gradio-ui/ui.py

import gradio as gr
import requests
from pydantic import BaseModel
from diffusers.utils import load_image


SDXL_LORA_API_URL = 'http://127.0.0.1:8000/api/v1/product-diffusion/sdxl_v0_lora_inference'

# Define the InpaintingRequest model
class InpaintingRequest(BaseModel):
    prompt: str
    num_inference_steps: int
    guidance_scale: float
    negative_prompt: str
    num_images: int
    mode: str

def generate_sdxl_lora_image(prompt, negative_prompt, num_inference_steps, guidance_scale, num_images, mode):
    # Prepare the payload for SDXL LORA API
    payload = InpaintingRequest(
        prompt=prompt,
        negative_prompt=negative_prompt,
        num_inference_steps=num_inference_steps,
        guidance_scale=guidance_scale,
        num_images=num_images,
        mode=mode
    ).model_dump()
    
    response = requests.post(SDXL_LORA_API_URL, json=payload)
    response_json = response.json()
    url = response_json['url']
    
    image = load_image(url)
    return image

with gr.Blocks(theme='gradio/soft') as demo:
    with gr.Tab("SDXL LORA TEXT-TO-IMAGE"):
        with gr.Row():
            with gr.Column(scale=1):
              
                    prompt = gr.Textbox(label="Prompt", placeholder="Enter your prompt here")
                    negative_prompt = gr.Textbox(label="Negative Prompt", placeholder="Enter negative prompt here")
                   

               
                    with gr.Column(scale=1):
                        num_inference_steps = gr.Slider(minimum=1, maximum=1000, step=1, value=20, label="Inference Steps")
                        guidance_scale = gr.Slider(minimum=1.0, maximum=10.0, step=0.1, value=7.5, label="Guidance Scale")
                        num_images = gr.Slider(minimum=1, maximum=10, step=1, value=1, label="Number of Images")
                        mode = gr.Dropdown(choices=["s3_json", "b64_json"], value="s3_json", label="Mode")
                        generate_button = gr.Button("Generate Image")
            
            
            image_preview = gr.Image(label="Generated Image", height=512, width=512,scale=1,show_download_button=True,show_share_button=True,container=True)
        
        generate_button.click(generate_sdxl_lora_image, inputs=[prompt, negative_prompt, num_inference_steps, guidance_scale, num_images, mode], outputs=[image_preview])
    

demo.launch()