is because of overtraining. All images are 1024X1024

Dreamer@TDFilmStudio•9/8/24, 9:33 AM

here another example. Faint vertical lines are visible if you zoom in

Maxivi•9/8/24, 9:38 AM

@Dr. Furkan Gözükara https://huggingface.co/OpenGVLab/InternVL2-8B/tree/main

this is the best captioner i found for flux, tested with 20 images, got my best results with them

OpenGVLab/InternVL2-8B at main

Maxivi•9/8/24, 9:45 AM

26B also nice, think could work on 4bit

MMaxivi @Dr. Furkan Gözükara https://huggingface.co/OpenGVLab/InternVL2-8B/tree/main th...

Dreamer@TDFilmStudio•9/8/24, 9:54 AM

Thhis is the top ranked in VLM models benchmarks

DDreamer@TDFilmStudio Thhis is the top ranked in VLM models benchmarks

Maxivi•9/8/24, 9:55 AM

Yep even the 8b model tops many of the other ones

MMaxivi @Dr. Furkan Gözükara https://huggingface.co/OpenGVLab/InternVL2-8B/tree/main th...

Dreamer@TDFilmStudio•9/8/24, 9:55 AM

Did you use any app or tool to load this model?

DDreamer@TDFilmStudio Did you use any app or tool to load this model?

Maxivi•9/8/24, 9:55 AM

i manually used the huggingface space

Dreamer@TDFilmStudio•9/8/24, 9:56 AM

nice, hope @Dr. Furkan Gözükara can create a set up like Cogvlm 2

druhl•9/8/24, 10:21 AM

Why does my terminated instance keep rising back up on Massed Compute?

druhl•9/8/24, 10:39 AM

Guess it finally died this time lol.

MC•9/8/24, 11:08 AM

Hey any advices for training a Flux style Lora with 100 images? For now I have been using Ostris Colab.

MMC Hey any advices for training a Flux style Lora with 100 images? For now I have ...

Maxivi•9/8/24, 11:28 AM

use onetrainer now

Maxivi•9/8/24, 11:28 AM

its fixed and the loras work on forge and comfy

MC•9/8/24, 11:34 AM

Any rec. Settings?

truellusion•9/8/24, 2:58 PM

I can't find a solution how to generate a queue of 100 different promts in order. The context window allows you to enter a lot of text with promts where they are placed in one line, but only the last one is launched

Ttruellusion I can't find a solution how to generate a queue of 100 different promts in order...

druhl•9/8/24, 3:28 PM

What are you using? SwarmUI generator?

druhl•9/8/24, 3:29 PM

Use seperator || to separate the prompts.

Dreamer@TDFilmStudio•9/8/24, 3:30 PM

@Dr. Furkan Gözükara Locking horns with GPT-4V https://huggingface.co/OpenGVLab/InternVL-Chat-V1-5

OpenGVLab/InternVL-Chat-V1-5 · Hugging Face

Ddruhl So, I just terminate the massed compute instance and it all goes away correct? �...

Furkan Gözükara SECourses•9/8/24, 3:41 PM

yes . you can use hugging face for backup which i use

Xxlads It’s not showing up for me

Furkan Gözükara SECourses•9/8/24, 3:41 PM

you need to generate if folder missing

AAIGambino whats the prompt here? lol love it

Furkan Gözükara SECourses•9/8/24, 3:43 PM

shared the prompts here , will put into our zip file as well : https://gist.github.com/FurkanGozukara/12c837939ca3fc59abc04d8bd44876ab

Gist

100 flux prompts.txt

100 flux prompts.txt. GitHub Gist: instantly share code, notes, and snippets.

Pprotector131090 Hello! whats the news on Flux Dreambooth training in Kohya on 24 vram ? possible...

Furkan Gözükara SECourses•9/8/24, 3:43 PM

i think yes and hopefully gonna start tests today

DDreamer@TDFilmStudio I have trained 3 custom Flux loras till now and everytime, I am getting scan lin...

Furkan Gözükara SECourses•9/8/24, 3:44 PM

did you notice that in my grids?

Furkan Gözükara SECourses•9/8/24, 3:44 PM

i saw such thing when only i used 2x upscale of swarmui

MMaxivi @Dr. Furkan Gözükara https://huggingface.co/OpenGVLab/InternVL2-8B/tree/main th...

Furkan Gözükara SECourses•9/8/24, 3:45 PM

do you use system prompt? if so what is it

FFurkan Gözükara SECourses do you use system prompt? if so what is it

Maxivi•9/8/24, 4:10 PM

just caption the image with details, colors, makeup, items, objects, emotions, art style, drawing style and objects but do not add any description or comment. do not miss any item in the given image, very simple sentance.

Maxivi•9/8/24, 4:10 PM

they have a demo interface

Maxivi•9/8/24, 4:10 PM

https://internvl.readthedocs.io/en/latest/get_started/local_chat_demo.html

Maxivi•9/8/24, 4:12 PM

@Dr. Furkan Gözükara its quite nice actually, able to capture details better as well. i've tried longer prompts too works quite nice

Maxivi•9/8/24, 4:12 PM

i think it would be better than joy and cog for style training too

FFurkan Gözükara SECourses you need to generate if folder missing