PixCell-Backend / llm.py
aizerner's picture
Update llm.py
f2f0e4e verified
Raw History Blame Contribute Delete
3.02 kB
import os
import google.generativeai as genai
from typing import Dict, Any
from dotenv import load_dotenv
load_dotenv()
GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
if not GEMINI_API_KEY:
raise RuntimeError("GEMINI_API_KEY environment variable is not set. Please set it in your environment or .env file.")
# Set a working US proxy (test and replace; this one is from free lists as of Dec 2025)
# Find fresh ones at free-proxy-list.net or proxylister.com — filter for US, HTTP, high uptime
#os.environ['HTTP_PROXY'] = 'http://198.199.86.11:80'
#os.environ['HTTPS_PROXY'] = 'http://198.199.86.11:80'
# Get API key from environment variable
# Check GC LOL -molt
def build_prompt_from_counts(class_counts: Dict[str, int], sample_type: str = "Blood smear", stain: str = "Giemsa", magnification: str = "1000x") -> str:
"""
Build a prompt for AI analysis based on detection results.
Args:
class_counts: Dictionary mapping class names to counts
sample_type: Type of sample (default: "Blood smear")
stain: Type of stain used (default: "Giemsa")
magnification: Microscope magnification (default: "1000x")
Returns:
Formatted prompt string for AI analysis
"""
summary = "Detected features in sample:\n"
for cls, count in class_counts.items():
summary += f"- {count} {cls.replace('_', ' ')}(s)\n"
summary += f"Sample type: {sample_type}\nStain: {stain}\nMagnification: {magnification}\n"
return f"""
You are a medical pathologist. Based on the following detection results from a microscopic image, write a concise diagnostic summary and suggest next steps.
{summary}
Output:
- Diagnostic Summary:
- Recommendation:
"""
def analyze_detections(class_counts: Dict[str, int], sample_type: str = "Blood smear", stain: str = "Giemsa", magnification: str = "1000x") -> Dict[str, Any]:
"""
Analyze detection results using Gemini AI.
Args:
class_counts: Dictionary mapping class names to counts
sample_type: Type of sample
stain: Type of stain used
magnification: Microscope magnification
Returns:
Dictionary containing analysis results
"""
try:
# Configure the API key
genai.configure(api_key=GEMINI_API_KEY,transport='rest')
#print("Gemini configured with proxy successfully.")
# Initialize the Gemini model
model = genai.GenerativeModel('gemini-2.5-flash')
# Build the prompt
prompt = build_prompt_from_counts(class_counts, sample_type, stain, magnification)
# Generate content from the model
response = model.generate_content(prompt)
return {
"success": True,
"analysis": response.text,
"detection_summary": class_counts
}
except Exception as e:
return {
"success": False,
"error": str(e),
"detection_summary": class_counts
}