ad astra per asperga 
(added colab instructions)
Colab:
-
Open automatic1111's colab: https://colab.research.google.com/drive/1Iy-xW9t1-OQWhb0hNxueGij8phCyluOh
-
Replace this cell:
#@title Normal 1.4 model
# get a token from https://huggingface.co/settings/tokens
user_token = "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" #@param {type:"string"}
user_header = f"\"Authorization: Bearer {user_token}\""
!wget --header={user_header} https://huggingface.co/CompVis/stable-diffusion-v-1-4-original/resolve/main/sd-v1-4.ckpt -O model.ckpt
With:
!apt install megatools
!megadl "https://mega.co.nz/#!FxclSKoL!EFSM4nLlXMuOvBLkoZNmtOH4Y8oycjrU7h2Hn6mKl1k"
-
Run (it'll take a few minutes to download)
-
Refer to Marsey as
a sks catin your prompt.a photo of a sks cat as a gardeneretc
Download the weights: https://models.rdra.ma/
Development on Stable Diffusion has been happening at a wild pace, and there's already written a working implementation of DreamBooth for it. Whereas Textual Inversion refines your prompt, DreamBooth is able to finetune the model itself to teach it new concepts. I did one run with 87 Marseys and another with just 3 upscaled ones. So far the upscaled run (u1000.ckpt) is looking best.
See the Textual Inversion posts for a comparison with that. DreamBooth is far better at keeping her colors consistent and producing recognizable images. Decent results were something like 1 in 50 before, now they're nearly every image.
a photo of Marsey
a photo of a spaceman Marsey in outer space
a photo of Marsey as a gardener
I love that it creates real Marseys too 
a photo of Marsey running away from a dinosaur
Missing the 
. Might be overfitting and improve with better regularization in DreamBooth.
a photo of Marsey cooking in the kitchen
These models also work with Deforum to create animations:
There's still plenty of low-hanging fruit to suggest these will get better quickly - the DreamBooth repo doesn't do regularization correctly, and we'll get a free bump in quality when SDv1.5 is released. The plague of AIposting continues
Jump in the discussion.
No email address required.
Holy shit we're on the cusp of replacing our own Marsey artists, no wonder they're pissing and pooping themselves.
Jump in the discussion.
No email address required.
I canβt fricking wait to incorporate this into marseycrossing are you joking
of course for marseyartists that canβt codeβ¦
Jump in the discussion.
No email address required.
also even these aren't perfect, this tool is best seen as a tool for artists. for instance this lil dude (C:2, R:3) is cute but he out of frame
Jump in the discussion.
No email address required.
More options
Context
More options
Context
More options
Context
my mind is literally blown right now, this is insane. it has the style, the structure, the shape perfect.
currently downloading the model, can you just swap out the original SD model with the marsey model? if so I could use AUTOMATIC1111's webui
also, why is the model so big? its like 3x larger than the SD model lmao
also also, models.rdra.ma, when did that happen, lol
edit: column 2 row 3 is actually really cute
Jump in the discussion.
No email address required.
Yeah it's a drop in replacement so automatic's UI should work. The extra size is from the "ema" version of the SD model which has extra metadata for training
I was high one night and got rdra.ma off njalla, I'm glad I'm getting some use out of it lmao
Jump in the discussion.
No email address required.
Can you use the usual pruning script on that model?
I'm talking about shit like https://github.com/harubaru/waifu-diffusion/blob/main/scripts/prune.py
I use a "custom" .bat (https://pastebin.com/R15jTB8U) .py (https://pastebin.com/W4mrB0nf) where you just have to edit the last line of the .py for your model. I think it needs to be in the root webui folder but I'm not certain. It works on Waifu-Diffusion so it might work on Marsey-Diffusion
Jump in the discussion.
No email address required.
Oh perfect, thanks, I'll give that a try in the future
Jump in the discussion.
No email address required.
Also, if you want to do crossovers between models (like waifu-diffusion and marsey-diffusion for anime marseys), it's dead easy to do, you just need to add the weights lmao
There's an automated script .bat (https://pastebin.com/0J9TPEcs) along with the .py (https://pastebin.com/056xeuVm). This one doesn't require changing anything, just type what's needed in cmd.
It will ask you for a ratio between the two models, there's no general rule on merges yet so just guess.
Jump in the discussion.
No email address required.
More options
Context
More options
Context
based af
Jump in the discussion.
No email address required.
Darn that's a LOT of unused shit, most models "only" get halved in size
Jump in the discussion.
No email address required.
More options
Context
More options
Context
More options
Context
More options
Context
More options
Context
cc
@HeyMoon cc 
@X (formerly chiobu) (could you swap out my pin in /h/marsey?)
if anyone has a prompt idea lmk and I'll try to generate it
Jump in the discussion.
No email address required.
-Marsey at a blm riot
-Marsey on trial for warcrimes
-marsey in a tim burton movie
-marsey executing POWs in saigon
-god-emporer marsey
-marsey shocked and scared when confronting a photorealistic cat
-marsey catching and eating a carp
Jump in the discussion.
No email address required.
editing the others in in a sec
the tooth it gives shocked Marsey is cute
Jump in the discussion.
No email address required.
This is gold dude
Jump in the discussion.
No email address required.
More options
Context
Jump in the discussion.
No email address required.
How do you plead?
Jump in the discussion.
No email address required.
More options
Context
More options
Context
This frickin rules
Jump in the discussion.
No email address required.
More options
Context
zoz
Jump in the discussion.
No email address required.
zle
Jump in the discussion.
No email address required.
zozzle
Jump in the discussion.
No email address required.
More options
Context
More options
Context
More options
Context
Hahaha holy shit this is awesome. The blm riots are perfect
Jump in the discussion.
No email address required.
More options
Context
More options
Context
More options
Context
can you do george bush marsey
Jump in the discussion.
No email address required.
tried a few, but style transfer seems trickier with dreambooth than textual inversion. I wonder if combining the two would work, hmm
Jump in the discussion.
No email address required.
More options
Context
take #2:
prompt2prompt is amazing with DB
. you gotta try it if you haven't (automatic's UI has a couple versions, this and this, the second one is the "img2img alt" script)
just need to figure out if it can be trained for style transfer too
Jump in the discussion.
No email address required.
i may be dumb but where is the version for automatic1111? this is neat btw
Jump in the discussion.
No email address required.
for doggettx's version, just use it in the prompt (
digital art of a [cat:sks cat] holding a gun). the other one is in the scripts dropdown under "img2img alternative test". negative prompts like "poorly drawn" and "bad art" seem to help tooJump in the discussion.
No email address required.
Oh shit, its already in the code? Nice I will def try that thanks
Jump in the discussion.
No email address required.
More options
Context
More options
Context
More options
Context
More options
Context
More options
Context
More options
Context
Born just in time to see good ai generated marseys
Jump in the discussion.
No email address required.
More options
Context
INFINITE MARSEYS
Jump in the discussion.
No email address required.
More options
Context
These are incredible
Jump in the discussion.
No email address required.
More options
Context
Holy shit these are amazing, especially the astronaut Marsey
Current HP: 8/75
Current Mana: 130/250
Inventory:
* 50 Gald
Jump in the discussion.
No email address required.
You could go right in there and re-design an "Always Has Been" meme with AI space Marsey.
Jump in the discussion.
No email address required.
More options
Context
More options
Context
Holy shit these are amazing, maybe you could try some prompts of different artists from https://generrated.com/ ?
Jump in the discussion.
No email address required.
can't believe I forgot to try this. I think it kind of works, I'll have to try again on artists that focus on animals in the future
Caspar David Friedrich -
Edward Hopper -
Howard Arkley -
Jump in the discussion.
No email address required.
those edward hopper marseys are perfect
Jump in the discussion.
No email address required.
More options
Context
More options
Context
More options
Context
I hate how good these are.
Jump in the discussion.
No email address required.
More options
Context
Wowwee mister this is pretty nifty
Not like my dad
Jump in the discussion.
No email address required.
More options
Context
πAbsolutely ππΌamazingππΏ
Jump in the discussion.
No email address required.
More options
Context
Snapshots:
https://web.archive.org/https://i.imgur.com/hm3JJg6.mp4
https://archive.ph/?url=https://i.imgur.com/hm3JJg6.mp4&run=1 (click to archive)
https://ghostarchive.org/search?term=https://i.imgur.com/hm3JJg6.mp4 (click to archive)
https://models.rdra.ma/:
https://web.archive.org/https://models.rdra.ma/
https://archive.ph/?url=https://models.rdra.ma/&run=1 (click to archive)
https://ghostarchive.org/search?term=https://models.rdra.ma/ (click to archive)
working implementation:
https://web.archive.org/https://github.com/XavierXiao/Dreambooth-Stable-Diffusion
https://archive.ph/?url=https://github.com/XavierXiao/Dreambooth-Stable-Diffusion&run=1 (click to archive)
https://ghostarchive.org/search?term=https://github.com/XavierXiao/Dreambooth-Stable-Diffusion (click to archive)
https://web.archive.org/https://i.rdrama.net/images/16841312643438563.webp
https://archive.ph/?url=https://i.imgur.com/T5PAT1d_d.webp?maxwidth=9999&fidelity=high&run=1 (click to archive)
https://ghostarchive.org/search?term=https://i.imgur.com/T5PAT1d_d.webp?maxwidth=9999&fidelity=high (click to archive)
https://web.archive.org/https://i.rdrama.net/images/1684131265366033.webp
https://archive.ph/?url=https://i.imgur.com/7UCyptp_d.webp?maxwidth=9999&fidelity=high&run=1 (click to archive)
https://ghostarchive.org/search?term=https://i.imgur.com/7UCyptp_d.webp?maxwidth=9999&fidelity=high (click to archive)
https://web.archive.org/https://i.rdrama.net/images/16841312669646115.webp
https://archive.ph/?url=https://i.imgur.com/Mlj7IHD_d.webp?maxwidth=9999&fidelity=high&run=1 (click to archive)
https://ghostarchive.org/search?term=https://i.imgur.com/Mlj7IHD_d.webp?maxwidth=9999&fidelity=high (click to archive)
https://web.archive.org/https://i.rdrama.net/images/16841312684157832.webp
https://archive.ph/?url=https://i.imgur.com/i1ReyQm_d.webp?maxwidth=9999&fidelity=high&run=1 (click to archive)
https://ghostarchive.org/search?term=https://i.imgur.com/i1ReyQm_d.webp?maxwidth=9999&fidelity=high (click to archive)
https://web.archive.org/https://i.rdrama.net/images/1684131269856919.webp
https://archive.ph/?url=https://i.imgur.com/xoOkLal_d.webp?maxwidth=9999&fidelity=high&run=1 (click to archive)
https://ghostarchive.org/search?term=https://i.imgur.com/xoOkLal_d.webp?maxwidth=9999&fidelity=high (click to archive)
Jump in the discussion.
No email address required.
More options
Context
finally a thread that needs commendations, well done!
-wpd user
Jump in the discussion.
No email address required.
More options
Context
comparison of characters and styles in different marsey attempts. "sks cat" is dreambooth version, "fmarsey" is float-trip's original, "smarsey2" is my (shit) attempt at using textual-inversion
smaller model: https://mega.nz/file/AWsDGIAC#3EXe-ChMJaIBOlale2--ld6usX55Q9H6iqLVnRHlvNA
Jump in the discussion.
No email address required.
textual inversion is no good with this. I just trained embeddings off of the new model here - https://models.rdra.ma/embeddings/
elephant in the style of *(old model/embeddings)elephant in the style of *(new model/embeddings)could be related to this: https://github.com/XavierXiao/Dreambooth-Stable-Diffusion/issues/15
at some point it might be worth commissioning Marsey in 3-5 more standard poses for the AI to learn from. a few issues (like her tail) seem to stem from the dataset
Jump in the discussion.
No email address required.
More options
Context
version that has what the originals would look like unmodified by style
comparison of holding a sword
Jump in the discussion.
No email address required.
[row] marsey holding [column]
Jump in the discussion.
No email address required.
More options
Context
Comparison of CFG scales
lol she be gooning with friends
Jump in the discussion.
No email address required.
horny marsey
Jump in the discussion.
No email address required.
porn
Jump in the discussion.
No email address required.
More options
Context
More options
Context
More options
Context
More options
Context
More options
Context