Software Engineering Courses (SECourses)•3y ago

Hi, I'm banging my head against the wall trying to install Dreambooth to a RunPod. Wondering if anyo

Hi, I'm banging my head against the wall trying to install Dreambooth to a RunPod.
Wondering if anyone else is running into a problem with today's Dreambooth (I see there was an update today: https://github.com/d8ahazard/sd_dreambooth_extension

I've been following your "Ultimate RunPod" tutorial @Furkan Gözükara SECourses to the letter.

So, I've attached an image of the Pod I've chosen.

Next is the container logs:

NateOP•3/26/23, 6:32 AM

2023-03-26T06:28:43.440593662Z Launching Web UI with arguments: --port 3000 --xformers --listen --enable-insecure-extension-access

NateOP•3/26/23, 6:33 AM

2023-03-26T06:28:50.339983137Z Calculating sha256 for /workspace/stable-diffusion-webui/models/Stable-diffusion/SDv1-5.ckpt: 4c86efd062ea772972dd068c3870773f1fd9a37ba9a288ee79ec848448ce7785
2023-03-26T06:28:50.340315707Z Loading weights [4c86efd062] from /workspace/stable-diffusion-webui/models/Stable-diffusion/SDv1-5.ckpt
2023-03-26T06:28:51.295544766Z Creating model from config: /workspace/stable-diffusion-webui/configs/v1-inference.yaml

NateOP•3/26/23, 6:34 AM

2023-03-26T06:28:35.243923440Z Installing xformers

NateOP•3/26/23, 6:35 AM

2023-03-26T06:28:40.362709032Z Installing requirements for Web UI
2023-03-26T06:28:43.440593662Z Launching Web UI with arguments: --port 3000 --xformers --listen --enable-insecure-extension-access
2023-03-26T06:28:50.339983137Z Calculating sha256 for /workspace/stable-diffusion-webui/models/Stable-diffusion/SDv1-5.ckpt: 4c86efd062ea772972dd068c3870773f1fd9a37ba9a288ee79ec848448ce7785
2023-03-26T06:28:50.340315707Z Loading weights [4c86efd062] from /workspace/stable-diffusion-webui/models/Stable-diffusion/SDv1-5.ckpt
2023-03-26T06:28:51.295544766Z Creating model from config: /workspace/stable-diffusion-webui/configs/v1-inference.yaml
2023-03-26T06:28:51.296084472Z LatentDiffusion: Running in eps-prediction mode
2023-03-26T06:28:51.436024406Z DiffusionWrapper has 859.52 M params.
2023-03-26T06:28:54.256998580Z Applying xformers cross attention optimization.
2023-03-26T06:28:54.264969229Z Textual inversion embeddings loaded(0):
2023-03-26T06:28:54.268481070Z Model loaded in 5.5s (calculate hash: 1.5s, load weights from disk: 0.9s, create model: 0.5s, apply weights to model: 0.6s, apply half(): 0.8s, load VAE: 0.9s, move model to device: 0.3s).
2023-03-26T06:28:54.679792888Z Running on local URL: http://0.0.0.0:3000
2023-03-26T06:28:54.679813069Z
2023-03-26T06:28:54.679818209Z To create a public link, set

share=True

share=True

launch()

launch()

.
2023-03-26T06:29:07.508571846Z [I 2023-03-26 06:29:07.508 ServerApp] 302 GET /?token=[secret] (@100.64.0.21) 0.79ms
2023-03-26T06:29:07.917108303Z [I 2023-03-26 06:29:07.916 LabApp] Generating new user for token-authenticated request: dc6dff40f6eb481786e80e3335d6778a
2023-03-26T06:29:19.115777894Z [W 2023-03-26 06:29:19.115 LabApp] Could not determine jupyterlab build status without nodejs

NateOP•3/26/23, 6:36 AM

So, there are differences here in the logs compared to the tutorial, the SD model here is SDv1-5, compared to the pruned model in the tutorial. And...
Xformers - there is no indication to which version.
..not sure if any of that matters! But just noting.

NateOP•3/26/23, 6:40 AM

My webui-user.sh also shows differences. The COMMANDLINE_ARGS do not specify my model, no reference to a *.ckpt
In the tutorial the model is specified as the pruned v1-5 version.

NateOP•3/26/23, 6:47 AM

I added the --share argument

NateOP•3/26/23, 7:01 AM

I also changed the port to 3010

Next step in the tutorial says to save and webui-user.sh and then shut down the terminal - NOTE: at this stage in the process there is no terminal running, so I can't shut one down. However, I continue with the tutorial.

Next step is to launch the UI so, in the correct dir I start the relauncher.

I have gradio link now and can see the UI.

I follow the instructions and yes, I can generate images.

Next is to install Dreambooth from the extensions tab in the UI.
The one available is
"Dreambooth training based on Shivam Shiaro's repo, optimized for lower-VRAM GPUs.
Added: 2022-11-07"

NOTE: this is not the latest Dreambooth on Github

However, I install it and yes we see
"Installed into /workspace/stable-diffusion-webui/extensions/sd_dreambooth_extension. Use Installed tab to restart." in the UI.

Then I restart my terminal.

At this point I'm presented with a new gradio link for the UI.

Below that I get this error:

Error executing callback app_started_callback for /workspace/stable-diffusion-webui/extensions/sd_dreambooth_extension/scripts/api.py
Traceback (most recent call last):
File "/workspace/stable-diffusion-webui/modules/script_callbacks.py", line 101, in app_started_callback
c.callback(demo, app)
File "/workspace/stable-diffusion-webui/extensions/sd_dreambooth_extension/scripts/api.py", line 524, in dreambooth_api
model_cfg: DreamboothConfig = Body(description="The config to save"),
NameError: name 'DreamboothConfig' is not defined

NateOP•3/26/23, 7:02 AM

And so the extension has not been installed, there is no Dreambooth tab

NateOP•3/26/23, 7:21 AM

So, I try restarting the UI from the extensions tab. Then I try to relaunch the UI with "python relauncher.py"

As it runs I see the following issues:

Launching Web UI with arguments: -f --port 3010 --xformers --listen --enable-insecure-extension-access --share

//then we see a list of all the bracketed arguments, too long to copy and paste here...//

launch.py: error: unrecognized arguments: -f
Relauncher: Process is ending. Relaunching in 2s...
Relauncher: Launching...
Relaunch count: 1

And then it just loops around like that trying to launch but doensn't

NateOP•3/26/23, 8:14 AM

I also tried the tutorial "How to install new dreambooth extension on RunPod...."
- again, following the tutorial to the letter, picking a new Pod with 62 GB RAM this time.
- when I get to the part where we "kill the instance" I type fuser -k 3000/tcp ...but nothing happens.

NateOP•3/26/23, 8:16 AM

and relaunching, again, it starts with a Web UI argument with -f
(!!!) Not sure why it's doing that?

Launching Web UI with arguments: -f --port 3000 --xformers --listen --enable-insecure-extension-access

Again I'm left again with:

launch.py: error: unrecognized arguments: -f

NateOP•3/26/23, 8:18 AM

this is how the whole lot looks.

Any help would be greatly appreciated. So far I've had no luck installing Dreambooth

NNate Hi, I'm banging my head against the wall trying to install Dreambooth to a RunPo...

Furkan Gözükara SECourses•3/26/23, 10:29 AM

follow this tutorial it works great https://youtu.be/oI0GEnrySS0

YouTubeSECourses - Software Engineering Courses

How To Install New DreamBooth Extension On RunPod - Automatic1111 W...

RunPod: https://bit.ly/RunPodIO. Discord : https://bit.ly/SECoursesDiscord. How to install and start using #DreamBooth extension on #RunPod SD Web UI template. If I have been of assistance to you and you would like to show your support for my work, please consider becoming a patron on

https://www.patreon.com/SECourses

DreamBooth extension: ht...

How To Install New DreamBooth Extension On RunPod - Automatic1111 W...

Furkan Gözükara SECourses•3/26/23, 10:30 AM

if you still cant make it , if you are our patreon supporter i can setup for you via google remote desktop

FFurkan Gözükara SECourses look what is coming

Cypo•3/26/23, 10:31 AM

it is impressive

Cypo•3/26/23, 10:31 AM

good work as always!

CCypo it is impressive

Furkan Gözükara SECourses•3/26/23, 11:09 AM

the only part is eyes. i am trying if any deflicker there that can improve

Stevens•3/26/23, 11:39 AM

I have davinci and will try to run deflicker on it but i will have to reboot my computer and i am avoiding it

SStevens I have davinci and will try to run deflicker on it but i will have to reboot my ...

Furkan Gözükara SECourses•3/26/23, 11:40 AM

ah whenever you can

Original message was deleted

Cypo•3/26/23, 12:34 PM

@Furkan Gözükara SECourses again

CCypo @Furkan Gözükara SECourses again 😅

Furkan Gözükara SECourses•3/26/23, 12:50 PM

thanks :d

Furkan Gözükara SECourses•3/26/23, 12:52 PM

real vs animation

Furkan Gözükara SECourses•3/26/23, 12:52 PM

couldnt find free deflicker though

Furkan Gözükara SECourses•3/26/23, 12:52 PM

none working good

Furkan Gözükara SECourses•3/26/23, 12:52 PM

Furkan Gözükara SECourses•3/26/23, 12:59 PM

preparing script

Furkan Gözükara SECourses•3/26/23, 12:59 PM

Greetings everyone. I am excited to share with you my latest project – a tutorial for how to turn a real video into any style animation or anime like corridor crew did using a fully automated process. For the past 5 days, I have been working on perfecting the workflow, and I am thrilled to present my findings to you today.
I want to highlight that the animation you are currently viewing has been made solely using the Autoatmci1111 Web UI and Stable Diffusion 1.5 version. No hand-picked frames, no photoshop, no inpainting, or famous deflicker of Davinci Resolve were used. Furthermore, I have now more experience to make it even better. This means that the animation has enormous potential for further improvement.