Prompt Extend

Summary

• Introduction
• Parameters
• Examples
• Usage for developers
• Reference
• License

Introduction

Prompt Extend is an innovative approach that aims to enhance the capabilities of language models and improve their response generation. It involves extending the initial prompt or query by providing additional context or specifications to guide the model's understanding and generate more accurate and relevant responses.

Model Description

The text generation model is designed to generate appropriate style cues based on the main idea provided in a prompt. This model utilizes GPT-2, a powerful language model, which has been trained on a dataset of stable diffusion prompts.

Parameters

Inputs

• input - (text): The input of the model is a main idea provided in the prompt. This main idea can be a sentence or a short paragraph, representing the core concept, topic, or general content that the model will generate text about.

Output

• output - (text): The output of the model is the generated text based on the main idea of the prompt. This text contains suitable style cues in line with stable diffusion prompts. These style cues may include language usage, sentence structure, vocabulary, expression, or any other stylistic elements commonly found in stable diffusion prompts.

Examples

input	output
giant dragon flying in the sky	giant dragon flying in the sky, flying in the sky, shooting fire by makoto shinkai and ferdinand knab magic the gathering illustration city 4k Trending on artstation HQ, deviantart, kimi no na wa, simoni, 8k hd
peaceful village landscape	peaceful village landscape, by makoto shinkai and thomas kinkade, trending on artstation, cgsociety cinematic lighting, sunny weather, colorful, very wide river, pastel colors, realistic trees, visually stunning visuals, top down view, bright sun rays, realistic sunset
a coconut laying on the beach	a coconut laying on the beach, hyper realism, cinematic, volumetric lighting, intricate complexity, extremely detailed, photorealistic, film still cinematic lighting, post-processing, unreal engine, concept art, smooth, sharp focus, a mix of styles

Usage for developers

Please find below the details to track the information and access the code for processing the model on our platform.

Requirements

torch
transformers

Code based on AIOZ structure

from transformers import pipeline
import torch, os

...
def do_ai_task(
        input: Union[str, Path],
        model_storage_directory: Union[str, Path],
        device: Literal["cpu", "cuda", "gpu"] = "cpu",
        *args, **kwargs) -> Any:
    """Define AI task: load model, pre-process, post-process, etc ..."""
    # Define AI task workflow. Below is an example
    device = "cuda" if torch.cuda.is_available() else "cpu"
    model = os.path.abspath(model_storage_directory + "/...")
    pipe = pipeline('text-generation', model=model, device=device)
    with torch.no_grad():
        result = pipe(input+',', num_return_sequences=1)[0]["generated_text"]
    return str(result)

Reference

This repository is based on and inspired by Partho Das's work. We sincerely appreciate their generosity in sharing the code.

License

We respect and comply with the terms of the author's license cited in the Reference section.