Тёмный

Stable Diffusion Crash Course for Beginners 

freeCodeCamp.org
Подписаться 10 млн
Просмотров 246 тыс.
50% 1

Learn how to use Stable Diffusion to create art and images in this full course. You will learn how to train your own model, how to use Control Net, how to use Stable Diffusion's API endpoint, and more.
The ethical implications of AI in art are significant. This course aims to educate on the technical aspects of the tool. We emphasize the importance of using it responsibly and respecting the rights of artists.
✏️ Course created by ‪@lynnzheng‬
🔗 Install stable diffusion: github.com/AUT...
🔗 Download models: civitai.com/
🔗 Download VAE for Model "Counterfeit": huggingface.co...
🔗 Download Embeddings: huggingface.co...
🔗 Train a Lora: civitai.com/mo...
🔗 Install Control Net - github.com/Mik...
(Use --enable-insecure-extension-access extensions)
🎉 Thanks to our Champion and Sponsor supporters:
👾 davthecoder
👾 jedi-or-sith
👾 南宮千影
👾 Agustín Kussrow
👾 Nattira Maneerat
👾 Heather Wcislo
👾 Serhiy Kalinets
👾 Justin Hual
👾 Otis Morgan
--
Learn to code for free and get a developer job: www.freecodeca...
Read hundreds of articles on programming: freecodecamp.o...

Опубликовано:

 

24 сен 2024

Поделиться:

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист
Посмотреть позже
Комментарии : 277   
@lynnzheng
@lynnzheng Год назад
Hey everyone. Lynn here. Had lots of fun making this video and hope you enjoyed it :) We created the course hoping to teach you about the tools and we gently request that you do not use it for evil causes :)
@lynnzheng
@lynnzheng Год назад
We at freeCodeCamp respect the work of artists and acknowledge that AI-generated art using Stable Diffusion is a tool that can be used to enhance creativity, but it does not replace the value of human creativity. Happy coding!
@gpk_2544
@gpk_2544 Год назад
Wow thanks for this amazing dedication you have put in ! Thanks for your contribution 🙏🏻❤❤❤
@LaichuTV
@LaichuTV Год назад
thank you soooooooooo much! now I can create more charactor images for my story ❤
@hienvinhle1797
@hienvinhle1797 Год назад
Thank you ❤
@casul-4891
@casul-4891 Год назад
No 😈
@ajay0909
@ajay0909 Год назад
This course shows how to use the diffusion models, it doesn't have anything about the model building. This is a good course for anyone who wants to know about stable diffusion
@LordMembrane
@LordMembrane 6 месяцев назад
Begginers with stable diffusion, experienced with Linux and programming.
@ichiroutakashima4503
@ichiroutakashima4503 Год назад
I'll BE DAMNED. Something I knew but clearly has no idea how it works is now being featured with a tutorial with FCC! I will forever be grateful with this community ever since my web development days and changing my future!
@simonboucher5170
@simonboucher5170 7 месяцев назад
I was looking for someone who would show me how to use different options and tools after i was able to setup my local web ui. Thank you a lot. I think i will be able to make better images now that the different options have been explained and demonstrated to me.
@moisesmichelow2104
@moisesmichelow2104 Год назад
Finally! I can be an artist without having any art skills.
@khal3300
@khal3300 Год назад
not really you'd be more of a prompter , AI is doing all the art and creative part .
@HigherOrderBeing
@HigherOrderBeing Год назад
Haha me too better than picaso
@HigherOrderBeing
@HigherOrderBeing Год назад
​@@khal3300 whatever 🗿
@khal3300
@khal3300 Год назад
@@HigherOrderBeing just facts buddy , AI is here to stay we all know it , but stop coping into thinking you're an artist because you can type prompts lmao , or that AI art is art , is just generating images from an endless database of real art , art requires abstract thought , when AI becomes sentient and makes art of it's own initiative then we can call it AI art .
@moshmoshpitpit
@moshmoshpitpit Год назад
Someone didn't watch the video
@inkwyvern5171
@inkwyvern5171 Месяц назад
Stable difussion 100% replaces the value of human creativity. It's amazing!
@sugaruling
@sugaruling Год назад
Finally, a course that will help me in life ❤
@plinketharry7469
@plinketharry7469 2 месяца назад
This was great guys i massively looking forward to more ai, stable diffusion and hopefully llms. You're making a difference out here
@parama_padida_dei
@parama_padida_dei 10 месяцев назад
Starts form 7:30
@jojobizadTRASH
@jojobizadTRASH Год назад
I wish i had a good computer for this. The Bing Stable Diffusion is insane
@tigullavikas4516
@tigullavikas4516 11 месяцев назад
wt do mean by good computer, can u be specific, i want to learn this project, can i use 8gigs ram + 2 gigs graph power.
@jojobizadTRASH
@jojobizadTRASH 11 месяцев назад
@tigullavikas4516 requires a lot more than that I'd I'm going to be honest. Anything that's 12gb and beyond should be pretty decently for both GPU and RAM
@harshjain8345
@harshjain8345 9 месяцев назад
Cloud is an option
@GersonMartinez-m3b
@GersonMartinez-m3b 9 месяцев назад
I wish it would've explained the process of installing it more. You just went I'm installing it, it's installed. I also wish you would've gone into more detail about the specs needed to install
@minhdao7603
@minhdao7603 Год назад
Finally! I can coding with 1 hand
@chinhcao8820
@chinhcao8820 Год назад
💀💀
@ThePessimisticPothead
@ThePessimisticPothead Год назад
Wow that artwork looks very familiar
@spraynpray
@spraynpray 2 месяца назад
I watched the first 4 minutes of this video. I want my 4 minutes back.
@laurenceturpin1409
@laurenceturpin1409 8 месяцев назад
Hey Lynn Just wanted to let you know I really enjoyed the course. Thank you for making it.
@korontos9038
@korontos9038 Год назад
btw only the a1111 front end of stable diffusion is banned in google colab , other front ends like comfy or stable swarm are not .
@DiogenesTheCynic.
@DiogenesTheCynic. Год назад
Comfy isn't good at all, don't know why you suggest using that.
@lynnzheng
@lynnzheng Год назад
Thanks for the suggestions! Will look into them :)
@korontos9038
@korontos9038 Год назад
@@DiogenesTheCynic. its very very complicated but its more effective and capable than any other stable diffusion front end
@korontos9038
@korontos9038 Год назад
@@lynnzheng you're welcome :)
@DiogenesTheCynic.
@DiogenesTheCynic. Год назад
Broken and complicated mess you mean, and not something you want to suggest to beginners who just want to type and click generate. You would be scaring them away from wanting to learn.
@CDCDC697
@CDCDC697 10 месяцев назад
Any recommendation for someone who is looking to start working with Stable Diffusion and has a macbook pro? Does it even work on macbooks? I don't want to purchase an extra Windows laptop just for running Stable Diffusion. I want to use the video animation part.
@edwardn8165
@edwardn8165 11 месяцев назад
Great vid but lots of things are unclear for the noobs, quite fast pace with installation not explained how to instal certain things and wehre they were taking from couldnt follow properly, feel like for more advanced users rather beginners
@The_SUN1234
@The_SUN1234 10 месяцев назад
it doesnt just feel...it is.
@Tom_Cii
@Tom_Cii Год назад
I do want to point out that this is a massive gray area both legally and morally in terms of stealing other people's work if using other people's art for feeding your model, as there is 0 compensation nor consent given usually with the big art ai models.
@pandyssianrat
@pandyssianrat Год назад
​@@LizerinoCopying and pasting is different than observing and copying. Also, there wouldn't be any issues if people used these for personal stuff. Instead they use it to gain traction/money.
@CrazyWeridoRH
@CrazyWeridoRH Год назад
Yeah I’m unsubbing from this channel. I’m not going to support people who support plagiarism tech.
@Singularity606
@Singularity606 Год назад
@@CrazyWeridoRH You don't know what plagiarism is.
@CrazyWeridoRH
@CrazyWeridoRH Год назад
@@Singularity606 If you put my work into a machine and have that machine pop out my work with just a different background and lineart with a new hat or etc, and then have the gall to say you "made it" when all you did was the same skill it takes to google something, then I'd be pretty insulted and would consider it plagiarism.
@spectri1990
@spectri1990 9 месяцев назад
I might be brain dead, but this tutorial is super bad. They say instal scroll down don't show what links to click what order to download things. They just go straight to the script downloading. When you go to the site it's super confusing to know what to click to download but even when you find what you think is a download, you it has an error.
@liamcoutou8437
@liamcoutou8437 2 месяца назад
I'm 7 months late but... It's not that hard to type "git clone", "install requirements" and then run a file that says start
@spectri1990
@spectri1990 2 месяца назад
@liamcoutou8437 what the hell are you talking about. if that's all you needed to do, this video wouldn't exist. Also, she could've just done what you said in the video
@tigullavikas4516
@tigullavikas4516 11 месяцев назад
can i use vae from 2.5 version and rest models from 3.0? im following the guide, so im a bit confused :)
@lhxperimental
@lhxperimental 10 месяцев назад
Hmm..I see a lot of magic happening, I have no clue how anything is working. I have too many questions about its working. I understand it is not possible cover everything in one hour and this is a good high level summary of the possibilities. Thank you!
@JenniferBland
@JenniferBland Год назад
@lynnzheng Can you share the specs of the laptop you are using for building models? Windows/Mac, memory, storage, GPU?
@zachadolphe3633
@zachadolphe3633 9 месяцев назад
Running the stability diffusion program on an nvidia rtx 3070 takes about 10-20 seconds to run and do. So pretty good and that gpu isn't anything crazy.
@GodlyGale
@GodlyGale Месяц назад
loved
@simpleshun
@simpleshun Год назад
if we are getting ai art course might as well add a course on how to exit vim.
@The_SUN1234
@The_SUN1234 10 месяцев назад
i understood nothing, and you using the linux didnt help at all.
@HikikomoriDev
@HikikomoriDev 6 месяцев назад
Something tells me that Pela from Star Rail was modeled and textured by AI processes, a lot of these AI models look a lot like her in a way.
@gjsxnobody7534
@gjsxnobody7534 Год назад
where is the "how to connect my app to API" part? I can't find it
@nah131
@nah131 Год назад
AI tools are the future tools!! I have been using Stable Diffusion since April this year. There are tons of extension that we can use in SD. Not to mention, the latest update SDXL 9.0 is dope too.
@brandon_wallace
@brandon_wallace Год назад
Thanks Lynn.
@JorgeMartinez-xb2ks
@JorgeMartinez-xb2ks Год назад
Thanks Lynn, best tutorial about the subject.
@DanielRodrigues-bx6lr
@DanielRodrigues-bx6lr 8 месяцев назад
The commandline arguments aren't working, it just keeps giving me an unrecognized arguments error
@bestonyoutube
@bestonyoutube Год назад
Is there a guide for Windows too? Also for training your own models?
@deadlyraver4454
@deadlyraver4454 10 месяцев назад
I too want this. I'm stuck trying to figure out where to put the various resources.
@shugucchi
@shugucchi 10 месяцев назад
Using the same site you can install the package for windows so it uses your gpu only, which is how i did it. You download the zipped file, run the updater, then run and eveything else should be the same. training your own model is something i can't help with right now as that also my next step I want to learn.
@deadlyraver4454
@deadlyraver4454 10 месяцев назад
@@shugucchi thanks. I did figure it out not very long after posting and I'm trying like hell to stop creating eldritch demonic horrors now. The second you beat one stage, another one appears. 😆
@xade8381
@xade8381 Год назад
A Fullstack course with Nim 2.0 please
@sfj4959
@sfj4959 Год назад
It’s pretty cool 🎉
@SameerGudekarVlogs
@SameerGudekarVlogs Год назад
I want to learn this coding and all .. Please tell me the starting point
@squiddymute
@squiddymute 8 месяцев назад
“it won’t get technical” .. but requires machine learning background 😅
@polynomial
@polynomial 8 дней назад
This is linux training. It would have been useful to put that in the title sowe know it is for tech heads.
@driftingsodium2186
@driftingsodium2186 Год назад
I truly wonder if the artists gave consent to their art being used to train this model :")
@raidensh0gun
@raidensh0gun Год назад
They did not, and the day some law gets passed in the EU or the Americas, which prohibits the usage of an artitst's art to train said models, it can give many companies a run for their money. But then again, with the model's current capabilities, one can always create new training data to keep the model up to date. atleast for the next 5 years I believe.
@itsmimic007
@itsmimic007 Год назад
We'll they didn't. Copyrighted artworks were used without the creator's permission.
@jgalvan09
@jgalvan09 Год назад
Who cares, do you text a popular artist with a particular niche just to draw too?
@astreakaito5625
@astreakaito5625 Год назад
@@jgalvan09 not the same thing
@jgalvan09
@jgalvan09 Год назад
@@astreakaito5625 oh suuuure lol “ not the same thing “ sure buddy
@abdullahmertozdemir9437
@abdullahmertozdemir9437 Год назад
A full course on Julia or Zig please
@nazar1744
@nazar1744 Год назад
We are all here for the Waifu
@painlesslynx5550
@painlesslynx5550 Год назад
8:25 OMG LMAO IT'S CALLED SAFEBOORU. That screen scared the crap out of me LOL.
@BurGurr
@BurGurr 7 месяцев назад
💀
@rewanthnayak2972
@rewanthnayak2972 Год назад
I am waiting for tutorial on Fine Tuning a LLM
@imeaniguess.6963
@imeaniguess.6963 10 месяцев назад
14:38 I recognize that face, it’s Kim Chaewon. 😂 I don’t know how I feel about that. 😵‍💫
@yeanchan665
@yeanchan665 Год назад
I have a question. Can I share the ai module with friends??
@fintech1378
@fintech1378 Год назад
Can you cover the API too
@molyyang2738
@molyyang2738 10 месяцев назад
Great video but how and where do I get linux? I get it's like a virtual machine but I am like having an error on my side trying to get that started for me to try this :( ....
@divadolleh
@divadolleh 10 месяцев назад
RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check
@FRareDom
@FRareDom Год назад
very good tutorial
@benox1
@benox1 Год назад
Why arent you using ComfyUI?
@korontos9038
@korontos9038 Год назад
cuz she aint mentally challenged
@legendj.a291
@legendj.a291 11 месяцев назад
Keep it up
@tanyalalaby937
@tanyalalaby937 2 месяца назад
how do you do this on pc windows do you have another video for that?
@dantrrrrr
@dantrrrrr Год назад
love this
@arifkuyucu
@arifkuyucu 8 месяцев назад
Excuse me but aren't there resources for lazy people like me that contain ready-made prompts (with negatives), examples listed according to image types, and that we can learn by copying and pasting? I would prefer to have access to these resources instead of learning how to trim hands, feet and hair. What's the point of using AI if I can spend minutes correcting a simple image? I go and use photoshop etc.
@SeanKenyon-b6s
@SeanKenyon-b6s Месяц назад
nice
@mystikyogi
@mystikyogi Год назад
is there any api available which can be used to convert my selfies into ai art ?
@waleedkhalid9608
@waleedkhalid9608 Год назад
Nice 👍
@AdvaitPisal
@AdvaitPisal 11 месяцев назад
My laptop has Intel UHD GPU installed (which graphic driver 30.something) . Is it enough for performing the tasks shown is the video ? as I have my engineering project based on stable diffusion pls help
@tgazo5874
@tgazo5874 10 месяцев назад
I don't think this is a crash course for beginners. You seem to skip many steps.
@sushant.neupane
@sushant.neupane Год назад
Can you bring a good course on prompt engineering. Thank you for everything you do for the students and Learners.
@lynnzheng
@lynnzheng Год назад
Yes :) We will add that idea to the queue.
@VulgarTruth
@VulgarTruth 5 месяцев назад
Hey, I want to buy a laptop to start learning stable diffusion.. is gygabyte aorus x17 good for its price or what would you recommend? Cheers
@BlueprintBro
@BlueprintBro 3 месяца назад
If your focus is going to be ai art and ai stuff in general to my knowledge your rig should have plenty of vram and a beefy gpu for a good workflow. Most importantly vram. You seem to have a sizeable budget from the laptop you mentioned so imo i would recommend getting a laptop that has at least 16gb of vram and generally high end specs, but focus on vram
@liamcoutou8437
@liamcoutou8437 2 месяца назад
​@@BlueprintBro The only cuda laptop GPU with 16 GB vram is the 4090
@beichooo
@beichooo Год назад
Nice
@fun_iqp
@fun_iqp Год назад
It wouldn't been bad some more context, but nice overall.
@reaur5853
@reaur5853 25 дней назад
16:06 Training LoRA
@snowman1185-v
@snowman1185-v Год назад
YES!! Thank you
@Detrue
@Detrue Год назад
how much power my pc needs for this? Gpu and those things?
@TerryDavisMedia
@TerryDavisMedia Год назад
doesn't work on intel-based-graphics for anyone wondering.
@gabrielgomesmabiala6707
@gabrielgomesmabiala6707 Год назад
Great!!
@Giyu-sl4nf
@Giyu-sl4nf 3 месяца назад
I didn't get VAE or STABLE DIFFUSION file instead i got .SAFETENSOR why ???
@liamcoutou8437
@liamcoutou8437 2 месяца назад
That is a stable diffusion file
@rashedulkabir6227
@rashedulkabir6227 Год назад
Which companies will offer me jobs if I become expert at creating AI art with Stable Diffusion?
@poudlardo
@poudlardo Год назад
why so many dislikes?
@frogfox9577
@frogfox9577 Год назад
Someone know a good free cloud GPU ? because i have a radeon 570x and i don't think than is okay ..... it's a 4Vram ...
@brainstormingsharing1309
@brainstormingsharing1309 Год назад
🔴 Keep it up! 👍
@praveensuryapraveen4258
@praveensuryapraveen4258 9 месяцев назад
How to install in mac???
@AC0_0
@AC0_0 Год назад
les gooooo
@unknown-kb5wg
@unknown-kb5wg Год назад
Only high end pc spec can try this?
@treepalds3978
@treepalds3978 4 месяца назад
Yeah, it is not for low end pc, i've tried ai generating image with my gt 1030 and it was terrible. Unless you'll be the one that make this ai for low end pc
@sunnybabar0
@sunnybabar0 10 месяцев назад
Is it free?
@H0noki
@H0noki 7 месяцев назад
forgive me artists
@FaiGal
@FaiGal Год назад
First programming tutorial I feel like is gonna be a sin if I follow it. My brother is literally an artist who graduated from a good art school and I'd feel like I sh*t on all his hard work if I used this. But nonetheless this was very interesting.
@watynecc3309
@watynecc3309 Год назад
You're a good brother and a lovely person have a great day
@fotrj
@fotrj Год назад
What the hell? You let random strangers on the internet gaslight you into believing that you DON'T DESERVE to create good art easily. It is like making instant noodles and feeling you are offending some chef in China.
@astridalaniz8580
@astridalaniz8580 Год назад
​@@fotrjthis is a deeply unserious opinion and also that's not what gaslighting means lmao
@fotrj
@fotrj Год назад
@@astridalaniz8580 Unserious because you feel called out? Hahaha pathetic 🤣. Maybe you need to recheck what gaslighting is because making people feel guilty when they have nothing to feel guilty about is EXACTLY what it means.
@astridalaniz8580
@astridalaniz8580 Год назад
@@fotrj I just did and that's literally not what it means. The ai datasets are created with unethical means and using the ai to create images is also unethical. It's not gaslighting to have a moral compass.
@pastuh
@pastuh Год назад
1 year too late. lol
@lucienthurston366
@lucienthurston366 Год назад
How so?
@hck1bloodday
@hck1bloodday Год назад
we didn't have control net or LoRa a year ago
@samiturunen5335
@samiturunen5335 10 месяцев назад
Could be more compact...
@fgfanta
@fgfanta 11 месяцев назад
Thank you! (but pity the voice sounds like she is over the phone).
@itsmelukario5969
@itsmelukario5969 Год назад
@paulcousins6535
@paulcousins6535 8 дней назад
way too complicated...even if this is free. im using web based ai tools, but they cost too much
@risex1594
@risex1594 Год назад
Now showing how to make "art" with a neural network trained with other artists work without their consent. Nice
@aammssaamm
@aammssaamm Год назад
That's a reality now with everything. I doubt it can be fixed with laws. Handcraft, live music, etc. were hit hard before with other machines.
@freecodecamp
@freecodecamp Год назад
We agree that the ethical implications of AI in art are significant. This course aims to educate on the technical aspects of the tool. We emphasize the importance of using it responsibly and respecting the rights of artists.
@skyhappy
@skyhappy Год назад
Artists learn from each other. This is no different. Great artists steal.
@aammssaamm
@aammssaamm Год назад
@@skyhappy Exactly, artists, but this is a machine, not an artist.
@astridalaniz8580
@astridalaniz8580 Год назад
​@@freecodecamp the fact that you know it's unethical and still chose to make a tutorial about it actually makes it worse
@eugenmalatov5470
@eugenmalatov5470 9 месяцев назад
are there no timestamps?
@ranggayogiswara5148
@ranggayogiswara5148 Год назад
Lesss gooo
@Laucron
@Laucron Год назад
You mean AI image mashing
@misaonthefly
@misaonthefly 4 месяца назад
This is just a follow along... she doesn't explain any of the tools... thumbs down
@cheryldahl9192
@cheryldahl9192 6 месяцев назад
This is not for beginners.
@dw8200
@dw8200 9 месяцев назад
Great content, but the walkthrough is terrible.
@TowerEugenia-b2t
@TowerEugenia-b2t Месяц назад
like
@maksimsaikin3721
@maksimsaikin3721 Год назад
How to make my anime gf?
@CatholicWeeb
@CatholicWeeb Год назад
*Takes out Notebook*
@FloKorp86
@FloKorp86 9 месяцев назад
Hi, I've always enjoyed your courses, but this one seems quite odd. I'd consider generating images from other people's copyrighted material ethically ambiguous and I'd have expected that conversation to get resolved before you jump head first into an endorsement without acknowledging the consequences for real life human artists who are affected by their work essentially being stolen, just so you can show people how to pump out an infinite stream of images for free.
@M_7844
@M_7844 Год назад
Only NVIDIA graphic cards are supported. If you have AMD you are running on CPU (very slow). Windows only: Don't close CMD window once web UI is available. Don't type or interact directly in CMD window (you can "pause" the processing)
@kusali11
@kusali11 Год назад
You can run on Apple M1 and M2 as well.
@joshbarros1995
@joshbarros1995 Год назад
First❤
@JaCKal_646f67
@JaCKal_646f67 Год назад
Oh god 🤣💀
@jgalvan09
@jgalvan09 Год назад
😂 this video will trigger Artist talking about “ AI taking over artist job *old man voice* damn AI taking all of our jobs “ 🤣
@neociber24
@neociber24 Год назад
Maybe is because I'm not an artist, but still don't see this taking artists jobs, maybe junior jobs but an artist still have total control over the canvas, even with ControlNet there are limitations.
@DKYtut
@DKYtut Год назад
@@neociber24 digital artists already struggle with finding jobs, and no doubt there are small business projects that now prefer generated art over art made by a local artist.
@w1-w2-w3
@w1-w2-w3 Год назад
It won't because there are so many issue with intellectual property rights problem. Many websites don't accept AI generated art including Steam. If you want to use AI generated art, you need to hire the artist first and let them drew the base art style you wanted. And then train the Ai with that art style. I think that is one of the legal ways to use AI generated art.
@Namrec_Molai
@Namrec_Molai Год назад
We will become useless, i heard from somewhere gov will give us free money its crazy for you but who knows
@ughestrada
@ughestrada Год назад
Hollywood execs want to use AI to copy people's images and use them in whatever media they want without crediting or compensating them in any way, so uh, yes, AI will be taking people's jobs and will be used by the already wealthy to make more money. Also, if we're gonna automate everything, and everyone will be out of a job, then HOW are we gonna pay for anything?
Далее
How Stable Diffusion Works (AI Image Generation)
30:21
Просмотров 149 тыс.
When Players Sacrifice for Team ❤️
00:32
Просмотров 7 млн
THEY'RE EATING THE DOGS DANCE REMIX!
00:10
Просмотров 175 тыс.
НОВЫЙ РОЛИК УЖЕ НА КАНАЛЕ!
00:14
Просмотров 599 тыс.
Stable Diffusion Prompt Guide
11:23
Просмотров 32 тыс.
LORA training EXPLAINED for beginners
27:33
Просмотров 103 тыс.
How I'd Learn AI in 2024 (if I could start over)
17:55
Просмотров 993 тыс.
Why Does Diffusion Work Better than Auto-Regression?
20:18