# ちちぷい プロンプト研究 aiart stablediffusion aiartworks ai aiイラト automatic1111 sd creating images photo design nft nftartist scifiaiart cyborg aiイラスト customaitraining synthara stablediffusion2 aiartcommunity ghosted controlnet

Progress: Initial settings for embedding 3D have been completed
進捗　3Dの組み込み用初期設定ができた

0 9
2024-12-17

新しい人が増えたので、趣味でやっているプロンプト・モデル収集活動の紹介をします。

モデル比較
https://t.co/zsr8vCI0dm

Embedding(反呪文)比較
https://t.co/zqItu7voa3

サンプラー比較
https://t.co/y20jIN5g4s

クオリティタグ比較
https://t.co/CEq0nVPsLC

https://t.co/Y27aTncxre

43 203
2023-06-22

A test. My negative prompt changed a lot since my early days and I also now us negative embeddings to increase quality, but sadly a lot of the interesting style I used to get is lost. Same prompt 3 times with different negatives:

6 43
2023-04-04

今日初めて知ったんだけど
badhandv4 っていう embeddings があって、それをネガティブに入れると、何もしなくても手がマシになるという代物。

1枚目が何も指定無し。
2枚目が (badhandv4:0.3) をネガティブに追加したもの。

多少絵は変わるけど手が意図した形になった。

33 167
2023-04-04

2 x Two different real photoshoots turned into trained Embedding
2 x Different Ai image collections turned into trained embedding
Together own a personalized art machine that spits out mindblowing pieces:D Did this for
4 contest :D

2 17
2023-04-04

C++20 フレームワーク Siv3D v0.6.8 をリリースしました🎉
・OpenAI Embeddings API
・ファイルを外部アプリで開く機能
・絵文字を Unicode 15.0 に更新
・アイコン 700 種追加
・ユーザの名前や言語の取得
など 8 件の新機能、1 件の改善、2 件のバグ修正。
リリースノート: https://t.co/97WQYKnuKA

6 36
2023-03-31

i was wrong but im still salty they didn't like me embedding my own ads on their site and deleted all my shit lol

but i see they're going heavier on ads again i guess

0 1
2023-03-25

[GPT-8R33] Stable Diffusion 1.5 vanilla, no embeddings + Photoshop. Just messing around a but with some good old sci-fi themed projects.

0 5
2023-03-21

そのままMAXまでupscale
1枚目：embeddingなしからupscale
2枚目：embeddingありからupscale

途中で補正したおかげで明らかな違い。

1 2
2023-03-20

iti latent upscaleで使ってみた。
効果ありますね。若干絵柄変わっちゃいますが。
1枚目：元画像
2枚目：embeddingなし
3枚目：embeddingあり

大きく破綻してたものが許容範囲にはなった。
https://t.co/D7dxo5NwAs

0 7
2023-03-20

Over the past 3-ish months, I've been using my works to train/guide Stable Diffusion 2.1 via custom embedding creation! It is so captivating to be able to capture aspects of my art.
Here are some from the "fungus skyscraper" set.

1 8
2023-03-17

Mountain (Arknights)
(without using Lora

不用embedding或者lora画特定人物其实是老早就试过的方法了，弄过神威的图。只不过山真的好难弄🥹

231 1497
2023-03-11

Gen'ing some images to go with slides for talk Tuesday 's , re. neural audio synthesis; audio production via embeddings; the links between notation, encoding, compression, MIDI, perception, etc.

1 4
2023-03-03

[Wind Angel] Something new today with vanilla stable diffusion. No embeddings.

0 5
2023-03-02

latent walk (spherical linear interpolation between word embeddings on CLIP's latent hypersphere) vs multi-cond guidance
the effect is similar, but pose changes more during latent walk

3 20
2023-02-26

This is a result without AddNet and embeddings. 2nd image is a line-art image with ControlNet from 1st image.

0 3
2023-02-26

my embedding does mostly black and white generations... it's limited, but always interesting.

with i'm able to add color + fx to them without changing the original form too much, which is pretty wild.

0 5
2023-02-26

Yep, my thoughts exactly. Honestly making embeddings work is pretty hard ! I really really want to be able to generate faces like this

0 1
2023-02-25

Intermediate results after 15 hours of training.
Clothes need more time to bake in but due to the limitation of clip some of them will probably fail.
A better way would be to do embedding + model training.

0 0
2023-02-25