@settima_ai

As a result of trying to add various elements, it is unstable and has a lot of gacha elements. I think it takes patience and love to output a good image lol

色々な要素を入れようとした結果、不安定で、ガチャ要素が多いです、良い画像を出力するには根気と愛が必要だと思いますｗ

@settima_ai

When using two different characters, LORA, it is necessary to check Latent. Have you done that?

For Attention, mix, and for Latent, place each region.

Do you understand this?

2つの異なるキャラクターLORAを使用する場合、Latentにチェックを入れる必要が有りますが、それは行っていますか？

Attentionの場合は混合し、Latentはリージョン毎に配置します

これで解りますか？

@settima_ai

As the PC is currently training, images cannot be generated.I have attached links to output examples using three different character LORAs in the past.However, different LORAs are placed in different regions. Then, each number occupies GPU memory.If it is 12G of memory, the limit is three LORAs, and I think it is best to limit it to two for normal use.

現在PCに学習をさせている為、画像の生成が出来ません、過去に、3つの異なる、キャラクターのLORAを使用した出力例のリンクを付けておきます、ただし、異なるLORAを別リージョンに配置すると個数毎にGPUのメモリを占有します、12Gのメモリであれば、3つのLORAが限界で、通常使用では2つまでに留めるのが良いと思います。

https://civitai.com/images/2529708

@settima_ai 

I'll write the simplest example.

Please check this box.

Please also check active

Then at the prompt

<lora:AtaruMoroboshi001:0.7>,2 people,

AtaruMoroboshi,1boy,

ADDCOL,

<lora:ShutaroMendou001:0.7>,2 people,

ShutaroMendou,1boy,

-------------------------------

最も簡単な例を書きます。

このチェックを入れて下さい。

アクティブにもチェック入れてね

その後、プロンプトに

<lora:AtaruMoroboshi001:0.7>,2 people,

AtaruMoroboshi,1boy,

ADDCOL,

<lora:ShutaroMendou001:0.7>,2 people,

ShutaroMendou,1boy,

@settima_ai

The link didn't work, so I've pasted it on Rei.

リンクが繋がらなかったのでReiに張り付けてあります。

Rei (Urusei Yatsura) LORA - v1.0 | Stable Diffusion LoRA | Civitai

@settima_ai 

After I finished learning, I uploaded a sample after the explanation.

学習が終わったので、説明の後にサンプルをアップしました。

Hello, I think someone else has made one, but is it not available?

@settima_ai 

I confirmed that you corrected the sentence, but I am very angry with you. I have a certain respect for people who are working hard for the development of AI for free. I put the LORA I created in the trash and pressed the delete button. It's very unfortunate.

Image post by jibunsagasinotabi | Civitai

@settima_ai 

You said that the LORA that someone else created was trash, so I threw it in the trash too.

@settima_ai 

I was planning to release UY_AtaruMoroboshiXLpony001, but I canceled and scrapped it because you made abusive remarks. That's all.