clockclock's picture
Update app.py
4bdb54f verified
raw
history blame
3.05 kB
# app.py (Final Corrected Version)
import gradio as gr
import torch
from diffusers import AutoPipelineForInpainting
from PIL import Image
import time
# --- Model Loading ---
print("Loading model for low-RAM CPU environment...")
model_id = "runwayml/stable-diffusion-inpainting"
try:
pipe = AutoPipelineForInpainting.from_pretrained(model_id, torch_dtype=torch.float32)
pipe.enable_model_cpu_offload()
print("Model loaded successfully with CPU offloading enabled.")
except Exception as e:
print(f"An error occurred during model loading: {e}")
raise e
# --- Prompts ---
DEFAULT_PROMPT = "photorealistic, 4k, ultra high quality, sharp focus, masterpiece, high detail"
DEFAULT_NEGATIVE_PROMPT = "blurry, pixelated, distorted, deformed, ugly, disfigured, cartoon, watermark"
# --- Inpainting Function (Correct Signature) ---
def inpaint_image(image_and_mask, user_prompt, guidance_scale, num_steps, progress=gr.Progress(track_tqdm=True)):
# The input is now a dictionary with 'image' and 'mask' keys
image = image_and_mask["image"].convert("RGB")
mask = image_and_mask["mask"].convert("RGB")
if image is None or mask is None:
raise gr.Error("Please upload an image and draw a mask on it first!")
if user_prompt and user_prompt.strip():
prompt = user_prompt
negative_prompt = DEFAULT_NEGATIVE_PROMPT
else:
prompt = DEFAULT_PROMPT
negative_prompt = DEFAULT_NEGATIVE_PROMPT
print(f"Starting inpainting on CPU...")
result_image = pipe(
prompt=prompt, image=image, mask_image=mask, negative_prompt=negative_prompt,
guidance_scale=guidance_scale, num_inference_steps=int(num_steps)
).images[0]
return result_image
# --- UI ---
with gr.Blocks(theme=gr.themes.Soft()) as demo:
gr.Markdown("# 🎨 AI Image Fixer (Stable Version)")
gr.Warning("‼️ PATIENCE REQUIRED! Generation can take 15-30 minutes on free hardware.")
with gr.Row():
with gr.Column(scale=2):
# This component returns a dictionary when tool='brush'
input_image = gr.Image(label="1. Upload & Mask Image", source="upload", tool="brush", type="pil")
prompt_textbox = gr.Textbox(label="2. Describe Your Fix (Optional)", placeholder="Leave empty for a general fix")
with gr.Accordion("Advanced Settings", open=False):
guidance_scale = gr.Slider(minimum=0, maximum=20, value=8.0, label="Guidance Scale")
num_steps = gr.Slider(minimum=10, maximum=50, step=1, value=20, label="Inference Steps")
with gr.Column(scale=1):
output_image = gr.Image(label="Result", type="pil")
submit_button = gr.Button("Fix It!", variant="primary")
# The `inputs` list is simple. The function signature must match what Gradio provides.
submit_button.click(
fn=inpaint_image,
inputs=[input_image, prompt_textbox, guidance_scale, num_steps],
outputs=output_image
)
if __name__ == "__main__":
demo.launch()