feat: Implement AI-powered embeddings and vector similarity search system
This commit is contained in:
Aherobo Ovie Victor
@@ -0,0 +1,220 @@
|
||||
"""Groq LLM integration for DS Task AI News"""
|
||||
import os
|
||||
from typing import List, Dict, Any, Optional
|
||||
from groq import Groq
|
||||
from config import settings
|
||||
|
||||
class GroqLLMService:
|
||||
def __init__(self):
|
||||
self.client = None
|
||||
self.model = "llama3-8b-8192" # Default Groq model
|
||||
|
||||
# Initialize Groq client if API key is available
|
||||
if settings.groq_api_key:
|
||||
try:
|
||||
self.client = Groq(api_key=settings.groq_api_key)
|
||||
print("✅ Groq LLM service initialized")
|
||||
except Exception as e:
|
||||
print(f"⚠️ Groq initialization failed: {e}")
|
||||
self.client = None
|
||||
else:
|
||||
print("⚠️ Groq API key not provided")
|
||||
|
||||
def is_available(self) -> bool:
|
||||
"""Check if Groq service is available"""
|
||||
return self.client is not None
|
||||
|
||||
def summarize_article(self, article: Dict[str, Any]) -> Optional[str]:
|
||||
"""Generate a summary for an article"""
|
||||
if not self.is_available():
|
||||
return None
|
||||
|
||||
try:
|
||||
title = article.get('title', '')
|
||||
content = article.get('content', '')
|
||||
|
||||
prompt = f"""
|
||||
Please provide a concise summary of this news article in 2-3 sentences:
|
||||
|
||||
Title: {title}
|
||||
Content: {content}
|
||||
|
||||
Summary:
|
||||
"""
|
||||
|
||||
response = self.client.chat.completions.create(
|
||||
messages=[
|
||||
{"role": "user", "content": prompt}
|
||||
],
|
||||
model=self.model,
|
||||
max_tokens=150,
|
||||
temperature=0.3
|
||||
)
|
||||
|
||||
summary = response.choices[0].message.content.strip()
|
||||
return summary
|
||||
|
||||
except Exception as e:
|
||||
print(f"Error generating summary: {e}")
|
||||
return None
|
||||
|
||||
def analyze_sentiment(self, article: Dict[str, Any]) -> Optional[str]:
|
||||
"""Analyze sentiment of an article"""
|
||||
if not self.is_available():
|
||||
return None
|
||||
|
||||
try:
|
||||
title = article.get('title', '')
|
||||
content = article.get('content', '')
|
||||
|
||||
prompt = f"""
|
||||
Analyze the sentiment of this news article. Respond with only one word: "positive", "negative", or "neutral".
|
||||
|
||||
Title: {title}
|
||||
Content: {content}
|
||||
|
||||
Sentiment:
|
||||
"""
|
||||
|
||||
response = self.client.chat.completions.create(
|
||||
messages=[
|
||||
{"role": "user", "content": prompt}
|
||||
],
|
||||
model=self.model,
|
||||
max_tokens=10,
|
||||
temperature=0.1
|
||||
)
|
||||
|
||||
sentiment = response.choices[0].message.content.strip().lower()
|
||||
|
||||
# Validate response
|
||||
if sentiment in ['positive', 'negative', 'neutral']:
|
||||
return sentiment
|
||||
else:
|
||||
return 'neutral' # Default fallback
|
||||
|
||||
except Exception as e:
|
||||
print(f"Error analyzing sentiment: {e}")
|
||||
return None
|
||||
|
||||
def extract_keywords(self, article: Dict[str, Any]) -> Optional[List[str]]:
|
||||
"""Extract key topics/keywords from an article"""
|
||||
if not self.is_available():
|
||||
return None
|
||||
|
||||
try:
|
||||
title = article.get('title', '')
|
||||
content = article.get('content', '')
|
||||
|
||||
prompt = f"""
|
||||
Extract 3-5 key topics or keywords from this news article. Return them as a comma-separated list.
|
||||
|
||||
Title: {title}
|
||||
Content: {content}
|
||||
|
||||
Keywords:
|
||||
"""
|
||||
|
||||
response = self.client.chat.completions.create(
|
||||
messages=[
|
||||
{"role": "user", "content": prompt}
|
||||
],
|
||||
model=self.model,
|
||||
max_tokens=50,
|
||||
temperature=0.3
|
||||
)
|
||||
|
||||
keywords_text = response.choices[0].message.content.strip()
|
||||
keywords = [kw.strip() for kw in keywords_text.split(',') if kw.strip()]
|
||||
|
||||
return keywords[:5] # Limit to 5 keywords
|
||||
|
||||
except Exception as e:
|
||||
print(f"Error extracting keywords: {e}")
|
||||
return None
|
||||
|
||||
def generate_insights(self, articles: List[Dict[str, Any]]) -> Optional[str]:
|
||||
"""Generate insights from multiple articles"""
|
||||
if not self.is_available() or not articles:
|
||||
return None
|
||||
|
||||
try:
|
||||
# Create a summary of article titles
|
||||
titles = [article.get('title', '') for article in articles[:10]] # Limit to 10 articles
|
||||
titles_text = '\n'.join([f"- {title}" for title in titles])
|
||||
|
||||
prompt = f"""
|
||||
Based on these recent news headlines, provide 2-3 key insights about current trends or themes:
|
||||
|
||||
Headlines:
|
||||
{titles_text}
|
||||
|
||||
Key Insights:
|
||||
"""
|
||||
|
||||
response = self.client.chat.completions.create(
|
||||
messages=[
|
||||
{"role": "user", "content": prompt}
|
||||
],
|
||||
model=self.model,
|
||||
max_tokens=200,
|
||||
temperature=0.4
|
||||
)
|
||||
|
||||
insights = response.choices[0].message.content.strip()
|
||||
return insights
|
||||
|
||||
except Exception as e:
|
||||
print(f"Error generating insights: {e}")
|
||||
return None
|
||||
|
||||
def enhance_article(self, article: Dict[str, Any]) -> Dict[str, Any]:
|
||||
"""Enhance article with AI-generated metadata"""
|
||||
enhanced_article = article.copy()
|
||||
|
||||
if self.is_available():
|
||||
# Add summary
|
||||
summary = self.summarize_article(article)
|
||||
if summary:
|
||||
enhanced_article['ai_summary'] = summary
|
||||
|
||||
# Add sentiment
|
||||
sentiment = self.analyze_sentiment(article)
|
||||
if sentiment:
|
||||
enhanced_article['sentiment'] = sentiment
|
||||
|
||||
# Add keywords
|
||||
keywords = self.extract_keywords(article)
|
||||
if keywords:
|
||||
enhanced_article['ai_keywords'] = keywords
|
||||
|
||||
return enhanced_article
|
||||
|
||||
def batch_enhance_articles(self, articles: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
|
||||
"""Enhance multiple articles with AI features"""
|
||||
enhanced_articles = []
|
||||
|
||||
for article in articles:
|
||||
enhanced = self.enhance_article(article)
|
||||
enhanced_articles.append(enhanced)
|
||||
|
||||
return enhanced_articles
|
||||
|
||||
# Test function
|
||||
if __name__ == "__main__":
|
||||
# Test Groq integration
|
||||
groq_service = GroqLLMService()
|
||||
|
||||
if groq_service.is_available():
|
||||
print("✅ Groq service is available")
|
||||
|
||||
# Test with sample article
|
||||
sample_article = {
|
||||
"title": "AI Technology Advances in Healthcare",
|
||||
"content": "Recent developments in artificial intelligence are transforming the healthcare industry with new diagnostic tools and treatment methods."
|
||||
}
|
||||
|
||||
enhanced = groq_service.enhance_article(sample_article)
|
||||
print(f"Enhanced article: {enhanced}")
|
||||
else:
|
||||
print("⚠️ Groq service not available (API key needed)")
|
||||
Reference in New Issue
Block a user