Welcome to sdupdates Discussions! #1
Replies: 7 comments 16 replies
-
Beta Was this translation helpful? Give feedback.
-
|
Thank you for all the kind words 😄 My only socials for right now are listed in my bio, but I'll probably end up making more. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Yea definitely, I'll add it right now. Thanks for the link and training info |
Beta Was this translation helpful? Give feedback.
-
|
Hi, I uploaded AnythingV3.0.ckpt (Chinese NAI Resume Training Model) to Huggingface, I also pruned it to run on 4G. Do you think you could add it to the list? In this repository there are 3 files, unpruned model, pruned model, and VAE Maybe pickled, I'm still learning how to unpickle it. Here's a comparison between using animefull-finalpruned.ckpt vs AnythingV3.0-pruned.ckpt, I'm using latest pull of Automatic1111's Stable Diffusion Web UI to generate both images.
Thank you! |
Beta Was this translation helpful? Give feedback.
-
Where was the source for that one? Was it in the original NAI leak? |
Beta Was this translation helpful? Give feedback.
-
|
The left one is from the original nai leak, pruned model. The right one is on sdupdates2#newsfeed today, 11/8+11/7, looks like the chinese community did resume training on the original nai leak so you can get more detailed results. |
Beta Was this translation helpful? Give feedback.
-
Is there a link to the original post? I can read Chinese and I'm interested in this project, but can't find the OP, maybe the name is too generic. |
Beta Was this translation helpful? Give feedback.
-
|
https://www.bilibili.com/read/cv19603218 |
Beta Was this translation helpful? Give feedback.
-
|
Hi, thanks for making this, I'll add it in right now |
Beta Was this translation helpful? Give feedback.
-
|
Hello, can we maybe organize a Google Drive repo for the usual models such as SD, WD, etc? Not sure about the licenses and permissions though. In Google Drive, we can add a shortcut of someone's file into our drive, and it won't take space. The shortcut can be used in Google Colab just like normal file. I was doing this with SD 1.4 but SD 1.5 doesn't have a Google Drive link. |
Beta Was this translation helpful? Give feedback.
-
|
Hi, yea that sounds like a good idea for people who want to use colab but don't want to wait to upload the model every time. It also gives people another mirror to download from if they dislike torrenting/mega/whatever other mirrors there are. The only issue might be takedowns? (like NAI), though I'm not too sure how that works so take this with a grain of salt. I'll be sure to add in the repo if someone creates it. If no one wants to, I can create it sometime in the future |
Beta Was this translation helpful? Give feedback.
-
|
Just found this when surfing on twitter, maybe it's what we called English Grimoire, I think it's useful for people who don't understand Japanese, Chinese, and Korean spellbook https://lunarmimi.net/freebies/novelai-anime-girl-prompt-guide/ |
Beta Was this translation helpful? Give feedback.
-
|
Sorry for the late reply, I was busy for the past few days. I'll add it in, but the link seems to have died. Thanks for linking it though |
Beta Was this translation helpful? Give feedback.
-
|
Nevermind, it works now. Thanks again |
Beta Was this translation helpful? Give feedback.
-
Seems to work for me |
Beta Was this translation helpful? Give feedback.
-
|
I made an embedding for stable diffusion 2.0 based on greg rutkowski's work. The idea is to illustrate that you can get close to his style just from textual inversion despite his artwork not existing in the training data. This was the case with the 1.5 models as well, except the previous clip model was trained on him. https://www.reddit.com/r/StableDiffusion/comments/z5hlqa/greg_rutkowski_embedding_for_automatic1111/ |
Beta Was this translation helpful? Give feedback.
-
|
I'll add this when organizing the embeddings section, thanks for making this (and sorry for the late reply) |
Beta Was this translation helpful? Give feedback.
-
|
Hi! I write a colab notebook based on Japanese Fine-tuning Guide that can run in 12GB VRAM , by Kohya_ss, it's called Kohya Trainer. This might be a good choice for everyone who wants to be an AI trainer but doesn't get enough VRAM Do you think you could add it to the list? |
Beta Was this translation helpful? Give feedback.
-
|
I also got permission from Kohya_ss, to be able to publish this notebook (if anyone asking about am i stealing someone code):
|
Beta Was this translation helpful? Give feedback.
-
|
It runs well so far, I use it to fine-tune a model and create Hitokomoru Diffusion and other experimental model, it already available on huggingface if anyone interested |
Beta Was this translation helpful? Give feedback.
-
|
Thank you in advance! |
Beta Was this translation helpful? Give feedback.
-
|
Ah interesting, I'll be sure to add this and add proper credit when organizing the Training section. Thanks for linking and making this (and sorry for the late reply). |
Beta Was this translation helpful? Give feedback.
-
Welcome to sdupdates
This is a place where you can discuss anything relating to SD or sdupdates, like:
Beta Was this translation helpful? Give feedback.
All reactions