genpixel

Sleeping

App Files Files Community

genpixel / app.py

jn1xia

Update app.py

2f08fbd verified 10 months ago

raw

history blame contribute delete

6.48 kB

	import streamlit as st
	from PIL import Image
	import requests
	from io import BytesIO
	# import tensorflow as tf
	# import tensorflow as tf
	import streamlit as st
	from diffusers import StableDiffusionPipeline
	from diffusers import DiffusionPipeline
	import torch
	from accelerate import Accelerator
	#model_id = "CompVis/stable-diffusion-v1-4"
	#pipe = StableDiffusionPipeline.from_pretrained(model_id)


	# Ensure the model is using the CPU
	#pipe = pipe.to("cpu")

	image_html = ""
	accelerator = Accelerator()

	# Function to display an example image
	def display_example_image(url):
	response = requests.get(url)
	img = Image.open(BytesIO(response.content))
	st.image(img, caption='Generated Image', use_column_width=True)

	#function to generate AI based images using Huggingface Diffusers
	def generate_images_using_huggingface_diffusers(text):

	from diffusers import DiffusionPipeline

	#pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0")
	#pipe.load_lora_weights("strangerzonehf/Flux-Midjourney-Mix2-LoRA")

	pipe = DiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4")


	# pipe = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4", torch_dtype=torch.float32)
	# pipe = pipe.to("cuda")
	pipe = pipe.to(accelerator.device)
	prompt = text
	image = pipe(prompt,num_images_per_prompt=3)
	return image

	# Placeholder function for generating images (replace this with your actual generative AI code)
	def generate_images(prompt, num_images=3):
	# This is a placeholder function. Replace it with your actual image generation code.
	# Here, we are just returning the same example image multiple times for demonstration.
	image_url = pipe(prompt, num_images_per_prompt=num_samples, num_inference_steps=50, guidance_scale=7.5).images # Replace with a valid image URL
	response = requests.get(image_url)
	img = Image.open(BytesIO(response.content))
	image_html = image_url
	return [img] * num_images


	title_center = """
	<style>
	.title{
	text-align: center
	}
	</style>

	"""
	# Title of the app
	st.markdown(title_center, unsafe_allow_html=True)

	title_container = """
	<h1 class="title">AutoFloor</h1>
	"""

	st.markdown(title_container, unsafe_allow_html=True)
	# Text input for user prompt
	user_prompt = st.text_input("Enter your prompt here:")

	# file= st.file_uploader ("Unggah file Gambar", type = ["jpg", "png"])

	# model = tf.keras.models.load_model('L00005_HL512_bagus.h5')

	st.markdown("""
	<style>.element-container:has(#button-after) + div button {
	margin: 0 auto;
	display: block;
	}</style>""", unsafe_allow_html=True)


	st.markdown('<span id="button-after"></span>', unsafe_allow_html=True)
	# Generate and display images in a 3x3 grid
	if st.button('Generate Images', type="primary"):
	if user_prompt:
	st.write(f"Prompt: {user_prompt}")
	#image_url = "https://wpmedia.roomsketcher.com/content/uploads/2022/01/06145940/What-is-a-floor-plan-with-dimensions.png" # Replace with a valid image URL
	# Generate images based on the user's prompt
	#generated_images = generate_images(user_prompt, num_images=3)
	image_output = generate_images_using_huggingface_diffusers(user_prompt)
	st.info("Generating image.....")
	st.success("Image Generated Successfully")
	st.image(image_output, caption="Generated by Huggingface Diffusers")

	html_code = f"""
	<!DOCTYPE html>
	<html lang="en">
	<head>
	<meta charset="UTF-8">
	<meta name="viewport" content="width=device-width, initial-scale=1.0">
	<style>
	.zoomable-image {{
	transition: transform 0.3s ease;
	cursor: pointer;
	}}
	</style>
	<script src="https://unpkg.com/[email protected]/dist/panzoom.min.js"></script>
	</head>
	<body>
	<div id="image-container" style="text-align: center;">
	<img id="zoomable-image" class="zoomable-image" src="{{image_url}}" alt="Zoomable Image" style="max-width: 100%; height: auto;">
	</div>
	<script>
	document.addEventListener('DOMContentLoaded', (event) => {{
	const image = document.getElementById('zoomable-image');
	const panzoomInstance = panzoom(image, {{
	maxZoom: 3,
	minZoom: 1,
	bounds: false,
	boundsPadding: 0.1
	}});

	image.addEventListener('click', () => {{
	const currentTransform = image.style.transform;
	if (currentTransform.includes('matrix')) {{
	panzoomInstance.zoomAbs(0, 0, 1);
	}} else {{
	panzoomInstance.zoomAbs(image.width / 2, image.height / 2, 3);
	}}
	}});

	image.addEventListener('dblclick', () => {{
	const xys = panzoomInstance.getTransform()
	if(xys.scale > 1) {{
	const fScale = 1 - xys.scale
	const fixeX = xys.x / fScale
	const fixeY = xys.y / fScale
	panzoomInstance.smoothZoomAbs(fixeX, fixeY, 1)
	}} else {{
	panzoomInstance.moveBy(-xys.x, -xys.y, true)
	panzoomInstance.smoothZoomAbs(xys.x, xys.y, 1)
	}}
	panzoomInstance.moveTo(0, 0)
	panzoomInstance.zoomAbs(0, 0, 1)
	}});
	}});
	</script>
	</body>
	</html>
	"""

	# # Embed the HTML and JavaScript into the Streamlit app
	# st.components.v1.html(html_code, height=300)


	# Display images in a 3x3 grid
	cols = st.columns(3)
	for i in range(3):
	# for j in range(3):
	# with cols[j]:
	# st.image(generated_images[i * 3 + j], use_column_width=True)

	# Display the image with zoom effect
	# container_style = """
	# <div class="container">
	# <img class="zoom" src="{}" style="width:100%;">
	# </div>
	# """.format(image_url)
	# st.markdown(container_style, unsafe_allow_html=True)
	st.components.v1.html(html_code, height=600)
	else:
	st.write("Please enter a prompt.")