模型/有用的高質量嵌入 - aid291

有用的高質量嵌入 - aid291

陳嘉欣 (Chan Ka-yan)

10/14/2026

12:25:34 AM

| Discussion

提示

將下載的負面文本嵌入文件放置於您的 Stable Diffusion 目錄下的 embeddings 文件夾中。

使用括號增強提示詞不等同於增加向量強度。

文本嵌入的向量數會影響提示詞可包含的詞元數量；超過 75 個詞元的提示詞部分可能會被忽略。

版本亮點

主觀評價：aid291 是一個專用於 AnimeIllustDiffusionV2.91 的特製 負面文本嵌入，能夠巨幅提高出圖質量，比其他任何該系列的負面文本嵌入都要好。它它在其他模型上的表現 尚未測試過。當輸入負面提示詞時，請將要添加其他普通的負面 tag，放在 "aid291" tags 之前，以免弱效 tag 效用被覆蓋。（例如，aid291, extra digits -> extra digits, aid291）

作用範圍：提高清晰度，消除偽影噪點，修正線條扭曲，添加細節，修正人體結構，改善手部細節，穩定構圖，輕微美顏

向量數：39

向量強度：30% 0.5, 70% 0.35

缺點：（可能）會把不必要的物體引入

Subjective evaluation: aid291 is a specially crafted negative text embedding for AnimeIllustDiffusionV2.91. It hugely improves aid291 model's performance, better than any other negative text embeddings I have released so far. However, its performance on other models has not been tested, but worth trying. When inputting negative prompts, it's recommended to put other general tags before 'aid291' tag to prevent it weaken other tags before it. (e.g., aid291, extra digits -> extra digits, aid291)

Effect: Improve sharpness, eliminate artifact and noise, correct line distortion, add details, correct human anatomy, improve hand structure, stabilize composition, and slightly beautify face.

Vector count: 39

Vector strength: 30% 0.5, 70% 0.35

Disadvantage: (May) Introduce unnecessary objects into

介紹（中文）

基本信息

該頁面為推薦用於 AnimeIllustDiffusion [1] 模型的所有文本嵌入（Embedding）。您可以從版本描述查看該文本嵌入的信息。

使用方法

您應該將下載到的負面文本嵌入文件置入您 stable diffusion 目錄下的 embeddings 文件夾內。之後，您只需要在填寫負面提示詞處輸入 badv4 即可。

評價參數

每個文本嵌入將從以下參數描述其特性：

正/負面文本嵌入：如果一個文本嵌入是正面文本嵌入，則您應該將它填寫在正面提示詞內發揮作用；反之，如果一個文本嵌入是負面文本嵌入，則您應該將它填寫在負面提示詞內發揮作用。

作用範圍：一個文本嵌入主要影響的畫面內容。

向量數：代表該文本嵌入佔多少個詞元。通常來說，您輸入提示詞中超過 75 個詞元的部分會被忽略。因此，一個文本嵌入的向量數越少，您填入它之後剩餘輸入提示詞的空間越多。

向量強度：一個詞元向量值的大小。向量強度越高，效果越強。普通詞元的向量強度為[-0.05,0.05]。不可將用括號增強提示詞等效於增強向量強度。

Introduction (English)

Basic Information

This page contains all the text embeddings recommended for use with the AnimeIllustDiffusion model [1]. You can view information about the text embeddings from the version description.

Usage

You should place the downloaded negative text embedding file in the embeddings folder of your Stable Diffusion directory. After that, you only need to enter "badv4" in the field for negative prompts.

Evaluation Parameters

Each text embedding is described by the following parameters:

Positive/Negative Text Embedding: If a text embedding is a positive text embedding, it should be used in positive prompts. Conversely, if a text embedding is a negative text embedding, it should be used in negative prompts.

Scope: The main content of the image that a text embedding will primarily affect.

Vector Count: The number of token that the text embedding represents. Generally, the part of a prompt containing more than 75 tokens will be ignored. Therefore, the fewer the number of vectors in a text embedding, the more space you will have left to input the prompt.

Vector Strength: The magnitude of a token's vector. The higher the vector strength, the stronger the effect. The vector strength of token is usually between [-0.05, 0.05]. Using parentheses to enhance prompts is not equivalent to increasing vector strength.