Stable diffusion animation tutorial. Deforum ALL settings explained. Make your own AI video!

In this tutorial I'll go through all the settings in Deforum for #stablediffusion so you easily can make your own AI videos.
Links:
colab.research.google.com/git...
huggingface.co/CompVis/stable...
docs.google.com/document/d/13...
Chat with me in our community discord: / discord
Support me on Patreon to get access to unique perks!
/ sebastiankamph
Ultimate Stable diffusion guide
• Stable diffusion tutor...
The Rise of AI Art: A Creative Revolution
• The Rise of AI Art - A...
Ultimate Animation guide in Stable diffusion
• Stable diffusion anima...
Dreambooth tutorial for Stable diffusion
• Dreambooth tutorial fo...
5 tricks you're not using
• Top 5 Stable diffusion...
Avoid these 7 mistakes
• Don't make these 7 mis...
How to fix live render preview:
• Stable diffusion gui m...
CHAPTERS
0:00 Introduction
1:21 Explaining settings
11:45 Setting the prompts
12:22 More settings explanation
15:00 Let's make an animation
16:50 Let's render!
20:57 Results

[@sometimelucky]

Extremely helpful tutorial! Thanks a lot, so helpful to get information on what the different parameters mean

[@sebastiankamph]

You're very welcome! Thank you for the kind words 💎

[@servicestuffs2549]

Thanks I’ve used this for awhile and still learned some stuff! Great job!

[@sebastiankamph]

Great to hear! And thank you :)

[@ajdimarucot23]

Really thorough explanation of each cell! Thanks for this.

[@sebastiankamph]

Happy to help! Let me know if you create something cool.

[@ajdimarucot23]

@@sebastiankamph I do share them on my IG but won't spam it here haha

[@sebastiankamph]

@@ajdimarucot23 you're welcome to and I'll have a look!

[@adamgdev]

Siiiiick video! Many thanks!! Would love more of these on Deforum.

[@enExima]

Thank You, good to have your tutorial at hand..

[@gkennedy_aiforsocialbenefit]

Very promising. Great tutorial thanks!

[@kamikaze-ub9kz]

TNice tutorials the best tutorial I've ever watched! Super informatic and love the softow in wNice tutorialch you taught everytNice tutorialng. Thanks a lot

[@Darklegioncz]

Nice jobe dude thx for all tutorials.

[@gabrielmontoya6494]

Thank you dude, this is right on point and helped me a lot.

[@sebastiankamph]

You're welcome, glad it helped!

[@JTVizionNetwork]

Excellent video thank you so much!

[@blackvx]

Thank you for enlightening us!

[@sebastiankamph]

Thank you for sending nice words my way! ⭐💎

[@aigeneral12k]

Thank you for the video! Do you think you could do an update video, with newly added settings? Thanks again

[@ADSADSStore]

ty cuz ive been having a hard ti getting started.

[@DargSix]

Top video man, thank you very much

[@davewaldmancreative]

Brilliant. Thanks, Sebastian!

[@sebastiankamph]

My pleasure!

[@DJAG97]

Very helpful and clear tutorial. Is there a way to run the same process on a silicone chip Mac ? Thank you

[@irpacynot]

That is farkin' awesome. Thank you!

[@sebastiankamph]

Thanks a lot! 😊

[@irishnftgal]

Thanks so much for this!

[@sebastiankamph]

Glad it was helpful! Thank you for taking the time to comment 😊

[@TwinsBeats]

That intro though! 🔥

[@sebastiankamph]

Wow, that was fast. Haha. Thanks buddy!

[@TwinsBeats]

@@sebastiankamph you bet! We are subbed with notifications on. 😁

[@StableDiffusion-gv7hf]

i had a question. As you mentioned to take all frames and use premier pro to make a video out of it. How does it give you that morphing or AI video shifting look in premier pro?? Please guide.

[@karan.awasthi]

Really good tutorial. Very easy to follow for beginners and explains pretty much what one needs to get it up and running. Does it allow users to upgrade the res from the default 512, like in a vertical format?

[@sebastiankamph]

Yes, but I recommend keeping it fairly low and upscaling afterwards.

[@NHZMusic]

@@sebastiankamph how to upscale later?

[@maitreyaprajapati4708]

thank you so much dude you're a god

[@sebastiankamph]

Champ, your presence here is a rainbow factory and that's a fact.

[@andrewstewart62]

Fumbled my way through this program on my own before finding this video. You cleared up some questions I had as for what each setting did so thanks for that. My one question that I’m having a hard time getting an answer to is how to get more cohesive results whether from stills or animations? Everything always seems so abstract but I see people pumping out videos where there is multiple objects in frame that each are doing something separate. I’m also seeing animations that make more sense in general and I’m curious if you can shed any light on settings to play with to get away from the abstract

[@yaylah7314]

you can do a composite of different animations with after effects for example. just render some animations out of this and create a montage there

[@tilius.xyz_]

use chat gpt to help you write improve your prompts, make them long and accurate, keep testing to the desired preference.

[@tereanselm]

I've managed to get better results with increasing strength to 0.5 or more in init settings under Run (where you change to W and H and so on)

[@aivideoyoutube]

Thank you

[@sebastiankamph]

Most welcome!

[@MrPlasmo]

Another great tutorial! thanks. I have a question - will using Deforum in my Google account use up my total Google storage? I barely have any storage space left and was wondering if running this will use up more of my storage space. thanks

[@sebastiankamph]

Thanks! And yes, it will. Each image saved will use up Google drive storage. But you can move these out from your Drive and use a local storage when you've finished rendering.

[@gadigila]

thanks for the video, its work also for mac?

[@vijaychoudhary9997]

great tut

[@sebastiankamph]

Thank you, Vijay!

[@hitlab]

great tutorial! Thank you. Question: is Deforum animations comparable to PYTTI animations? What's the difference?

[@sebastiankamph]

I'm not well versed in Pytti. I'd ask in the Deforum discord.

[@eloisamenasaravia9398]

Works well!! DANKEEE

[@sebastiankamph]

You're very welcome 😊

[@jerryjack6976]

great tutorial ! I am not sure if the model downloaded properly for me, it does not look like the same type of file as in the video, it just appears blank with the same title. I can not enter it as a file like you do in the tutorial

[@InducingMusic]

Great tutorial 🔥 Do u know what is the max resolution for video, hopefully HD? Also, I got the error "You have not accepted the license for this model.".

[@sebastiankamph]

Depends on your gpu. But can easily run and upscale to get full hd and even 4k. You need to accept the license in Huggingface for the model.

[@Collants]

WOW, thanks A LOT. Greetings from Chile

[@sebastiankamph]

You're very welcome! Sweden sends its regards to Chile! 🔥🔥

[@KyleKesterson-sw2yz]

Thank you for this! Would you be able to do an updated version that walks through the steps of connecting Colab to Stable Diffusion 2.0? Down/uploading the new packet to Drive is spitting out errors in the custom checkpoint path. Appreciate any guidance here!

[@traviswilliam2490]

this!

[@HoppingMadMedia]

Thanks for the great tutorials! So, I've done all this, but now my question is, once I'm ready to send all these images to Premiere Pro for editing, how do I upscale/up-res the video/frames for exporting a video without losing quality or becoming pixelated? I can't seem to find a good tutorial on that.

[@sebastiankamph]

You're welcome! I'd recommend running a batch upscaler like Gigapixel and when you have your upscaled frames, import them in Premiere as an image sequence. Just remember to interpret the footage as the frame rate you worked with.

[@HoppingMadMedia]

@@sebastiankamph Thanks so much, man! I'll give this a go!

[@sorijin]

I've been upscaling from 382x640 and and 512x832 using the Extras/ Batch from Directory using just Lanczos as upscaler 1, very clean upscales as far as I can tell

[@barickfoster]

hi sir, for some reason , when i run the program i have many folders created with image sequences instead of just one main folder for all the images sequences; i would like to know what i did wrong , or checked for that to happen , thanks

[@BjarkiH]

Hi, great tutorial!. But I sometimes get a error in the NVIDIA cell that says "NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running." Do you know what I can do? I tried updating to the most recent driver and have restarted my computer but it is this there

[@ashsxlvr]

youre a legend 🐐

[@sebastiankamph]

Why thank you, that's very kind 😊

[@yanex4631]

sir please make part 2 with better animation🙏, and I have subscribed to your channel👍

[@user-hl7uz6jq5o]

Hello, and what actions can slow down the animation and make it a little smoother, I am grateful in advance

[@optimoos]

@Sebastian Kamph I'm trying to automate camera motion settings -> angle, zoom, translation, rotation... it's turning into a chore, and I lack special creative effects that come with randomness... Are you aware of a tool or anything that could make this easier (well, automatic :D)? thanks

[@andresmarcial6506]

Thanks for the tutorial. I tried it a couple of times but it takes hours to render the animation. Do you know why it happened? I used a resolution of 1080 x 1920 in image settings. Do you think the slow image rendering might be due to that high resolution?

[@sebastiankamph]

Yes, you must go much much lower. Then upscale afterwards.

[@andresmarcial6506]

@@sebastiankamph thank you very much for your reply

[@Kontor23]

Is it possible to set a second image for the interpolation, so that you not only set an init_image but also the goal image for the interpolation. So you could precisely end at a certain image instead of working with the animation_prompts?

[@sebastiankamph]

I believe only 1 init image is possible now, but you can prepare what images you want interpolated by testing out the prompts preemptively.

[@naturalloghouse]

can I run this directly from google drive, or do I need to have google drive downloaded to my MacBook as an app?

[@MrPlasmo]

Another question - when creating an animation using Deforum, is it using your local computer resources (i.e. do you need a good graphics card), or are you using google's gpu resources (if so, can you run multiple instances using different google accounts)? thx

[@sebastiankamph]

If you're running Deforum in Google Colab, then you are using Google's hardware and not your own so your computer can rest easy :) You can run multiple instances of Google Colab, but Google might temporarily (hours up to a day) but you in Colab jail, preventing gpu access :)

[@MrPlasmo]

@@sebastiankamph thanks. It looks like I ran out of GPU usage so can no longer connect for a while :( Is there a way to download the deforum locally so I don't have to use google gpu?

[@siddharthaxX]

and you can even use this notebook on mobile too

[@kk47shooter91]

thank you so much Sebastian Kamph, my only question is . when you want to start the animation from your own still image, what are the settings? I Checked use_init and I attached my jpg to the ini_image but it still stars as a robot..

[@kk47shooter91]

ok I figured it out. I forgot to set the strength to 1

[@stolencoats63]

I watched your stable diffusion tutorial video and installed it on my computer. Now I want to do animations like you do but this tutorial starts off with deforum in Google collab which I have no idea about. Is there a way to do this on my pc with the stable diffusion I installed following your tutorial? I just built a beast of a machine so maybe it can handle it? Ryzen 9 5900x, 64gb ram, rtx 3080ti, nvme drive etc.

[@stolencoats63]

@@dyland.8395 thank you!

[@michalwislicki6422]

Dear Sebastian, do you think its possible to use images, or links as animation_prompts? I want to marge my artworks but i can't figure out how... Much obliged for your help

[@sebastiankamph]

There's only the option to start with an init image. However you can interpolate between images but it's not as good.

[@michalwislicki6422]

Dear @@sebastiankampht thanks a lot for opening door to completely different dimension of creation! you da best

[@anirudhsays1534]

Thanks a lot for this, really helpful, would be great if you could plz do the same with warp fusion? Its really producing great videos off late.

[@sebastiankamph]

Thanks, I have been looking at it a bit. Thank you for the idea.

[@nickzardiashvili624]

Complete noob here. I'm not sure what this /content/drive/MyDrive/AI/models is referring to, where do I find/make this My Drive folder?

[@faededreality]

fantastic video! thanks! i followed everything but i still get the following error: error: OpenCV(4.6.0) /io/opencv/modules/imgproc/src/color.cpp:182: error: (-215:Assertion failed) !_src.empty() in function 'cvtColor' any idea on how to solve it?

[@danielpotes8508]

ROCKSTAR!

[@gradientboy]

can I design a UI with inline tooltips for extended usability?

[@learnanythingforfree2131]

Your video is super helpful. But unfortunately the Google Colab Workspace is super buggy! Its annoying for beginners. I get so many runtime errors.

[@chani_vlogs]

haha I'm with you there

[@gorkemgulan]

This animation runs so quick. 8 min for 500 frames. I spent once 11 hours for a 8sec animation and best couple hours with s colab pro+ account. How come? Is it because you skip steps in the middle. Which setting was that??

[@bekirdonmeez]

I have a question I run all but no idea I got instead a video its render picture

[@TEROUZ]

Can I use original video footage to render it this way? what link for a tutorial can I check out? thanks

[@sebastiankamph]

Yes, check this video on video input: kzbin.info/www/bejne/fqulinZ9naubnpo

[@TEROUZ]

@@sebastiankamph Thank you for the reply, checking it out

[@3pm479]

Thank you so much for the video. I am having issue in generating the video, the cell says some error, IDK how to fix it

[@sebastiankamph]

What's the error? I also uploaded an how to install Deforum locally, just minutes ago if you want to run it on your computer.

[@3pm479]

@@sebastiankamph In the Colab version while generating video it says "ValueError: max() arg is an empty sequence", they have changed the code and now its not working. In the Automatic1111 Colab Deforum shows error "No Module Named py3d_tools" , Please fix this. I have even asked to deforum discord help but no results.

[@DreamsinFrames1899]

Same here. Is there any fix?

[@andersblum]

Lmk if u find a fix

[@HoppingMadMedia]

I can't get deforum to generate images at 16:9 (640x360). Even when I change the settings in the google collab for image output to 640x360, the images still come out 640x320, and so when I go to upscale the images, I can't get true 16:9 unless I have horizontal black bars running along the top and bottom of the video.

[@sebastiankamph]

Yeah, this is an issue currently. I usually run 640x320, upscale beyond 1080 and crop parts out.

[@Critical-Vibrations]

The reason is because SD needs resolution to be in increments of 64....if you render larger images, you can get closer to 16:9 aspect ratio, thereby needing to crop less

[@gabriellekamph]

👏🙏

[@akstudios5450]

Enable the prebuilt GPU in colab to speed up the rendering. Thank me later 😊

[@css2a]

Hi sir! when i run it show error this: TypeError: render_animation() missing 1 required positional argument: 'uncond_prompts'

[@priceyt6839]

I get the error of running out of cuda memory. How much do I need? I have colab pro.

[@Calachi]

I keep getting an error in Python Definitions: FileNotFoundError: [Errno 2] No such file or directory: './externals/Next_ViT/classification/nextvit.py’

[@fortuna9131]

hey! i was rendering an animation on deforum, until frame 1200 all ok, but when i resumed the animation the last time i had an error because I typed 3 in spite of .3 on the rotation so it was so trippy from frame 1200 to 2400 (ending of animation). is there any way i can delete the frames i dont want and resume the animation from an exact frame? i tried to remove the frames i dont want from drive(so i deleted frames 1200 to 2400 in the drive folder)The thing is that when i try to resume the animation again there is nothing rendering. Any idea? this is the message I recive: "Saving to /content/drive/MyDrive/AI/StableDiffusion/2023-03/StableFun/20230315110619_* " but it never render the animation just says that message and gets de green tick without any render any help pls? thanks!

[@sebastiankamph]

Should work just deleting frames and resuming. I have done it multiple times.

[@lockswap]

have you considered taking images from these animations, then upscaling with img2imgHD for the traditional Stable Diffusion? 🤔 Have had trouble getting Deforum to work consistently also get this error most of the time. TypeError: Object of type set is not JSON serializable

[@YKutsuya]

Your error simply means your JSON isn't in the correct format. you probably have a , too much somewhere or too little.

[@blackvx]

I have a RTX 3080 12.0GB and I have a hard time with the 3d animation if I go beyond 896px X 512px. Does anyone know what would it take to run 3D at 1280X720? Thank you!

[@sebastiankamph]

It's better to run it at something like 640x380 and then upscale afterwards

[@dix_xib]

Kung.

[@wesleymessina5713]

Is there a way to connect Google collab Deforum notebook to my local GPU?

[@correctopinion4708]

there isn't, big shame!

[@andersblum]

"ValueError: max() arg is an empty sequence" when I try to load the video in "skip_video_for_run_all" how to fix?

[@rlemat7843]

Did you manage to find an answer to this problem? cause I'm in the same situation.

[@andersblum]

@@rlemat7843 I just ran it all again and it worked, also uncheck the box before u render the video, fixed it for me

[@andersblum]

uncheck the "render steps" box before clicking go

[@onlyvibhor]

can you make this video again with webui?

[@seencapone]

You said the strength schedule was the most important setting for animations but you didn't explain it very well or in depth. What are the useful value ranges? What does entering a lower or higher number actually result in practice? Why did you enter in the (Steps) x (Strength) calculation and then half the result?

[@sebastiankamph]

Sample steps - (sample steps*strength_schedule) = subsequent frames number of steps. I'd play around with 0.4-0-7 and learn from that. Low value = higher subsequent steps ie. more changes to your images.

[@rayamc607]

As with all things AI it generates almost random stuff and you pick and choose till you get what you want or just stick with what you get. It will be so much better when these things are controllable by better means than text prompts. What needs to be worked on now is the human to machine interface so that you get what you have in your head from the AI and not the pick and choose you have now. It's great what it can do but it's still early days and still very limited in someways and at the same time explosive

[@CHillaxUniverse]

evrything depends on prompts you use..learn better promts or learn how AI works LOL

[@rayamc607]

@@CHillaxUniverse Put another way - it doesn't get what's in my head on the screen ... it's guiding me more than me guiding it. Getting better though

[@hsyncnbayraktar]

When I change the Image Settings to different then 512x512, I get "out of memory" error, how can we change the resulation of animation ?

[@sebastiankamph]

You need to run with lower resolutions and upscale in another software. I usually run something close to 640x320 and then just upscale afterwards.

[@hsyncnbayraktar]

@@sebastiankamph thank you very much, I will try right now

[@vifx1]

my man please a need some help!

[@zombiefisting544]

where can i find a vid on v0.5 by chance? i cant find the one your using on this video i can only find v0.5

[@sebastiankamph]

99% is the same. Some new options in 0.5 but this is still valid.

[@zombiefisting544]

@@sebastiankamph thank you so much I'll deff try your full tutorial when I get home.

[@gobi_dqbt]

Do I need to install Python / Stable Diffusion before hand?

[@sebastiankamph]

Only if you want to run it locally. This guide was made with Google Colab, which requires no local installation.

[@gobi_dqbt]

@@sebastiankamph Ok great! Can't wait to see the first results. Until now : nothing :D

[@Processador_DE_BaixaQualidade]

@@gobi_dqbt got something?

[@gobi_dqbt]

@@Processador_DE_BaixaQualidade yessir, used colab.research.google.com/drive/1_MckQnU0mCF8FJ7PoV21nhVvoOeotHGy?usp=sharing for it :) gonna drop the vid the 30 december with my new single

[@mahmood392]

i have this issue.. where it just becomes plain with no detail for a while in... zoomed in too far or something I'm not sure. Im using the Audio Zoom frames also.

[@mahmood392]

it shows my subjets for a bit and then goes plain for awhile then someone pops back out again... its not detailed the entire time.

[@sebastiankamph]

@@mahmood392 Probably too much zoom. Try to stay closer to 1.

[@mahmood392]

@@sebastiankamph im using the audio frames for zoom.. the power is set to 4. this is usually always the case. i have even done 1.02 and i still get this blank empty places often..

[@sebastiankamph]

@@mahmood392 How high and low are your minimum and maximum zoom values?

[@mahmood392]

@@sebastiankamph the height iv seen is 1.04 and lowest is 1.00. And I set my strength at 0.65? 🤷‍♂️

[@itsErmando]

Check out some of my Daily Animation Challenge in here : kzbin.info/www/bejne/rZawmpWJeKZqd8k&ab_channel=Ermando

[@isidoraaxis6179]

on the "Load Settings" im getting the following error │ in :117 NameError: name 'DeforumAnimArgs' is not defined any one knows what to do?

[@isidoraaxis6179]

I re run the collab, and it worked ;)

[@yasatips4300]

For example, finding out wNice tutorialch notes fit into a particular chord (chords make up the basis of most soft) - you don't need to even know

[@cliveagate]

Another 'Top Class' tutorial

[@sebastiankamph]

Thanks again, Clive! And a super thank you for joining up and supporting my work here. You're a real superstar ⭐

[@cliveagate]

@@sebastiankamph my pleasure and thank you for sparking my imagination with your singing animation video.

[@visionsbytj3984]

Awesome video & thank you immensely for taking the time to make it. Ran into an issue that I'd appreciate your input on: Came across these errors which halted the process -> NameError: name 'args' is not defined, NameError Traceback (most recent call last) & 31 else: # render images for a video Any light you can give to get out of this tunnel?

[@sebastiankamph]

Thanks! Hmmm, no not sure. Try saving your settings and restarting.

[@captain9foxx]

I followed the video but my still does work keeps saying there a error and idk how to fix it

[@andersblum]

Me too, please lmk if you find a fix. "ValueError: max() arg is an empty sequence"

[@marcosfraguela]

Does anybody know wheather it's legal or not to use the generated images for commercial purpouses?

[@sebastiankamph]

According to SD's license, yes.

[@ash.r3753]

" max() arg is an empty sequence " I am getting this error on trying to view the video. Any solutions?

[@andersblum]

I get this too. Please lmk if you find a solution

[@Xishnik94]

i keep getting loss of sampling steps and resolution after the first frame... what am i doing wrong?>

[@sebastiankamph]

Check strength_schedule settings. It is explained in the guide. 99% that's the issue.

[@Xishnik94]

@@sebastiankamph Thank you! I did hear you say it shortly after posting my comment :)! 100% appreciate these guides!

[@babuinobabuino6870]

max() is an empty sentence ??? what am i supposed to do?? i get that on the render step

[@andersblum]

I get this too. Please lmk if you find a solution

[@tarazz]

why i get this error " NameError: name 'args' is not defined" while try to render video?

[@andersblum]

I get this too. Please lmk if you find a solution

[@tarazz]

@@andersblum no(

[@marmaladesky]

where do i get a token authorization

[@sebastiankamph]

If you're not saving the model, you can get it through Huggingface. Check the settings and create your Write access token.

[@EvladIonescu]

What are your pc specs?

[@sebastiankamph]

Ryzen 9 5900x, 32GB ram, RTX 3080

[@EvladIonescu]

@@sebastiankamph I have an amd 8350 970gtx with 16 gb ram 1600 but I managed to get 20 second text to img. I will get a 3060 with 12 gb and I think I will golden :D

[@selftransforming5768]

[@user-wv2cq3kx8b]

я читала оф инструкцию поняла что сначала нужно Питон выучить, а потом видео делать

[@CyborganicDreamscape]

i love that month old videos are outdated in this space. this shit is gonna get weird in a year.

[@sebastiankamph]

I wouldn't really call this one outdated as 0.5 just added some settings. So all this is still correct and valuable knowledge. But I understand your sentiment and it's wild trying to keep up! 😅

[@dadlord689]

It really would make more sense if you could draw something meaningful. If there was a way to produce animation of character jumping, fighting or doing anything in-place - that would be valuable. This... well, it is sickeningly.

[@sebastiankamph]

Check my latest video for that

[@dadlord689]

@@sebastiankamph I will! Thanks )

[@myronjones8113]

@@sebastiankamph which video?

[@sebastiankamph]

@@myronjones8113 The one about singing animation

[@llupescu]

I keep getting this error when I try to do an animation: ValueError Traceback (most recent call last) in 24 fname = f"{path_name_modifier}_%05d.png" 25 all_step_dirs = [os.path.join(args.outdir, d) for d in os.listdir(args.outdir) if os.path.isdir(os.path.join(args.outdir,d))] ---> 26 newest_dir = max(all_step_dirs, key=os.path.getmtime) 27 image_path = os.path.join(newest_dir, fname) 28 print(f"Reading images from {image_path}") ValueError: max() arg is an empty sequence I don't understand why, can you help me, please?

[@sebastiankamph]

Looks like you deleted a value somewhere. Feel free to post your settings in our Discord channel.

[@llupescu]

@@sebastiankamph Thank you!

[@ORGANICMINDART]

Hi .. thanks for the tutorial. but somehow i got the error " --------------------------------------------------------------------------- NameError Traceback (most recent call last) in 66 67 # config path ---> 68 ckpt_config_path = custom_config_path if model_config == "custom" else os.path.join(models_path, model_config) 69 if os.path.exists(ckpt_config_path): 70 print(f"{ckpt_config_path} exists") NameError: name 'models_path' is not defined " .. can you show me how to fix this error ?

[@sebastiankamph]

Did you put your model.ckpt in the models folder?

[@ORGANICMINDART]

@@sebastiankamph yes i put it , but maybe cos i not rename it into model.ckpt .? i will try again rename it the sd-v1-4.ckpt into model.ckpt

[@ORGANICMINDART]

can i chat you in Discord ?

[@SentienceGeneration]

Is this free as long as I don't ask for too many frames/ scaling?

[@sebastiankamph]

Yep

[@SentienceGeneration]

@Sebastian Kamph I ended needing to buy Collab Pro, not sure how the compute credits work though. I'm hoping it's 1 per video I make..