LoRA extractions is not as good as the full model but it creates a compromise between size and quality and it can be used along other LoRA model with a full model. Credit goes the creator of the checkpoint in version description
Sorry, i do not accept anymore request link(s) in my comments "Suggestions for a model extraction?"
04.0417.23 Name version is limited to 10 25 character, make it impossible to see full name... see below "List of all LoRA model".
How to use it in AUTOMATIC1111 WebUI?
Place the LoRA model in .\stable-diffusion-webui\models\Lora then use <lora:filename:multiplier> in the prompt.
About the comparison:
Picture order: First = LoRA, Second = Fine-tuned Model, Third = Base Model
DM = Network Dimension ( 120=lower quality/low size, 320=higher quality/higher size ) same for CD = Conv Dimension (Rank)
[CD is optionally if you wish to use it, install the extension: https://github.com/KohakuBlueleaf/a1111-sd-webui-locon (How? Extensions -> Install from URL)]
Obviously Version "1024CV1024" or in short "1024" use more VRAM
I did a large comparison DM+CV heresample pictures does not use vea and LoRA model with SDv1.4
every model from (Stable Diffusion Base Model: mostly sd-v1-4.ckpt, model base is determined by highest % affinity) was extracted using Kohya_ss (Save precision: fp16)
Name format: <Type><DM> <CV><Name>
List of all LoRA model:
LORA1024CV1024experience_80 - LORA320experience_80
LORA1024CV1024cardosAnimated_v20 - LORA320cardosAnimated_v20
LORA1024CV1024flat2DAnimerge_v10 - LORA320flat2DAnimerge_v10
LORA1024CV1024realismEngine_v10 - LORA320realismEngine_v10
LORA1024CV1024vectorArt_pulpVectorBeta - LORA320vectorArt_pulpVectorBeta
LORA1024CV1024hasdx_hasedsdx - LORA320hasdx_hasedsdx
LORA1024CV1024verisimilitude_V14 - LORA320verisimilitude_V14
LORA1024CV1024galaxytimemachinesForyou_v10 - LORA320galaxytimemachinesForyou_v10
LORA1024CV1024galaxytimemachinesGTM_v3Pruned - LORA320galaxytimemachinesGTM_v3Pruned
LORA1024CV1024hassakuv1m
LyCORIS1024CV1024illuminatiDiffusionV1_v11 - LORA320illuminatiDiffusionV1_v11
LORA1024CV1024xrikishisUnstable_alpha50 - LORA320xrikishisUnstable_alpha50
LORA1024CV1024diivesStyle_10 - LORA320diivesStyle_10
LORA1024CV1024elldrethsImagination_v10 - LORA320elldrethsImagination_v10
LORA1024CV1024digitalVividMemoriesAnime_10 - LORA320digitalVividMemoriesAnime_10
LORA1024CV1024generalPurpose - LORA320generalPurpose
LORA1024CV1024dreamlikeDiffusion10_10 - LORA320dreamlikeDiffusion10_10
LORA1024CV1024epicDiffusion11 - LORA320epicDiffusion11
LORA1024CV1024protogenInfinityX86 - LORA320protogenInfinityX86
LORA1024CV1024protogenNovaExperimentalX80 - LORA320protogenNovaExperimentalX80
LORA1024CV1024protogenX58RebuiltScifi - LORA320protogenX58RebuiltScifi
LORA1024CV1024modernDisney_v1 - LORA320modernDisney_v1
LORA1024CV1024revAnimated_v11 - LORA320revAnimated_v11
LORA1024CV1024revAnimated_v122 - LORA320revAnimated_v122
LORA1024CV1024comicBabes_v1 - LORA320comicBabes_v1
LORA320CD128perfectWorld_v2Baked - LORA120CD128perfectWorld_v2Baked -LORA1024CD1024perfectWorld_v2Baked
LORA14DM320realisticVisionV20_v20 - LORA14DM120realisticVisionV20_v20 - LORA1024CV1024dreamshaper_4NoVae
LORADM320realisticVisionV13_v13 - LORADM120realisticVisionV13_v13
LoRADM320dreamshaper_4NoVae - LoRADM120dreamshaper_4NoVae
LORA14DM320grapefruitv41 -LORA14DM120grapefruitv41
LORA1024CV1024berrygold11 - LORA14DM120BerryGold-1-1 - LORA14DM320BerryGold-1-1
LORA14DM120chilloutmix_NiPrunedFp32Fix - LORA14DM320chilloutmix_NiPrunedFp32Fix - LORA1024CV1024chilloutmix_NiPrunedFp32Fix
LORA14DM120deliberate_v2 - LORA14DM320deliberate_v2 - LORA1024CV1024deliberate_v2
LORA14DM120HSPU-1.1 - LORA14DM320HSPU-1.1 - LORA1024CV1024hspu_11
LORA14DM120HSPUCar1-1 - LORA14DM320HSPUCar1-1 - LORA1024CV1024hspuCarV10
LORA14DM120Maiden - LORA14DM320Maiden - LORA1024CV1024maiden
LORA1024CV1024yiffye18Furry4Kinky
LORA14DM320protogenX53Photorealism_10 - LORA14DM120protogenX53Photorealism_10
LORA14DM320anything-v4.5-vae-swapped - LORA14DM120anything-v4.5-vae-swapped - LORA1024CV1024anything-v4.5-vae-swapped
LORA14DM320robo-diffusion-v1 - LORA14DM120robo-diffusion-v1 - LORA1024CV1024robo-diffusion-v1
LORA14DM320Silicon29 - LORA14DM120Silicon29 - LORA1024CV1024Silicon29
Description
FAQ
Comments (20)
Suggestions for a model extraction?
you are amazing..also you put sd1.4 in the info
https://civitai.com/models/20294/comic-babes
https://civitai.com/models/7240/meinamix
https://civitai.com/models/16164/icomix
https://civitai.com/models/6250/dosmix
https://civitai.com/models/7371/rev-animated
https://civitai.com/models/24/modern-disney
https://civitai.com/models/3867/protogen-x58-rebuilt-scifianime-official-release
https://civitai.com/models/4331/protogen-nova-experimental-model-official-release
https://civitai.com/models/4398/protogen-infinity-official-release
I dont have the bandwidth to download these models and I know loras are less of ideal but I'll appreciate it if it isn't a hassle
also will those work on 1.5 model?
https://civitai.com/models/3855/epic-diffusion
https://civitai.com/models/1274/dreamlike-diffusion-10
https://civitai.com/models/3762/general-purpose-model
https://civitai.com/models/6681/digital-vivid-memories-anime-and-digital-art
https://civitai.com/models/3866/elldreths-imagination-mix
I used to use these loras from Kappa_Neuro but for some reason I deleted them, I would appreciate if you could extract them.
Didn't see this comment before!
Here's a few I'm interested in:
GTM Ultimate Blend
https://civitai.com/models/4884/galaxytimemachines-gtmultimateblendv3-fantasyai
GTM Photo For You
https://civitai.com/models/25636/galaxytimemachines-foryou-photo-fantasyai
RPGv4
https://civitai.com/models/1116/rpg
Verisimilitude
https://civitai.com/models/10540/verisimilitude
HasDX
https://civitai.com/models/3758/hasdx
Thanks in advance!
@Sakura96 @grego1502 @RTY64 Added!
I cut off DM120 to replace it by (DM)1024CV1024 (description updated for CV). Part of the reason is that people didn't use it much.
base of all yours request models was sometimes mixing closely between v1.4 and v1.5 so somes extract got much undesirable v1.4-v1.5 but was ok..
I hope you will enjoy mixing your loRA :)
@herbert_west @FlamingoDelta Added! see in description "List of all LoRA model".
verisimilitude got me a good breath from waifu!
@DigitalDreamer Thank you for all of your work on this!!!
thanks for all the work you are doing for the community, for people who use google colab this is very helpful
@FlamingoDelta Added +Update revAnimated to v1.2.2
@DarkAgent Added CarDos Animated v2 + Experience v8.
@grego1502 hasdx is already in description "List of all LoRA model".
Im working on a new trained model Diffuse Texture so please be patient for realisticelves.
thx for your effort! was waiting for something like this since i cant do it and most model creators dont react to lora requests.
https://civitai.com/models/10961/lazymix-real-amateur-nudes
https://civitai.com/models/4404/sci-fi-diffusion-v10
https://civitai.com/models/9388/realscifi
https://civitai.com/models/16401/aatrok
https://civitai.com/models/4834/visiongen-realism-reborn
https://civitai.com/models/1136/postapocalypse
https://civitai.com/models/20617/color-fusion
https://civitai.com/models/26396/feverdream
https://civitai.com/models/56967/majicmix-lux
https://civitai.com/models/41865/majicmix-fantasy
https://civitai.com/models/56967/majicmix-lux
https://civitai.com/models/22922/lyriel?modelVersionId=50127
Little news not so dreamy as i will probably stop LoRA extract... I started this as experiments to see if like old extract like URPM(only 18+) turn out 320mb even for small training dataset like robo-diffusion-v1 is not enough and need 2GB (aka LORA1024CV1024) to do ok. I think there is better way to do so, seeing 400GB-LoRA of extract SDXL LoRA or lycoris FFusion by @idle does a fantastic job with small size/quality (here’s an idea when compare row 1:1 vs 2:2 500mb).
Given the more feed fulfillment requests I got each time I do, this gets a lot of attention where i would prefer to experiment (&LLM) and train medium/large datasets, nonetheless i was happy to discover new SD15 models from all requests!
Good work. As long as it is in line with the rules of the platform. Maybe it needs the explicit citing of the original authors.
Version description contain links to model author.
@DigitalDreamer Good ! Other LORA extractions in program, to add to this great collection ?
