Home
Creators
Search Recent Missing Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Register Favorites Keys Review DMs Logout
Community
Telegram ThePornDude
Home Creators Posts Import Register Favorites Logout

[Early Access] 🔥Colab🔥 Stable Diffusion InvokeAI Ultimate Google Drive Version (Patreon)

Published:
2023-10-11 19:38:30
Imported:
2026-06
Tags:
haven't archived this post yet. have a subscription? use the importer!

Content

If you are a newcomer, you may be overwhelmed by the vast number of notebooks and updates. You can simply use the latest version

____________________

Description

I've created a Google Colab notebook that will allow you to run Stable Diffusion without a paid subscription and any disconnection issues. As usual, this Colab notebook has practically no limitations.

You can download LoRAs, Checkpoints, Embeddings, ControlNet, VAEs, and more. Despite a free version, this notebook automatically downloads all models and extensions, saves your generations on your Google Drive, and even remembers your workflow.
Please, watch the tutorial for Invoke UI before using it:
https://youtu.be/lmpAopZ41K4


Updates (Archive)

upd1: if the first notebook doesn't work try this:
colab.research.google.com/drive/1i35vMxbA4bBPsWc2RwSs1MVDyCD35749
upd2: The issue with the displayed generation results is a consequence of a library update, and it's not related to the notebook. In the next few days, the developers of InvokeAI will fix everything, and I will update the notebook immediately. At the moment, to see the generation results, simply refresh the web page.
upd3: At the moment, I have managed to run Invoke, and at first glance, everything seems to be working without issues. However, it requires thorough testing. The link to the new notebook is available bellow. To work with it, you need to delete the previous version of Invoke from Google Drive and, if possible, the 'databases' folder with the models. If you already have many models in the folder, you can temporarily move them to another directory and then move them back later. And yes, there's good news - this is the latest version of Invoke with all the new features.🫡
Cloud version also updated and works properly
colab.research.google.com/drive/1i35vMxbA4bBPsWc2RwSs1MVDyCD35749
upd4: In the latest version of Invoke, the upscaler isn't working. This is due to Invoke itself, which have not been fixed by the developers yet. At the same time, I have made it possible to run the old (initial) version of Invoke where everything works. The link is available below
To work with it, you need to delete the previous version of Invoke from Google Drive and, if possible, the 'databases' folder with the models. If you already have many models in the folder, you can temporarily move them to another directory and then move them back later.
https://colab.research.google.com/drive/1JEoNBppgFrCItK2pzcA3L5B4tNx_i2NJ
upd5:The issue has been resolved, and the notebook has been updated. The link didn't appear due to issues with loca.lt servers, which happen very rarely. You could simply wait for them to become available again, but I've added a more reliable alternative, ngrok.com. You will need to register on the website and insert your token into the notebook. It's a bit more involved but definitely more reliable. When loca.lt becomes available again, you can use it once more, as there's not much difference between them.
🔥 https://colab.research.google.com/drive/17Szgistw41GBjLuQ-ZjXYGtKePCgXs4V

____________________

OVERALL (Updates regularly)

Currently, there are two versions of Invoke available:

🔗 Latest InvokeUI (unstable, at least upscalers don't work): colab.research.google.com/drive/1i35vMxbA4bBPsWc2RwSs1MVDyCD35749

🔥 Stable Version:  https://colab.research.google.com/drive/17Szgistw41GBjLuQ-ZjXYGtKePCgXs4V

P.S. Guys, this is the most troublesome notebook I've ever worked on. It's amazing, but something breaks every time, and it's often for reasons beyond my control. I apologize for any inconvenience; I'm trying to update everything as quickly as possible🫡

Files

Previews only

Google Colaboratory

Comments

Syntax

Unfortunately I get this in Step 3: /content/drive/MyDrive/InvokeAI Warning: Identity file /content/drive/MyDrive/id_rsa not accessible: No such file or directory. Warning: Permanently added 'remote.moe' (ED25519) to the list of known hosts. root __ .----.-----.--------.-----.| |_.-----.--------.-----.-----. | _| -__| | _ || _| -__| | _ | -__| |__| |_____|__|__|__|_____||____|_____|__|__|__|_____|_____| You somehow forgot to present a public key when trying to authenticate. Please continue by creating one: ssh-keygen remotemoe accepts any key - see ya! Received disconnect from 159.69.126.209 port 22:2: too many authentication failures Disconnected from 159.69.126.209 port 22 2023-10-13 08:21:21.934048: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT >> patchmatch.patch_match: INFO - Compiling and loading c extensions from "/usr/local/lib/python3.10/dist-packages/patchmatch". >> patchmatch.patch_match: WARNING - patchmatch failed to load or compile. >> patchmatch.patch_match: WARNING - Refer to https://github.com/invoke-ai/InvokeAI/blob/main/docs/installation/INSTALL_PATCHMATCH.md for installation instructions. [2023-10-13 08:21:30,580]::[InvokeAI]::INFO --> Patchmatch not loaded (nonfatal) /usr/local/lib/python3.10/dist-packages/torchvision/transforms/functional_tensor.py:5: UserWarning: The torchvision.transforms.functional_tensor module is deprecated in 0.15 and will be **removed in 0.17**. Please don't rely on it. You probably just need to use APIs in torchvision.transforms.functional or in torchvision.transforms.v2.functional. warnings.warn( An exception has occurred: /content/drive/MyDrive/db/configs/models.yaml not found == STARTUP ABORTED == ** One or more necessary files is missing from your InvokeAI root directory ** ** Please rerun the configuration script to fix this problem. ** ** From the launcher, selection option [7]. ** ** From the command line, activate the virtual environment and run "invokeai-configure --yes --skip-sd-weights" ** ** (To skip this check completely, add "--ignore_missing_core_models" to your CLI args. Not installing these core models will prevent the loading of some or all .safetensors and .ckpt files. However, you can always come back and install these core models in the future.) Press any key to continue...

marat_ai
>>119702627

Hi! I've just tested and it works. Please check your runtime according to the instructions; it should be set to T4. If the issue persists, try deleting your 'invoke' folder on Google Drive and install everything once again.

Baal

Hello I did everything like you showed in video, But I get this in third step

Baal

/content/drive/MyDrive/InvokeAI Warning: Identity file /content/drive/MyDrive/id_rsa not accessible: No such file or directory. Warning: Permanently added 'remote.moe' (ED25519) to the list of known hosts. root __ .----.-----.--------.-----.| |_.-----.--------.-----.-----. | _| -__| | _ || _| -__| | _ | -__| |__| |_____|__|__|__|_____||____|_____|__|__|__|_____|_____| You somehow forgot to present a public key when trying to authenticate. Please continue by creating one: ssh-keygen remotemoe accepts any key - see ya! Received disconnect from 159.69.126.209 port 22:2: too many authentication failures Disconnected from 159.69.126.209 port 22 2023-10-13 15:34:28.612956: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT >> patchmatch.patch_match: INFO - Compiling and loading c extensions from "/usr/local/lib/python3.10/dist-packages/patchmatch". >> patchmatch.patch_match: WARNING - patchmatch failed to load or compile. >> patchmatch.patch_match: WARNING - Refer to https://github.com/invoke-ai/InvokeAI/blob/main/docs/installation/INSTALL_PATCHMATCH.md for installation instructions. [2023-10-13 15:34:52,331]::[InvokeAI]::INFO --> Patchmatch not loaded (nonfatal) /usr/local/lib/python3.10/dist-packages/torchvision/transforms/functional_tensor.py:5: UserWarning: The torchvision.transforms.functional_tensor module is deprecated in 0.15 and will be **removed in 0.17**. Please don't rely on it. You probably just need to use APIs in torchvision.transforms.functional or in torchvision.transforms.v2.functional. warnings.warn( [2023-10-13 15:35:03,906]::[uvicorn.error]::INFO --> Started server process [4234] [2023-10-13 15:35:03,906]::[uvicorn.error]::INFO --> Waiting for application startup. [2023-10-13 15:35:03,907]::[InvokeAI]::INFO --> InvokeAI version 3.1.0 [2023-10-13 15:35:03,907]::[InvokeAI]::INFO --> Root directory = /content/drive/MyDrive/db [2023-10-13 15:35:04,092]::[InvokeAI]::INFO --> GPU device = cuda Tesla T4 [2023-10-13 15:35:04,099]::[InvokeAI]::INFO --> Scanning /content/drive/MyDrive/db/models for new models [2023-10-13 15:35:04,467]::[InvokeAI]::INFO --> Scanned 6 files and directories, imported 0 models [2023-10-13 15:35:04,478]::[InvokeAI]::INFO --> Model manager service initialized [2023-10-13 15:35:04,511]::[uvicorn.error]::INFO --> Application startup complete. [2023-10-13 15:35:04,512]::[uvicorn.error]::INFO --> Uvicorn running on http://127.0.0.1:9090 (Press CTRL+C to quit)

starwolf 27
>>119702627

I'm having this problem too, and deleting the InvokeAI folder on Google Drive and reinstalling didn't do anything.

Syntax
>>119724324

Please let us know how you fixed it? I am still having the same issue despite any attempts to solve it.

Syntax
>>119724324

Didnt change anything for me. Luckily the Public version works fine for me, so for now I can use that.

天天 80分

me too,Warning: Identity file /content/drive/MyDrive/id_rsa not accessible: No such file or directory.

marat_ai
>>119841125

try this: colab.research.google.com/drive/1i35vMxbA4bBPsWc2RwSs1MVDyCD35749?usp=sharing

天天 80分

--------------------------------------------------------------------------- CalledProcessError Traceback (most recent call last) in () 27 ssh_name = "id_rsa" 28 ssh_path = Path(os.path.dirname(os.getcwd())) / ssh_name ---> 29 gen_key(ssh_path) 30 clear_output() 31 1 frames /usr/lib/python3.10/subprocess.py in run(input, capture_output, timeout, check, *popenargs, **kwargs) 524 retcode = process.poll() 525 if check and retcode: --> 526 raise CalledProcessError(retcode, process.args, 527 output=stdout, stderr=stderr) 528 return CompletedProcess(process.args, retcode, stdout, stderr) CalledProcessError: Command '['ssh-keygen', '-t', 'rsa', '-b', '4096', '-N', '', '-q', '-f', '/content/id_rsa']' returned non-zero exit status 1.

Jujubiee

I'm not sure what I did but I can get all through the steps but now when I use the "password" Invokeai will get stuck on a black screen and never load. I cleared my history, removed the files from my google drive, restarted my laptop but no luck.

marat_ai
>>119930712

Which version of the notebook did you use? And what do you mean by "password"? If it's the IP, sometimes the loc.lt itself may be inaccessible, possibly even in your country. In that case, a VPN might help. In the case of using the first version, it shouldn't be like that, but for some reason, it doesn't work for some people.

Jujubiee
>>119930712

I'm using the upd1 link for the SD_InvokeAI_Ultimate_GD_ver_v3(loca.lt).ipynb version because the 1st version didn't work for me at all. Yes i'm sorry I meant the IP isn't loading anymore for me, it was working at one point. I'm currently using a VPN i'll try switching it's location. Thank you for all the help.

Jujubiee
>>119930712

I guess I didn't clear my google drive as well as I thought. I got everything working. Thanks again!

marat_ai
>>119930712

I'm glad that it works! You are welcome! Feel free to write here or in direct messages if you have any issues.

Abdullah Hasan

The notebook works fine but the generated file doesn't show up in the UI. I have to manually search the output in my drive. And keep pressing cancel when generating image, since the UI doesn't know the image has been generated. Any suggestions how to fix it?

Abdullah Hasan

Also I don't understand coding but is there any way to skip the first cell after we already used it? I don't know why skipping to the 3rd cell after I already download everything still turns into error

marat_ai
>>119995730

The issue with the displayed generation results is a consequence of a library update, and it's not related to the notebook. In the next few days, the developers of InvokeAI will fix everything, and I will update the notebook immediately. At the moment, to see the generation results, simply refresh the web page.

marat_ai
>>119995836

You still need to install dependencies and libraries; Colab removes them every time. It used to be different before

marat_ai
>>120028912

Thanks! I've explained the issue in the post update. I've just updated the notebook, and you can check it out. All the needed information is in the post updates.

marat_ai
>>119995730

I've just updated the notebook, and you can check it out. All the necessary information is in the post-updates. As for your question, I don't know. I'm just checking their GitHub.

Abdullah Hasan
>>119995730

Thanks for the update! Now the generated image is not an issue in UI. It's a bummer that Invoke cannot use Roop though

Syntax

I am experiencing "connection closed" (in the Step 3 runtime) a lot. It seemingly happens at random intervals during use. Anyone else having this issue?

marat_ai
>>120153392

I don't have it yet, could you please send the full error code if the error happens again?

CharlieJ0rdan

The upscaler does not even working.

Federico Cassetta

Does it now work to display images without updating the webUI?

marat_ai
>>120283548

If by 'updating the web UI' you meant refreshing the webpage, then yes, this issue has been resolved. Just use the latest version of the notebook

CharlieJ0rdan

At update 4: At step 3: The localtunnel link is missing. It was the link where you have to paste the password/enpoint

Preston

At update 5: At step 3: No final link provided to webUI /tools/node/bin/lt -> /tools/node/lib/node_modules/localtunnel/bin/lt.js + localtunnel@2.0.2 added 22 packages from 22 contributors in 1.964s /content/drive/MyDrive/InvokeAI 2023-10-21 12:10:15.369257: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT >> patchmatch.patch_match: INFO - Compiling and loading c extensions from "/usr/local/lib/python3.10/dist-packages/patchmatch". >> patchmatch.patch_match: WARNING - patchmatch failed to load or compile. >> patchmatch.patch_match: WARNING - Refer to https://github.com/invoke-ai/InvokeAI/blob/main/docs/installation/INSTALL_PATCHMATCH.md for installation instructions. [2023-10-21 12:10:29,918]::[InvokeAI]::INFO --> Patchmatch not loaded (nonfatal) /usr/local/lib/python3.10/dist-packages/torchvision/transforms/functional_tensor.py:5: UserWarning: The torchvision.transforms.functional_tensor module is deprecated in 0.15 and will be **removed in 0.17**. Please don't rely on it. You probably just need to use APIs in torchvision.transforms.functional or in torchvision.transforms.v2.functional. warnings.warn( [2023-10-21 12:10:37,016]::[uvicorn.error]::INFO --> Started server process [3491] [2023-10-21 12:10:37,017]::[uvicorn.error]::INFO --> Waiting for application startup. [2023-10-21 12:10:37,017]::[InvokeAI]::INFO --> InvokeAI version 3.1.0 [2023-10-21 12:10:37,017]::[InvokeAI]::INFO --> Root directory = /content/drive/MyDrive/db [2023-10-21 12:10:37,389]::[InvokeAI]::INFO --> GPU device = cuda Tesla T4 [2023-10-21 12:10:37,396]::[InvokeAI]::INFO --> Scanning /content/drive/MyDrive/db/models for new models [2023-10-21 12:10:37,815]::[InvokeAI]::INFO --> Scanned 6 files and directories, imported 0 models [2023-10-21 12:10:37,826]::[InvokeAI]::INFO --> Model manager service initialized [2023-10-21 12:10:37,857]::[uvicorn.error]::INFO --> Application startup complete. [2023-10-21 12:10:37,857]::[uvicorn.error]::INFO --> Uvicorn running on http://127.0.0.1:9090 (Press CTRL+C to quit) \Invoke finished loading, trying to launch localtunnel The password/enpoint ip for localtunnel is: _________________________________________________________ 130.211.113.5

Atelier
>>120316229

Hi, i still got the same issue from the comment above. im using the local tunnel on SD_InvokeAI_Ultimate_GD_ver_v6(loca.lt, ngrok).ipynb

marat_ai
>>120316229

Hi! As I mentioned, it's happening because of loca.lt servers. In this case, I added ngrok, which works much more reliably. For more details just check the latest update

Henry Wirawan

The link is not created, please help, thanks /content/drive/MyDrive/InvokeAI Warning: Identity file /content/drive/MyDrive/id_rsa not accessible: No such file or directory. root __ .----.-----.--------.-----.| |_.-----.--------.-----.-----. | _| -__| | _ || _| -__| | _ | -__| |__| |_____|__|__|__|_____||____|_____|__|__|__|_____|_____| You somehow forgot to present a public key when trying to authenticate. Please continue by creating one: ssh-keygen remotemoe accepts any key - see ya! Received disconnect from 159.69.126.209 port 22:2: too many authentication failures Disconnected from 159.69.126.209 port 22 2023-10-27 09:59:51.846536: E tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:9342] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered 2023-10-27 09:59:51.846601: E tensorflow/compiler/xla/stream_executor/cuda/cuda_fft.cc:609] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered 2023-10-27 09:59:51.851673: E tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:1518] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered 2023-10-27 09:59:54.940842: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT >> patchmatch.patch_match: INFO - Compiling and loading c extensions from "/usr/local/lib/python3.10/dist-packages/patchmatch". >> patchmatch.patch_match: WARNING - patchmatch failed to load or compile. >> patchmatch.patch_match: WARNING - Refer to https://github.com/invoke-ai/InvokeAI/blob/main/docs/installation/INSTALL_PATCHMATCH.md for installation instructions. [2023-10-27 10:00:52,181]::[InvokeAI]::INFO --> Patchmatch not loaded (nonfatal) /usr/local/lib/python3.10/dist-packages/torchvision/transforms/functional_tensor.py:5: UserWarning: The torchvision.transforms.functional_tensor module is deprecated in 0.15 and will be **removed in 0.17**. Please don't rely on it. You probably just need to use APIs in torchvision.transforms.functional or in torchvision.transforms.v2.functional. warnings.warn( [2023-10-27 10:01:49,179]::[uvicorn.error]::INFO --> Started server process [7897] [2023-10-27 10:01:49,179]::[uvicorn.error]::INFO --> Waiting for application startup. [2023-10-27 10:01:49,179]::[InvokeAI]::INFO --> InvokeAI version 3.3.0 [2023-10-27 10:01:49,180]::[InvokeAI]::INFO --> Root directory = /content/drive/MyDrive/db [2023-10-27 10:01:49,181]::[InvokeAI]::INFO --> Using database at /content/drive/MyDrive/db/databases/invokeai.db [2023-10-27 10:01:49,385]::[InvokeAI]::INFO --> GPU device = cuda Tesla T4 [2023-10-27 10:01:49,393]::[InvokeAI]::INFO --> Scanning /content/drive/MyDrive/db/models for new models [2023-10-27 10:01:54,847]::[InvokeAI]::INFO --> Scanned 9 files and directories, imported 0 models [2023-10-27 10:01:54,856]::[InvokeAI]::INFO --> Model manager service initialized [2023-10-27 10:01:54,965]::[InvokeAI]::INFO --> Pruned 0 finished queue items [2023-10-27 10:01:54,982]::[InvokeAI]::INFO --> Cleaned database [2023-10-27 10:01:54,982]::[uvicorn.error]::INFO --> Application startup complete. [2023-10-27 10:01:54,983]::[uvicorn.error]::INFO --> Uvicorn running on http://127.0.0.1:9090 (Press CTRL+C to quit)

marat_ai
>>120761985

This method is not reliable. Recently, loca.lt had problems with servers. Please use ngrok instead, which is available in the latest versions of notebooks

Henry Wirawan

Hi, I'm new to stable diffusion, I just learn that I need to install ADetailer extension to generate image with normal face, but I can't find that option on invokeAI

marat_ai
>>120898499

That extension is only for Automatic, which you can use via Kaggle. I also have a notebook for this check previous posts. In case of Unvoke, custom workflows could probably help

Syntax

I am now gettting following issue/ error when following the generated link in Step 3, Option 2 -ngrok Tunnel: ERR_NGROK_8012 Traffic was successfully tunneled to the ngrok agent, but the agent failed to establish a connection to the upstream web service at localhost:9090. The error encountered was: dial tcp 127.0.0.1:9090: connect: connection refused

Maykit Happin

I'm getting this error code when I am trying to edit the second step for the models and checkpoints. Do I need to edit these or is everything taken care of? "Could Not Fetch /Content/Invokeal/Invokeai/Configs/Initial_Models.Yaml From Backend X"

marat_ai
>>121778031

Hi! What version of the notebook are you using? If an error occurs, it's not a problem, the main thing is that the UI loads, additional models can be loaded through it.

Syntax

Hi marat_ai. The v6 notebook enables the use of InvokeAi 3.1.0. Is there any possibilty to update the notebook in order to use the latest version of InvokeAi 3.4?

marat_ai
>>123093065

Hi! Yes, sure, I think I could do that and update to the latest version. Just give me a little time.

marat_ai
>>123093065

I tried to update, but there is a collision with libraries. So now I need more time :)

Jujubiee

I keep running into ERR_NGROK_8012... Traffic was successfully tunneled to the ngrok agent, but the agent failed to establish a connection to the upstream web service at localhost:9090. The error encountered was: dial tcp 127.0.0.1:9090: connect: connection refused.