BoleroMix LyCORIS Extract版本v1.0 BM-1.41 LyCORIS r128 (ID: 948292)

BoleroMix LyCORIS Extract版本v1.0 BM-1.41 LyCORIS r128 (ID: 948292)

Version naming info: v1.0 BM-1.41 LyCORIS r128
Version of this model BoleroMix-version Model Type Network Dimension/Rank of the model

This is a LyCORIS extract of BoleroMix by bolero537

By using LoRA Block Weight (Inspire-Pack has LBW nodes for Comfy) and this LyCORIS I can use BoleroMix's style on other checkpoints (gallery has comparison grids) Upd.: I generated some images with this and the quality is pretty good, i might use them to make a style LoRA

For LyCORIS LBW use 0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1 to apply style; to slightly enhance the effect use weight = 1 or 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1 (comparison image)

I created this because I really like the aesthetic of BoleroMix and also like the 4th Tail model and I wanted BoleroMix's style on 4th Tail. The problem with it, or rather me, is that I don't know any artists that make up a similar style and it's just much more convenient if the model has a nice stable base style I know and like overall instead of switching around artist styles I picked out (might upload my wildcard file at some point). And a plus point is that it can be used on other models too, and probably on the on-site generator

Info: 0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1 will only apply the OUT layers (identical to the OUT12 block merge preset) of the LyCORIS model , in theory it applies all aesthetic/style from the model to a checkpoint but weight = 1 or setting all layers to 1 will get it closer to the original model of the extract while possibly losing composition/concept knowledge of the checkpoint its applied to (im not sure about this so correct me if im wrong).

SD WebUI LBW preset: FXL_OUT12:0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1
Usage:
<lora:BoleroMix_LycoExtract_r128:1:1:lbw=0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1>
<lora:BoleroMix_LycoExtract_r128:1:1:lbw=FXL_OUT12>

Note: The comparison grids (except the simple 1x2) have the Comfy workflow embedded into them.

Info on the extract: This extract was made using Kohya_ss GUI in the LoRA > Tools > Extract LyCORIS LoCon tab and used PonyXL v6 for the base model selection

I may try around with lower rank extracts and see how it affects the style, so far there wasnt much difference in style in comparison to higher rank, so far with the existing extracts (see comparison grid) i dont see much difference in style and that is why i uploaded the lowest rank lycoris. Also as already mentioned: I generated some images with this and the quality is pretty good, i might use them to make a style LoRA

(Information even if I haven't uploaded a LoRA version: XL LoRA has less layers so the equivalent would be 0,0,0,0,0,0,1,1,1,1,1,1, ALSO: if you see LBW stuff online and they have 17 LoRA/26 LyCORIS layers, they are for sd1.5. sd1.5 models have more layers than sdxl)

If the author (bolero) wants me to remove this model because it doesn't count as merge I will do so.
This model or merges are not allowed to be uploaded anywhere outside of CivitAI and merges on CivitAI must use the same license as this model/BoleroMix

If you know any artists that have a similar style to BoleroMix please leave a comment/DM me :>

描述:

Network/Conv dimension (rank) of 128

This extract was made using Kohya_ss GUI in the LoRA > Tools > Extract LyCORIS LoCon tab and used PonyXL v6 for the base model selection.
Mode: fixed; net dim 128; conv dim 128
no sparse bias

训练词语:

名称: BoleroMix_LycoExtract_r128.safetensors

大小 (KB): 936787

类型: Model

Pickle 扫描结果: Success

Pickle 扫描信息: No Pickle imports

病毒扫描结果: Success

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

BoleroMix LyCORIS Extract

资源下载
下载价格VIP专享
仅限VIP下载升级VIP
犹豫不决让我们错失一次又一次机会!!!
原文链接:https://1111down.com/1157745.html,转载请注明出处
由于网站升级,部分用户密码全部设置为111111,登入后自己修改, 并且VIP等级提升一级(包月提升至包季,包季提升到包年 包年提升至永久)
没有账号?注册  忘记密码?

社交账号快速登录