README.md

---
license: creativeml-openrail-m
language:
- en
widget:
- text: 1girl, fate
- text: 1boy, league of 
- text: 1girl, genshin
- text: 1boy, national basketball association
- text: 1girl, spy x
- text: 1girl, absurdres
tags:
- stable-diffusion
- anime
- anything-v4
- art
- arxiv:2210.14140
datasets:
- FredZhang7/anime-prompts-180K
---

## Fast Anime PromptGen

The main model (`pytorch_model.bin`) was trained on a dataset of **80,000** anime tags and for 3 epochs. I fetched the tags from the [Safebooru API endpoint](https://safebooru.donmai.us/posts/random.json), but only accepted the ones with **up_score ≥ 8** and without any [blacklisted tags](./blacklist.txt). 
I didn't release the V1 model because it only generated gibberish prompts. After trying all means to correct that behavior, I eventually figured that the cause of the gibberish prompts is not from the pipeline params, model structure or training duration, but rather from the random usernames in the training data. 
Here's the complete [prompt preprocessing algorithm](./preprocess.py).


Todo:
- upload Danbooru model

## Text-to-image Examples

Prefix *1girl* | [Generated *1girl* prompts](./anime_girl_settings.txt) | Model *Anything V4*

![](./anime_girls.png)

Prefix *1boy*  | [Generated *1boy* prompts](./anime_boy_settings.txt) | Model *Anything V4*

![](./anime_boys.png)

## Contrastive Search
```
pip install --upgrade transformers
```
```python
import torch
from transformers import GPT2Tokenizer, GPT2LMHeadModel, pipeline
tokenizer = GPT2Tokenizer.from_pretrained('distilgpt2')
tokenizer.add_special_tokens({'pad_token': '[PAD]'})
model = GPT2LMHeadModel.from_pretrained('FredZhang7/anime-anything-promptgen')

prompt = r'1girl, genshin'

# generate text using fine-tuned model
nlp = pipeline('text-generation', model=model, tokenizer=tokenizer)

# generate 10 samples using greedy search
outs = nlp(prompt, max_length=76, num_return_sequences=10, do_sample=True, repetition_penalty=1.2, temperature=0.7, top_k=4, early_stopping=True)

print('\nInput:\n' + 100 * '-')
print('\033[96m' + prompt + '\033[0m')
print('\nOutput:\n' + 100 * '-')
for i in range(len(outs)):
    # remove trailing commas and double spaces
    outs[i] = str(outs[i]['generated_text']).replace('  ', '').rstrip(',')
print('\033[92m' + '\n\n'.join(outs) + '\033[0m\n')
```

Output Example:

![](./contrastive_search.png)

Please see [Fast GPT PromptGen](https://huggingface.co/FredZhang7/distilgpt2-stable-diffusion-v2) for more info on the pipeline parameters.


## Tips

- If you feel like a generated anime character doesn't show emotions, try emoticons like `;o`, `:o`, `;p`, `:d`, `:p`, and `;d` in the prompt.
I also use `happy smirk`, `happy smile`, `laughing closed eyes`, etc. to make the characters more lively and expressive.

- Adding `absurdres`, instead of `highres` and `masterpiece`, to a prompt tends to increase the sharpness of a generated image.