Comments (21)
It seems that too many users are trying to use the colab. Currently looking into a solution to Google's rate limits
from diffusion-colabui.
A possible workaround if you have good internet is to download the file manually and upload it into your own Google Drive under /AI_models/novelai.zip
from diffusion-colabui.
The Colab will detect that and load it from your own drive
from diffusion-colabui.
You could also just copy the file to your own drive so you don't have to download then upload it.
from diffusion-colabui.
I'm aware that i can download it manually and upload it onto my drive, but my internet's way to slow for that. I've managed to work around it by changing the code up a bit and downloading the model from another one of my drives, long story short I couldn't use the google account I've used before, so I shared it model from there and used it on the other account. I'm more concerned for other people who don't have the model anywhere, that's why i made the issue
from diffusion-colabui.
You could also just copy the file to your own drive so you don't have to download then upload it.
Ah yes. I don't use drive much so I was unaware
from diffusion-colabui.
I'm aware that i can download it manually and upload it onto my drive, but my internet's way to slow for that. I've managed to work around it by changing the code up a bit and downloading the model from another one of my drives, long story short I couldn't use the google account I've used before, so I shared it model from there and used it on the other account. I'm more concerned for other people who don't have the model anywhere, that's why i made the issue
Perhaps we can compile a list of shared drive links to alternate between if one isn't working. I have no budget to host the model myself.
from diffusion-colabui.
I am temporairly resolving this by using a copy in another account
from diffusion-colabui.
Perhaps we can compile a list of shared drive links to alternate between if one isn't working. I have no budget to host the model myself.
Yeah, we could make some dummy accounts to host the model on. As i said earlier, my internet is way to slow to upload larger files anywhere, but one could probably exploit the script in the colab the same way as i did to get it onto the drives at high speeds. Alternatively, there are some free hosting services which also ignore DMCA's, but i have little to no idea on how does the code work, so i have no clue if its possible to work with non-google hosting
from diffusion-colabui.
Alternatively, there are some free hosting services which also ignore DMCA's
The issue is that the file is too large to upload onto those services
so i have no clue if its possible to work with non-google hosting
I have tried hosting the model on my own server with 1TB transfer limit. However, it was hit by a DDOS attack and now disabled (Payment wasn't connected or else I would've went bankrupt). The speeds are extremely slow outside Google or GitHub (10MB/s).
from diffusion-colabui.
It seems my student account has a Gsuite subscription provided by the school. That should increase the transfer limits by quite a lot. It's an old account and has no limits on storage.
from diffusion-colabui.
Alternatively, there are some free hosting services which also ignore DMCA's
The issue is that the file is too large to upload onto those services
so i have no clue if its possible to work with non-google hosting
I have tried hosting the model on my own server with 1TB transfer limit. However, it was hit by a DDOS attack and now disabled (Payment wasn't connected or else I would've went bankrupt). The speeds are extremely slow outside Google or GitHub (10MB/s).
I guess that's a lost cause then.
It seems my student account has a Gsuite subscription provided by the school. That should increase the transfer limits by quite a lot. It's an old account and has no limits on storage.
That's nice, hopefully it will be sufficient. If not, then dummy gdrive accounts are the only good option i can think of, if it will be needed I can make some. It's a good thing google doesn't go after dummies.
from diffusion-colabui.
I will reopen this if the error occurs again
from diffusion-colabui.
It seems gsuite is also limited
from diffusion-colabui.
The daily users has reached over ~500 which is way beyond the transfer limits
from diffusion-colabui.
Does anyone know of a reliable hosting service? I'm thinking IPFS but web3 isn't something I'm familiar with
from diffusion-colabui.
Does anyone know of a reliable hosting service? I'm thinking IPFS but web3 isn't something I'm familiar with
What about dummy GDrive accounts? It might be pretty tedious for the user to switch them around and look for one that's working, but according to some research the download limit per file is 2TB a day, so with the current amount of users being over 500 just a couple accounts would be sufficient.
from diffusion-colabui.
Well, it seems 500 was exceeded. I've been alternating between 5 accounts manually and all of them have been limited. Currently looking into hiding it in a huggingface repo
from diffusion-colabui.
Damn. The bandwidth is now running out minutes after update
from diffusion-colabui.
Well, it seems 500 was exceeded. I've been alternating between 5 accounts manually and all of them have been limited. Currently looking into hiding it in a huggingface repo
A huggingface repo would probably make sense
Damn. The bandwidth is now running out minutes after update
Oh lord, that's quick. You could try mega.nz, from what i know it doesn't limit the files directly, but the person downloading. The colab would probably be counted as a single user, but perhaps it could be worked around by requiring the user to log into their Mega account, if that's even possible?
from diffusion-colabui.
I have uploaded it onto huggingface. I'm not sure how long it'll take them to notice that it's pirated since it's zipped
from diffusion-colabui.
Related Issues (20)
- Hello, can you please update new plug-ins such as controlnet, thank you very much
- error HOT 8
- Ask HOT 7
- When switching models the code crashes
- PyTorch and torchvision were compiled with different CUDA versions HOT 1
- save_to_gdrive option leads to unrecognised arguments HOT 1
- Run In GDrive removed from "old" version HOT 1
- train.py issue
- Where is LorA menu HOT 1
- Problems with "conection errored out" HOT 2
- Excuse me, would it be possible to update some useful plugins? Thank you very much. HOT 3
- AttributeError: 'Options' object has no attribute 'extra_networks_card_height' HOT 1
- Error when running gradio HOT 12
- Hosting question HOT 1
- Namerror HOT 2
- Disconnecting itself HOT 9
- Disconnect issue
- Google Colab copy can't use extensions
- Civit Ai fix
- There's no save everything locally to GDrive like before? HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from diffusion-colabui.