Spaces:

Alphar3d
/

Venusaur

Sleeping

App Files Files Community

Venusaur / app.py

Alphar3d

Update app.py

c570a53 verified about 2 months ago

raw history blame contribute delete

No virus

3.17 kB

	import google.generativeai as genai
	from pathlib import Path
	import gradio as gr
	from dotenv import load_dotenv
	import os

	# Load environment variables from a .env file
	load_dotenv()

	# Configure the GenerativeAI API key using the loaded environment variable
	genai.configure(api_key=os.getenv("AIzaSyDGsO9HM1KDZ0BdZ1RGWP8lC2XR4A_Oz5w"))

	# Set up the model configuration for text generation
	generation_config = {
	"temperature": 0.4,
	"top_p": 1,
	"top_k": 32,
	"max_output_tokens": 4096,
	}

	# Define safety settings for content generation
	safety_settings = [
	{"category": f"HARM_CATEGORY_{category}",
	"threshold": "BLOCK_MEDIUM_AND_ABOVE"}
	for category in ["HARASSMENT", "HATE_SPEECH", "SEXUALLY_EXPLICIT", "DANGEROUS_CONTENT"]
	]

	# Initialize the GenerativeModel with the specified model name, configuration, and safety settings
	model = genai.GenerativeModel(
	model_name="gemini-pro-vision",
	generation_config=generation_config,
	safety_settings=safety_settings,
	)
	# Function to read image data from a file path


	def read_image_data(file_path):
	image_path = Path(file_path)
	if not image_path.exists():
	raise FileNotFoundError(f"Could not find image: {image_path}")
	return {"mime_type": "image/jpeg", "data": image_path.read_bytes()}

	# Function to generate a response based on a prompt and an image path


	def generate_gemini_response(prompt, image_path):
	image_data = read_image_data(image_path)
	response = model.generate_content([prompt, image_data])
	return response.text


	# Initial input prompt for the plant pathologist
	input_prompt = """
	You are an expertised doctor. The user will upload an image of the injury. Analyse the image and provide a complete analysis to assess for insurance claim estimation.

	Additionally, provide any relevant details such as:
	1. Type and severity of the injury (e.g., broken bone, laceration, burn).
	2. Location of the injury on the body.
	3. Total cost that the user can claim from the insurance company.

	Based on the provided image and information, the model will analyze the injury and estimate the amount that can be claimed from the insurance."
	Disclaimer:
	"Please note that the information provided is based on the image analysis and should not replace professionals."
	"""

	# Function to process uploaded files and generate a response

	def process_uploaded_files(files):
	file_path = files[0].name if files else None
	response = generate_gemini_response(
	input_prompt, file_path) if file_path else None
	return file_path, response

	# Gradio interface setup
	with gr.Blocks() as demo:
	file_output = gr.Textbox()
	image_output = gr.Image()
	combined_output = [image_output, file_output]

	# Upload button for user to provide images
	upload_button = gr.UploadButton(
	"Click to Upload an Image",
	file_types=["image"],
	file_count="multiple",
	)
	# Set up the upload button to trigger the processing function
	upload_button.upload(process_uploaded_files,
	upload_button, combined_output)

	# Launch the Gradio interface with debug mode enabled
	demo.launch(debug=True)