Stable Diffusion 2 with AUTOMATIC1111 GUI on Free Colab

[@1littlecoder]

Use this updated Colab colab.research.google.com/github/TheLastBen/fast-stable-diffusion/blob/main/fast_stable_diffusion_AUTOMATIC1111.ipynb

[@leafdriving]

I have a 4GB card.... and watch the world continue (for a day or two) without me.... and watching your video fixed that. Thank you.

[@1littlecoder]

Than you for sharing :)

[@leafdriving]

@@1littlecoder I have proposed this to about 20 KZbinrs. One took me seriously and grew quickly. Ask your viewers for help with *any* project you like OR Challange your audience to create something based on what you just showed them, highlighing the person you judge winner on your next channel eposide. Engagement Is KEY. Your simple 2month old generic comment DID bring me back ~ I would have come back for a challange too. Its probable that some of your viewers not only surpass your ability, but are eagar and willing to help anyways. I wonder how much this channel could do a a group, rather than an individual.

[@JimmyGunawanX]

If you got glitching, instead of 512 x 512, use 768 x 768, your CKPT might be set for that as default base size generations.

[@1littlecoder]

Thank you

[@emimeemio]

hi, how I can fix this error? RuntimeError: Parent directory /content/gdrive/MyDrive/sd/stable-diffusion-webui/models/Stable-diffusion does not exist. Something went wrong, try again

[@darkbilo2194]

i have purchased colob. When I run the last cell ( start stable diffusion) it runs for 2 minutes then automatically stops ( green mark show and webui stop working)

[@genAIration]

Dude, can u use deforum extension? I always get a py3d_tools erorr 😫😫😫

[@Q-Bits8]

How to add a VAE from google drive folder? There is no option to chose vae?

[@rolandcampos1836]

Can't load models from hugging face it always say wrong link

[@mak_av]

hey thx for the video! can i ask if you menaged to get some external scripts to work? i was trying to use the Animator.v6 script but when i restart 1111 the script is just not there. thx again

[@nisargpatel570]

brother, i want to create deforum sd videos using colab. Please help me🙏, searched a lot but didn't get solution yet.

[@1littlecoder]

Please check camenduru colab you'd get it there

[@nisargpatel570]

@@1littlecoder Thanks for your response, I tried it earlier too, while running any stable diffusion notebook on free version of colab it gets "runtime disconnected" in every 7-10 minutes. Why does this happen and please give solution of it.

[@erickovalski7253]

In Model Download/Load I always have the same error: "FileNotFoundError: [Errno 2] No such file or directory: '/content/gdrive/MyDrive/sd/stable-diffusion-webui/models/Stable-diffusion/model.ckpt'". What am I doing wrong?

[@1littlecoder]

If you have some link in the model ckpt textbox please remove it. Just select v2 from the drop-down

[@mynuggets_oz779]

" with a piercing using a ring in her nose " AI wouldnt know what you mean by nose ring because its just an onject, you need to explain what it is, or a good trick is using a chat AI to help give better prompts

[@1littlecoder]

Thanks for sharing

[@SNRTHEROCK]

Good Video Brother. Automatic Eleven Eleven, not Automatic one one one one

[@1littlecoder]

Thank you :)

[@3pm479]

I am not getting Gradio or Local link ot run it, I am getting this message "LatentDiffusion: Running in eps-prediction mode DiffusionWrapper has 865.91 M params. ^C"

[@Nekocki]

me too

[@1littlecoder]

Disconnect and Reset runtime and connect again and run. I guess you hit memory issues

[@Nekocki]

@@1littlecoder not worked

[@1littlecoder]

Try this Colab - colab.research.google.com/drive/1RR7lDRItnFeCwtuQ6rXws5NkkHB9NW4W?usp=sharing

[@Planetgriham]

i have an error in running start stable diffusion FileNotFoundError Traceback (most recent call last) in 60 srv= getoutput('cat /content/srvr.txt') 61 ---> 62 for line in fileinput.input('/usr/local/lib/python3.7/dist-packages/gradio/blocks.py', inplace=True): 63 if line.strip().startswith('self.server_name ='): 64 line = f' self.server_name = "{srv[8:]}" ' 1 frames /usr/lib/python3.8/fileinput.py in _readline(self) 340 pass 341 # The next few lines may raise OSError --> 342 os.rename(self._filename, self._backupfilename) 343 self._file = open(self._backupfilename, self._mode) 344 try: FileNotFoundError: [Errno 2] No such file or directory: '/usr/local/lib/python3.7/dist-packages/gradio/blocks.py' -> '/usr/local/lib/python3.7/dist-packages/gradio/blocks.py.bak' what should i do...???

[@Faabaap1]

I'm experiencing the same problem

[@GodX7x]

Bro what is On Start stable diffusion user and password??

[@1littlecoder]

When you create your SD Ui, If you want to protect it with a user name and password. So no one brute force and access your link. Back then Gradio links were relatively easy to guess, so user name and password was a thing.

[@antonioanonimo2521]

we need the inpaiting model :(

[@FLEXTORGAMINGERA]

Yh man

[@map-creator]

It is asking for Huggingface token

[@1littlecoder]

Ideally it shouldn't. Did you select 1.5 or 2.0 model?

[@StelianPopa]

For me, doesn't work from the few days... Until now was perfect... but now I get this error running the last cell "Start Stable-Diffusion" Traceback (most recent call last): File "/content/gdrive/MyDrive/sd/stable-diffusion-webui/webui.py", line 7, in from fastapi import FastAPI ModuleNotFoundError: No module named 'fastapi' Can you please help me what to do...?

[@URUC-Official]

Thank you, as someone who also don't have the resources, I am excited every time you bring out a video with the web and colab versions for us free surfers :) P.S. You forgot to set your Resolution to 768, that is why the images looked a bit wonky at times :D I have found that by using the 768 model and leaving the resolution at 512 it gives strange results at times. thanks again!

[@1littlecoder]

Thanks a lot! That's very glad to know these efforts matter :)

[@muhammadzazulirizki1000]

Brother, I still don't really understand, so everytime I want to run the automatic1111 again (via google collab), I should re-run every line of code from the top? or the next time I want to do it, I just need to run the last part ("Start Stable-Diffusion")?

[@kashif7329]

I also wanted to know that

[@chromafresh]

@@kashif7329 Run all again from the top.

[@eleonoramarree]

it doesn't work. I downloaded your first url link followed your instruction. but running the last module it doesn't make a local tunnel url. this is the last text I get when running all modules and didn't get an error anywhere. LatentDiffusion: Running in v-prediction mode DiffusionWrapper has 865.91 M params. ^C

[@1littlecoder]

Where did you run that ?

[@eleonoramarree]

@@1littlecoder in colab save it to my drive in google. what am I doing wrong ?

[@hatuey6326]

@@1littlecoder same issue work the first time but not the second time : and same message

[@robbdeeze]

@@1littlecoder Im getting the same message below. Copied to drive, ran each cell. thanks in advance LatentDiffusion: Running in v-prediction mode DiffusionWrapper has 865.91 M params. Downloading: 100% 3.94G/3.94G [01:20

[@robbdeeze]

Eureka!!! I beleive i found the solution, atleast it worked for me. Delete all thye colab notebooks in your "colab folder " in your drive. Then follow the video. Good luck

[@-sint-]

Does not work. We need a new link so that we can run it in a collab with different models.

[@worldscutestpuppies3221]

hello sir is there any colab link that i can use sd latest version

[@1littlecoder]

Maybe Try this and see - kzbin.info/www/bejne/n56vg5afpJiiqbc&pp=gAQBiAQB

[@AIGMateYT]

I'm getting error from both side

[@RuslanGumbatov-px6qv]

How cool is that! It works! Amazing! Thank you very much!

[@Nekocki]

LatentDiffusion: Running in eps-prediction mode DiffusionWrapper has 865.91 M params. ^C can someone help me?

[@1littlecoder]

I think you're hitting OOM. Can you disconnect and delete runtime and connect again and run from the start

[@selianereinki3528]

@@1littlecoder I have the same problem, I did the procedure but it didn't solve it, not even deleting the folder in gdrive

[@1littlecoder]

Try this Colab - colab.research.google.com/drive/1RR7lDRItnFeCwtuQ6rXws5NkkHB9NW4W?usp=sharing

[@selianereinki3528]

@@1littlecoder FileNotFoundError Traceback (most recent call last) in 60 srv= getoutput('cat /content/srvr.txt') 61 ---> 62 for line in fileinput.input('/usr/local/lib/python3.7/dist-packages/gradio/blocks.py', inplace=True): 63 if line.strip().startswith('self.server_name ='): 64 line = f' self.server_name = "{srv[8:]}" ' 1 frames /usr/lib/python3.8/fileinput.py in _readline(self) 340 pass 341 # The next few lines may raise OSError --> 342 os.rename(self._filename, self._backupfilename) 343 self._file = open(self._backupfilename, self._mode) 344 try: FileNotFoundError: [Errno 2] No such file or directory: '/usr/local/lib/python3.7/dist-packages/gradio/blocks.py' -> '/usr/local/lib/python3.7/dist-packages/gradio/blocks.py.bak'

[@technonatorr]

I downloaded dreambooth extension from extensions tab, it's downloaded but I can't see as a tab on the UI. Any idea?

[@mixitnow1932]

How about using your own images(your face) as a reference? Will it do.?. in what folder shoud I put it? Tnx in advance

[@tenebriocron]

Thanks, good video!

[@bourehimyoussef111]

Thank you. The model downloaded to my google drive is 2.43 GB, is that stable diffusion 2 or 1 because version 1 was about 4 GB, and how do I add additional models.

[@1littlecoder]

Are you using automatic 1111? Is it on colab or local?

[@sunilsamantray37]

Bro Start Stable-Diffusion step takes too long to complete in Co Lab is this OK?

[@1littlecoder]

Can you please try the latest The Last Ben colab?

[@akadekki]

FileNotFoundError Traceback (most recent call last) in 60 srv= getoutput('cat /content/srvr.txt') 61 ---> 62 for line in fileinput.input('/usr/local/lib/python3.7/dist-packages/gradio/blocks.py', inplace=True): 63 if line.strip().startswith('self.server_name ='): 64 line = f' self.server_name = "{srv[8:]}" ' 1 frames /usr/lib/python3.8/fileinput.py in _readline(self) 340 pass 341 # The next few lines may raise OSError --> 342 os.rename(self._filename, self._backupfilename) 343 self._file = open(self._backupfilename, self._mode) 344 try: FileNotFoundError: [Errno 2] No such file or directory: '/usr/local/lib/python3.7/dist-packages/gradio/blocks.py' -> '/usr/local/lib/python3.7/dist-packages/gradio/blocks.py.bak'

[@AlanHetfield14]

Same

[@1littlecoder]

That's strange. Can you share the colab link?

[@jonjoni518]

@@1littlecoder the notebook needs to be modified because colab has upgraded to python 3.8

[@cybermollusk]

Thank you for this video. It was a great, clear tutorial. One question I have; when you sent the image to inpainting, you put double parenthesis around the original prompt. What does that do?

[@GS-ef5ht]

Parenthesis adds emphasis to that ((((portion of the prompt)))) you can use multiple () also depending on how much importance you want to add

[@Peppermint_juice]

I didn't got a local Tunnel url, what i do

[@1littlecoder]

Try the colab from the pinned comment

[@LoneRanger.801]

Can we open fast SD in Jupyter/conda localy?

[@1littlecoder]

Technically it should work if you have an NVIDIA GPU

[@AlanHetfield14]

HELP! It doesn't work, I see the error: "LatentDiffusion: Running in eps-prediction mode DiffusionWrapper has 865.91 M params. Downloading100% 3.94G/3.94G [01:39

[@lazaroht8246]

me too

[@Peppermint_juice]

Same

[@1littlecoder]

Try this Colab -- colab.research.google.com/drive/1RR7lDRItnFeCwtuQ6rXws5NkkHB9NW4W?usp=sharing

[@AlanHetfield14]

@@1littlecoder Hello! With that colab I see another error. Sorry for the inconvenience. FileNotFoundError: [Errno 2] No such file or directory: '/usr/local/lib/python3.7/dist-packages/gradio/blocks.py' -> '/usr/local/lib/python3.7/dist-packages/gradio/blocks.py.bak' Entire error: FileNotFoundError Traceback (most recent call last) in 60 srv= getoutput('cat /content/srvr.txt') 61 ---> 62 for line in fileinput.input('/usr/local/lib/python3.7/dist-packages/gradio/blocks.py', inplace=True): 63 if line.strip().startswith('self.server_name ='): 64 line = f' self.server_name = "{srv[8:]}" ' 1 frames /usr/lib/python3.8/fileinput.py in _readline(self) 340 pass 341 # The next few lines may raise OSError --> 342 os.rename(self._filename, self._backupfilename) 343 self._file = open(self._backupfilename, self._mode) 344 try: FileNotFoundError: [Errno 2] No such file or directory: '/usr/local/lib/python3.7/dist-packages/gradio/blocks.py' -> '/usr/local/lib/python3.7/dist-packages/gradio/blocks.py.bak'

[@tonytitani410]

Can you use a custom trained ckpt model on any version of stable diffusion?

[@1littlecoder]

Nope. Base version should be matched

[@viveks.]

How to use our custom trained dreambooth model with stable diffusion 2

[@1littlecoder]

You can add that model to your gdrive and paste that path in the ckpt text field

[@viveks.]

@@1littlecoder so doing this would run my model on SD 2? I have trained my model few weeks ago. (Apologies if this sounds like fool, I am new here)

[@1littlecoder]

Was the model trained from sd 2 base model or 1.4/1.5? If it's trained from sd v2 then yes

[@viveks.]

@@1littlecoder was trained on 1.5 I guess. How do I train it on SD 2?

[@1littlecoder]

kzbin.info/www/bejne/fZybeXVrpc52htU This should guide you, but I didn't get good results. Try with a better learning rate like 4e-6

[@Ubanhistory]

from google.colab import drive drive.mount('/content/gdrive') ce code se trouvant sur le colab est trop suspect