Since the release of Anymagine XL 3.0, there is no need to update the model.
1.32 is the last update.
The Anymagine XL 3.0 is superior to this model.
version 1.3
I extracted LORA from previously merged models and added LORA and lycoris that I learned before to increase tag accuracy.
(kohakuXL extract + DPO extract + LORA learned personally + hassaku extract)
version 1.21
Deephentai-mix-v1.2 (This is a different model than the test version.)
+
Version 1.21 merges BluePencil 2.0.0 and block tuning to improve stability.
Version 1.0 was released for about a month with no consideration for stability.
just focused on merging a lot of information.
Because of this, it was very unstable.
The 1.21 model is a compilation of models created by people much better than me,
compensating for weaknesses in my model with block tuning,
and merged in other additional LORAs.
It is still being worked on for stability and tag compatibility.
I recommend my embedding
I recommend the default VAE.
positive: 2 or 3 big tags like style, fonts, etc. + 1 sentences describing the situation
Too many sentences can rather damage the image.
Intuitive and consistent tags and 1-2 clear sentences are optimal.
negative:
text, watermark, negative, negative_v2, negativeXL_D
It is a model that merged several checkpoint models and learned Lora after learning it.
(hentaiMixXLRoadTo_v21+_hassakuXLSfwNsfw_alphaV09-sdXL_v10_x0.5.fp16+ssd-1b-dalle-3-step00077700.fp16 + (kohakuXLBeta_beta7Pro+_reproductionSDXL_2v12-sd_xl_base_0.9-fp16_x0.25.fp16+ssd-1b-dalle-3-step00077700.fp16 - sd_xl_base_0.9-fp16) x 0.2) x deephentaiMix_v40_Stableversionx0.775+envyagendaxl02_v10x0.225.fp16 x 0.75 + bluePencilXL_v200 x 0.25 x deephentaiMix_v50 + (hentaiMixXLRoadTo_v3025DHCNSFW - sdXL_v10) x alpha(cosine) + Merge with Lora personally made
Description
FAQ
Comments (10)
There is some color issue with your model on A111. Not using any vae. preview during creation shows normal colors, but the final image become dull.
As far as I know, the checkpoint problem doesn't cause that problem
Accurately mark it as a merge model, not a Trained model
Although the number of mergers was not small, the basis of the model was lora training, and in writing, it was marked as training.
@colanamnam It's not made up of a purely hands-on model or lora
it is correct to be classified as a merged model.
It doesn't matter exactly which way it is written. I merged and trained, but I just marked it like that by the weight difference. There's no benefit that I can get from it in the first place.
@colanamnam It is not good to provide inaccurate information to those searching for models
Will the deleted lora be uploaded elsewhere? I've seen similar ones hosted on huggingface.
I didn't do it.
Details
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.