Will there be a Windows version? #125
Replies: 12 comments
-
@EugeoSynthesisThirtyTwo I am looking to run the extension locally as well. I was able to use the webUI to install this repo as an extension via url. The app didin't show up anywhere in the UI, should it show up as a tab or something? That part may not have been right. In any case, I installed jupiter lab in an effort to run in there. I am too new with jupiter to make any sense of what was going on. Maybe you can take a look and connect the dots. Install as an extension vie automatic1111 webui
One in jupiter I tried running all cells, but it was a mess. |
Beta Was this translation helpful? Give feedback.
-
I recommend WSL, even though it's still not really plug and play |
Beta Was this translation helpful? Give feedback.
-
If you want to run it locally on windows, why would you not use https://github.com/bmaltais/kohya_ss
This is a colab repo..
…________________________________
From: EugeoSynthesisThirtyTwo ***@***.***>
Sent: Saturday, June 3, 2023 6:48:09 PM
To: hollowstrawberry/kohya-colab ***@***.***>
Cc: Subscribed ***@***.***>
Subject: [hollowstrawberry/kohya-colab] Will there be a windows version ? (Issue #37)
I don't want to run the training on google colab, instead I use jupyter to use my own GPU power, so I set COLAB=False
I think having this option is a very good idea but everything is made for linux, the code is using linux commands (apt/sed) and the code is overly complicated for my usage. I just want something that works on jupyter and I end up spending hours understanding and modifying everything and it still doesn't work
Could you please make an option to choose between linux and windows please ?
—
Reply to this email directly, view it on GitHub<#37>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AIR4DCQ62XSPQAUUNJA3NXLXJN2FTANCNFSM6AAAAAAYZNJ7UQ>.
You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
Ok I will try both WSL and kohya_ss's project, thank you |
Beta Was this translation helpful? Give feedback.
-
For the wsl option, would that be to install jupiter under the new environment? |
Beta Was this translation helpful? Give feedback.
-
I tried both WSL and kohya_ss's project and none of these solutions work, I spent hours trying to make it work and there are just too many errors to be listed here |
Beta Was this translation helpful? Give feedback.
-
I would now recommend this new training GUI by my friend. https://github.com/derrian-distro/LoRA_Easy_Training_Scripts As for why I haven't made a Windows version of the colab, I tried, but too many parts depend on Linux commands as well as specific dependencies... |
Beta Was this translation helpful? Give feedback.
-
I Opt in for a Windows jupyter version that uses local folders instead of google drive. @hollowstrawberry If you are interessted i can give you the converted code. As its states now: It works ^^ |
Beta Was this translation helpful? Give feedback.
-
Hey @AsaTyr2018 I'm wanting to make a hybrid of this. I'd like to use fiftyone and do all dataset processing locally on Win10, then let Colab do the heavy lifting/training. I just forked this, so I haven't had a chance to look through the code yet. Just curious, would your modifications help me jump start this process at all? It sounds like it would, so you're willing to share, if you'd pull request here (or directly to my fork or the OP's fork, if hollowstrawberry doesn't want/isn't interested) I'd appreciate the head start! Thx. |
Beta Was this translation helpful? Give feedback.
-
@aleph23 I also work on a small "helping hand" tool that helps me with all the training. currently it use the native python gui. and ok. in vision of yours its just a small bug xD but hey anything that helps huge projects =) Its here: https://github.com/AsaTyr2018/Dataset-Helper |
Beta Was this translation helpful? Give feedback.
-
There is also the Extension version of Kohya_ss for A1111. I opted to use that, but my GPU uses shared memory so the training was way too slow for my liking. Still a very good alternative for those looking to train locally and who have A1111 for their gen platform. |
Beta Was this translation helpful? Give feedback.
-
I used this as well when I was trying to run my training locally. Very easy to use. Just Shared RAM for my GPU is the issue for me. |
Beta Was this translation helpful? Give feedback.
-
I don't want to run the training on google colab, instead I use jupyter to use my own GPU power, so I set COLAB=False
I think having this option is a very good idea but everything is made for linux, the code is using linux commands (apt/sed) and the code is overly complicated for my usage. I just want something that works on jupyter and I end up spending hours understanding and modifying everything and it still doesn't work
Could you please make an option to choose between linux and windows please ?
Beta Was this translation helpful? Give feedback.
All reactions