Software Engineering Courses (SECourses)•3y ago

Thank you so much for answering again! I’d be happy to test 🙂 I take it it’s in your Patreon?

Thank you so much for answering again!
I’d be happy to test

I take it it’s in your Patreon?

AIGambino•12/8/23, 10:19 PM

Has anyone gotten animatediff working with sdxl?

AIGambino•12/8/23, 10:34 PM

https://www.youtube.com/watch?v=EC9bE7Fot78 is this possible with my custom models I've made in SDXL?

Furkan Gözükara SECourses•12/8/23, 10:55 PM

code former added : https://www.patreon.com/posts/94098751

Furkan Gözükara SECourses•12/8/23, 10:55 PM

EEyeSpyBekoAI Thank you so much for answering again! I’d be happy to test 🙂 I take it it’s in...

Furkan Gözükara SECourses•12/8/23, 10:55 PM

yes added

AAIGambino https://www.youtube.com/watch?v=EC9bE7Fot78 is this possible with my custom mode...

Furkan Gözükara SECourses•12/8/23, 10:56 PM

i dont know atm

SSandstormleader hi did we ever get an up to date automatic1111 with working dreambooth for pc lo...

Furkan Gözükara SECourses•12/8/23, 10:56 PM

yes i have installer

Furkan Gözükara SECourses•12/8/23, 10:56 PM

installs working verison

Furkan Gözükara SECourses•12/8/23, 10:56 PM

but older one

virginia•12/8/23, 11:58 PM

@StarBuffer it is very hard to get exact images of model you trained from this shoe, in case of lora. expecially when your shoe has many stylized detials. checkpoint may be better. Otherwise you might need 50 images for training. sd 1.5 training is very bad. I suggest you not to use.

Pelolisu•12/9/23, 12:06 AM

@Dr. Furkan Gözükara my model trained on realistic was unable to get my face at all checkpoints. I didn't use Lion but 8bit Adam with the settings for 12GB Vram.

Pelolisu•12/9/23, 12:09 AM

images like this, and duplicated, came close but it's not me.

PPelolisu @Dr. Furkan Gözükara my model trained on realistic was unable to get my face at ...

Furkan Gözükara SECourses•12/9/23, 12:27 AM

you can also try our Kaggle SDXL DreamBooth

Furkan Gözükara SECourses•12/9/23, 12:27 AM

it will be probably better

Pelolisu•12/9/23, 12:35 AM

ok i will but i have a RTX 3060 Ti 8Gb + 16GB shared virtual

Pelolisu•12/9/23, 12:35 AM

Kaggle is better graphic card?

PPelolisu Kaggle is better graphic card?

Furkan Gözükara SECourses•12/9/23, 1:00 AM

yep

Furkan Gözükara SECourses•12/9/23, 1:00 AM

it is 15 gb

Furkan Gözükara SECourses•12/9/23, 1:00 AM

shared vram will be ultra slow

Pelolisu•12/9/23, 1:02 AM

i found that the problem of twins is often

Pelolisu•12/9/23, 1:02 AM

in realist vision

Pelolisu•12/9/23, 1:03 AM

because the better resolution is 512X768 but i have used 1024x768

Pelolisu•12/9/23, 1:07 AM

i used 5.1 but i have 2.0 too

Pelolisu•12/9/23, 1:52 AM

with 8bit adam and xformers my training took 2:55hs now with lion and without xformers it will take almost 7 hours with 20 images in the dataset i will try kaggle

Pelolisu•12/9/23, 1:55 AM

but I will also test how it performs on 8bit adam locally at 512x768 and then using a 4x upscaler to generate the images.

PPelolisu after uploading the image, the video and the pose are expected to appear, correc...

thewingag•12/9/23, 2:44 AM

did u install Magic in comfyui? sr, last night, i go to sleep? 1AM in my country

Pelolisu•12/9/23, 2:45 AM

Yes, I put it to run but it doesn't show any progress. It was in the attention message, can you tell if it takes a long time for the result to come?

Pelolisu•12/9/23, 2:46 AM

I had to look for an example openpose video to use with my image, but the result was black, I think it takes a long time to process, but it also didn't give any error messages.

Pelolisu•12/9/23, 2:49 AM

I need a comfyUI tutorial with Magic Animate to see if I didn't do anything wrong.

thewingag•12/9/23, 2:53 AM

ok.

Pflanzen die Tanzen•12/9/23, 4:22 AM

A short story for my niece and nephews:

Tools used:

Stable diffusion
GPT-4

The main goals were these:

give my niece a nice present
create a coherent story
create coherent characters

All in all a rather straight forward process, just putting together several techniques I've learned in the past year.

FiveSAI•12/9/23, 4:35 AM

Even Faster SDXL Dreambooth Training Full BF16 Enabled

FiveSAI•12/9/23, 4:36 AM

Still further testing but results so far amazing

Pelolisu•12/9/23, 5:03 AM

I'm crying with 8gb vram haha

Someone.in.144p•12/9/23, 5:24 AM

same heh

FFurkan Gözükara SECourses you need to give accurate folder path and it will work

Someone.in.144p•12/9/23, 5:25 AM

it worked :DDDD, thank you so much for the help man, u savd my school project

Someone.in.144p•12/9/23, 5:25 AM

PPelolisu Click to see attachment

thewingag•12/9/23, 8:17 AM

hmm, about 100Gb lol :v

virginia•12/9/23, 11:04 AM

@Pelolisu I usually get exact dozens of images from hundreds of or even thousands of created images on the basis of trained lora, Furkan said dreambooth checkpoint is better than lora. I have not tried yet. If a object is too detailed and styled, lora results are not good, time wasting, many images are not to the standards you require.