Imrayya/stable-diffusion-webui-Prompt_Generator
1
0
Fork 0
mirror of https://github.com/imrayya/stable-diffusion-webui-Prompt_Generator.git synced 2024-01-11 09:00:44 +01:00
Code Issues Releases Wiki Activity

Added Magic Prompt by Gustavosta

Browse Source
This commit is contained in:
Kareem
2023-01-15 16:25:58 +01:00
parent 557e4c969a
commit 5358234f55
3 changed files with 149 additions and 24 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+25 -6
View File
@@ -1,25 +1,44 @@
# Prompt Generator
Adds a tab to the webui that allows the user to generate a prompt from a small base prompt. Based on [FredZhang7/distilgpt2-stable-diffusion-v2](https://huggingface.co/FredZhang7/distilgpt2-stable-diffusion-v2). I did nothing apart from porting it to [AUTOMATIC1111 WebUI](https://github.com/AUTOMATIC1111/stable-diffusion-webui)
Adds a tab to the webui that allows the user to generate a prompt from a small base prompt. Based on [FredZhang7/distilgpt2-stable-diffusion-v2](https://huggingface.co/FredZhang7/distilgpt2-stable-diffusion-v2) and [Gustavosta/MagicPrompt-Stable-Diffusion](https://huggingface.co/Gustavosta/MagicPrompt-Stable-Diffusion). I did nothing apart from porting it to [AUTOMATIC1111 WebUI](https://github.com/AUTOMATIC1111/stable-diffusion-webui)
![image](https://user-images.githubusercontent.com/8998556/209890919-203463fe-4b25-4ba0-9b29-57b1744dfd0f.png)
## Installation:
## Installation
1. Install [AUTOMATIC1111's Stable Diffusion Webui](https://github.com/AUTOMATIC1111/stable-diffusion-webui)
2. Clone this repository into the `extensions` folder inside the webui
## Usage:
## Usage
1. Write in the prompt in the *Start of the prompt* text box
2. Click Generate and wait
## Parameters Explanation
## Parameters Explanation
- **Start of the prompt**: As the name, the start of the prompt that the generator should start with
- **Temperature**: A higher temperature will produce more diverse results, but with a higher risk of less coherent text
- **Top K**: The number of tokens to sample from at each step
- **Top K**: Strategy is to sample from a shortlist of the top K tokens. This approach allows the other high-scoring tokens a chance of being picked.
- **Max Length**: the maximum number of tokens for the output of the model
- **Repetition Penalty**: The penalty value for each repetition of a token
- **Repetition Penalty**: The parameter for repetition penalty. 1.0 means no penalty. See [this paper](https://arxiv.org/pdf/1909.05858.pdf) for more details. Default setting is 1.2
- **How Many To Generate**: The number of results to generate
- **Use blacklist?** Using `.\extensions\stable-diffusion-webui-Prompt_Generator\blacklist.txt`. It will delete any matches to the generated result (case insensitive). Each item to be filtered out should be on a new line. *Be aware that it simply deletes it and doesn't generate more to make up for the lost words*
## Models
There are two models provided:
### FredZhang7
Made by [FredZhang7](https://huggingface.co/FredZhang7) under creativeml-openrail-m license. Useful to get styles for a prompt. Eg: "A cat sitting" -> "A cat sitting on a chair, digital art. The room is made of clay and metal with the sun shining through in front trending at Artstation 4k uhd..."
### MagicPrompt
Made by [Gustavosta](https://huggingface.co/Gustavosta) under the MIT license. Useful to get more natural language prompts. Eg: "A cat sitting" -> "A cat sitting in a chair, wearing pair of sunglasses"
*Be aware that sometimes the model fails to produce anything or less than the wanted amount, either try again or use a new prompt in that case*
## Credits
Credits to both [FredZhang7](https://huggingface.co/FredZhang7) and [Gustavosta](https://huggingface.co/Gustavosta)