Sign In

KM_MixExperimentStage1

Updated: Apr 16, 2024

characterwomananime2.5dgirlsstyles

Download

1 variant available

fp32 SafeTensor

KM_MixExperimentStage1_realistic.safetensors

Full precision, largest file (pruned) • 2.3 GB

Verified:

Type
Checkpoint Merge
Stats

92

Reviews
Published

Apr 13, 2024

Base Model

SD 1.5

Hash
AutoV2
A0498F4122
default creator card background decoration
Followers - 220

220

Likes - 1979

2K

00630-3365480525-(masterpiece),best quality,1girl,hat,flower,sunflower,solo,long hair,dress,looking at viewer,straw hat,jewelry,petals,hand on he.png

I am conducting model blending to validate one of my hypotheses. Currently, due to the lack of model diversity, it has become difficult for me to blend sufficiently unique models from the existing materials. In conclusion, I need to move on to the next stage.

我正在进行模型混合以验证我的一个假设。目前,由于模型缺乏多样性,我很难从现有材料中混合出足够独特的模型。总之,我需要进入下一个阶段。

The next update will be delayed by 10 days, so I have prepared the latest mixed 9 models.

下一次更新将会延迟10天,因此我已经准备好了最新的混合9个模型。

Another model was abandoned from uploading due to lack of distinctive features.

还有一个模型因为缺乏特点而放弃上传了

None of my models have used lora hybrid, maybe later I will try to directly mix lora.

我的模型都没有使用Lora混合,也许以后我会尝试直接混合Lora。

It is worth noting that all Lora and embeddings will affect the image effect. If you want to understand a model, I think you should try to leave those Lora and embeddings.

值得注意的是,所有的Lora和embeddings 将会影响图像效果。如果你想了解一个模型,我认为你应该尝试离开那些Lora和embeddings



HAHAHA, how much do we really understand about AI now?