Pai Bloom 1b1 Text2prompt Sd

🚀 BeautifulPrompt

我们开源了一个自动Prompt生成模型，用户只需输入极其简单的Prompt，就能得到经语言模型优化的Prompt，从而更轻松地生成高颜值图像。

🚀 快速开始

代码示例

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained('alibaba-pai/pai-bloom-1b1-text2prompt-sd')
model = AutoModelForCausalLM.from_pretrained('alibaba-pai/pai-bloom-1b1-text2prompt-sd').eval().cuda()

raw_prompt = '1 girl'
input = f'Instruction: Give a simple description of the image to generate a drawing prompt.\nInput: {raw_prompt}\nOutput:'
input_ids = tokenizer.encode(input, return_tensors='pt').cuda()

outputs = model.generate(
    input_ids,
    max_length=384,
    do_sample=True,
    temperature=1.0,
    top_k=50,
    top_p=0.95,
    repetition_penalty=1.2,
    num_return_sequences=5)

prompts = tokenizer.batch_decode(outputs[:, input_ids.size(1):], skip_special_tokens=True)
prompts = [p.strip() for p in prompts]
print(prompts)

✨ 主要特性

• 输入简单：用户可以直接输入极其简单的Prompt。
• 优化输出：得到经过语言模型优化过的Prompt，有助于生成更美观的图像。

💻 使用示例

基础用法

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained('alibaba-pai/pai-bloom-1b1-text2prompt-sd')
model = AutoModelForCausalLM.from_pretrained('alibaba-pai/pai-bloom-1b1-text2prompt-sd').eval().cuda()

raw_prompt = '1 girl'
input = f'Instruction: Give a simple description of the image to generate a drawing prompt.\nInput: {raw_prompt}\nOutput:'
input_ids = tokenizer.encode(input, return_tensors='pt').cuda()

outputs = model.generate(
    input_ids,
    max_length=384,
    do_sample=True,
    temperature=1.0,
    top_k=50,
    top_p=0.95,
    repetition_penalty=1.2,
    num_return_sequences=5)

prompts = tokenizer.batch_decode(outputs[:, input_ids.size(1):], skip_special_tokens=True)
prompts = [p.strip() for p in prompts]
print(prompts)

📚 详细文档

作品展示

原始Prompt	BeautifulPrompt优化后的Prompt
prompt: taylor swift, country, golden, fearless,wavehair	prompt: portrait of taylor swift as a beautiful woman, long hair, country, golden ratio, intricate, symmetrical, cinematic lighting, highly detailed, digital painting, artstation, concept art, smooth, sharp focus, illustration
prompt: A majestic sailing ship	prompt: a massive sailing ship, epic, cinematic, artstation, greg rutkowski, james gurney, sparth

使用须知

⚠️ 重要提示

使用上述模型需遵守AIGC模型开源特别条款。

论文引用

如果您觉得该模型有用，请考虑引用以下论文：

@inproceedings{emnlp2023a,
  author    = {Tingfeng Cao and
            Chengyu Wang and
            Bingyan Liu and
            Ziheng Wu and
            Jinhui Zhu and
            Jun Huang},
  title     = {BeautifulPrompt: Towards Automatic Prompt Engineering for Text-to-Image Synthesis},
  booktitle = {Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: Industry Track},
  pages     = {1--11},
  year      = {2023}
}

📄 许可证

本项目采用Apache-2.0许可证。