Software Engineering Courses (SECourses)•11mo ago

chatgpt made me this graph from your [raw testing data ](https://gist.githubusercontent.com/FurkanGo

chatgpt made me this graph from your raw testing data

AAiInfluence Swarm stopped showing me how the image is made up

kareem9000•2/14/25, 1:16 PM

same

Kriyus•2/14/25, 1:23 PM

no CLIP/text encoder weights in checkpoint, the text encoder model will not be loaded.
CLIP/text encoder model load device: cuda:0, offload device: cpu, current: cpu, dtype: torch.float16
clip missing: ['text_projection.weight']

Kriyus•2/14/25, 1:25 PM

Even though my Clip files are installed, I'm getting a 'clip missing' error. Has anyone experienced this before, or do you know the solution?

Kkareem9000 same

AiInfluence•2/14/25, 3:54 PM

got an answer on swarm server.

Go on the left side of swarm on the search bar "Filter parameters" and search for the keyword preview. There's a tik option on the Swarm Internal tab that reads "No previews" make sure to uncheck it

AAiInfluence @Dr. Furkan Gözükara New SOTA TTS 🚨🚨🚨 https://github.com/Zyphra/Zonos 🚨🚨🚨�...

Timson•2/14/25, 4:35 PM

sounds amazing, much better than F5-TTS

TTimson sounds amazing, much better than F5-TTS

AiInfluence•2/14/25, 4:41 PM

check your DMS pls

AiInfluence•2/14/25, 4:41 PM

yeah but poor language support

AAiInfluence yeah but poor language support

Timson•2/14/25, 4:55 PM

F5-TTS had only 2 languages, zonos have 5

TTimson F5-TTS had only 2 languages, zonos have 5

AiInfluence•2/14/25, 4:55 PM

F5 could be finetuned tho

Mido•2/14/25, 6:27 PM

whats the optimal amount of steps to have for flux lora training? i have 175 images and 180 reg images

Mido•2/14/25, 6:31 PM

btw training is very slow, i am on 48 vram, how can i make it faster?

kareem9000•2/14/25, 7:32 PM

Where is the unified download for swarmui again?

AAiInfluence got an answer on swarm server. Go on the left side of swarm on the search bar "...

kareem9000•2/14/25, 7:32 PM

will try ithank u

kareem9000•2/14/25, 8:05 PM

anyone else getting this on windows

kareem9000•2/14/25, 8:06 PM

im using my SSD and still have this issue with swarm ui

kareem9000•2/14/25, 8:06 PM

kareem9000•2/14/25, 8:07 PM

@Furkan Gözükara SECourses this is using the unified downloader, i ran sawrm ui with launch-windows.bat

kareem9000•2/14/25, 8:07 PM

right before it crashes i see this

Kkareem9000 right before it crashes i see this

nicearoni•2/14/25, 8:21 PM

do you get the same error if you try a different model?

AAiInfluence @Dr. Furkan Gözükara New SOTA TTS 🚨🚨🚨 https://github.com/Zyphra/Zonos 🚨🚨🚨�...

Furkan Gözükara SECourses•2/14/25, 8:34 PM

ye i am following on my list

NNeuromasters Why are there streaks when generating on flux? or is that from the upscaler? How...

Furkan Gözükara SECourses•2/14/25, 8:34 PM

this happens when using lora but fine tuning fixes

MMido Hello Dr. Furkan, I tired literally everything, reinstalled everything, made a w...

Furkan Gözükara SECourses•2/14/25, 8:35 PM

if you are gold member i can connect and check your setup

Ttazztone chatgpt made me this graph from your [raw testing data ](https://gist.githubuser...

Furkan Gözükara SECourses•2/14/25, 8:35 PM

nice

MMido whats the optimal amount of steps to have for flux lora training? i have 175 ima...

Furkan Gözükara SECourses•2/14/25, 8:39 PM

getting faster gpu

Furkan Gözükara SECourses•2/14/25, 8:39 PM

and bigger batch size

Kkareem9000 Where is the unified download for swarmui again?

Furkan Gözükara SECourses•2/14/25, 8:40 PM

https://www.patreon.com/posts/114517862

Patreon

SwarmUI Easy Ultra Fast and Robust Unified Downloader for Stable Di...

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

Kkareem9000 anyone else getting this on windows

Furkan Gözükara SECourses•2/14/25, 8:40 PM

16 gb ram your issue

FFurkan Gözükara SECourses 16 gb ram your issue

kareem9000•2/14/25, 8:41 PM

Yeah i noticed

kareem9000•2/14/25, 8:41 PM

i guess i should just upgrade RAM inside my PC

Kkareem9000 i guess i should just upgrade RAM inside my PC

Furkan Gözükara SECourses•2/14/25, 8:42 PM

100%

Furkan Gözükara SECourses•2/14/25, 8:42 PM

go to 2x48 gb if your mobo supports

Furkan Gözükara SECourses•2/14/25, 8:42 PM

or 2x32

FFurkan Gözükara SECourses go to 2x48 gb if your mobo supports

kareem9000•2/14/25, 8:44 PM

2 ram sticks u mena?

kareem9000•2/14/25, 8:44 PM

mean*

Kkareem9000 2 ram sticks u mena?

Furkan Gözükara SECourses•2/14/25, 8:46 PM

yes

KKriyus no CLIP/text encoder weights in checkpoint, the text encoder model will not be l...

Kriyus•2/14/25, 8:47 PM

@Furkan Gözükara SECourses

KKriyus @Furkan Gözükara SECourses

Furkan Gözükara SECourses•2/14/25, 8:51 PM

i dont know from memory i use swarmui

CCharlie Blau Do you know why using a LoRA can cause this drop in quality? Generated with Comf...

J2_Nedo•2/14/25, 10:16 PM

i do think Timson have the right answer, i had similar results with overtrained Loras, but in the end, it is mostly the resolution.

Mido•2/14/25, 10:50 PM

trained a LoRA for flux exactly like the tutorial (for human images), the results are nothing like the trained pictures, and its not even a consistent character, just keeps pulling out new character as if i am not using any lora

Mido•2/14/25, 10:50 PM

and even tho i used hundreds of images, the size of each of the lora files was only 18MB