Remove Gemini API integration and related functionality from embedding and fact extraction modules

app/core/embedding.py

@@ -1,25 +0,0 @@
-import os
-
-# Replace local embedding model with Gemini API integration
-GEMINI_API_KEY = os.getenv("GEMINI_API_KEY", "")
-
-if not GEMINI_API_KEY:
-    raise ValueError("❌ Gemini API Key (GEMINI_API_KEY) has not been set yet")
-
-def generate_embedding(texts: list):
-    import requests
-    url = f"https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key={GEMINI_API_KEY}"
-    headers = {"Content-Type": "application/json"}
-    payload = {
-        "contents": [{"parts": [{"text": text}]} for text in texts]
-    }
-
-    try:
-        response = requests.post(url, headers=headers, json=payload)
-        response.raise_for_status()
-        data = response.json()
-        # Parse embeddings from the response (adjust based on Gemini API's actual response structure)
-        return [item.get("embedding", []) for item in data.get("contents", [])]
-    except Exception as e:
-        print(f"Error generating embedding via Gemini API: {e}")
-        return None

app/core/fact_extraction.py

@@ -1,55 +0,0 @@
-import os
-from app.core.device_setup import device
-from app.core.fact_management import save_user_fact
-from app.core.logging_setup import logger
-
-# Replace local model loading with Gemini API integration
-GEMINI_API_KEY = os.getenv("GEMINI_API_KEY", "")
-
-if not GEMINI_API_KEY:
-    raise ValueError("❌ Gemini API Key (GEMINI_API_KEY) has not been set yet")
-
-def query_gemini_for_entities(text: str):
-    import requests
-    url = f"https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key={GEMINI_API_KEY}"
-    headers = {"Content-Type": "application/json"}
-    payload = {
-        "contents": [{"parts": [{"text": text}]}]
-    }
-
-    try:
-        response = requests.post(url, headers=headers, json=payload)
-        response.raise_for_status()
-        data = response.json()
-        # Parse entities from the response (adjust based on Gemini API's actual response structure)
-        return data.get("entities", [])
-    except Exception as e:
-        logger.error(f"🚨 Error querying Gemini API for entities: {e}")
-        return []
-
-# Replace the NLP pipeline with Gemini API calls
-def extract_and_store_facts(message):
-    entities = query_gemini_for_entities(message)
-
-    # Extract name
-    name = next((entity['name'] for entity in entities if entity.get('type') == 'PERSON'), None)
-    if name:
-        save_user_fact("name", name)
-        logger.info(f"User name '{name}' stored in memory.")
-
-    # Extract location
-    location = next((entity['name'] for entity in entities if entity.get('type') == 'LOCATION'), None)
-    if location:
-        save_user_fact("location", location)
-        logger.info(f"User location '{location}' stored in memory.")
-
-    # Extract hobbies (simple keyword detection)
-    hobbies = []
-    hobby_keywords = ["love", "enjoy", "like", "hobby", "passion"]
-    for token in message.split():
-        if token.lower() in hobby_keywords:
-            hobbies.append(token)
-
-    if hobbies:
-        save_user_fact("hobbies", ", ".join(hobbies))
-        logger.info(f"User hobbies '{', '.join(hobbies)}' stored in memory.")

app/main_hf.py

@@ -1,4 +1,3 @@
-import os
 from dotenv import load_dotenv
 load_dotenv()
 

app/routes/chat_hf.py

@@ -6,7 +6,6 @@ from datetime import datetime, timezone
 from app.core.device_setup import device
 from app.core.message_saving import save_message
 from app.core.past_conversations import get_past_conversations
-from app.core.fact_extraction import extract_and_store_facts
 from app.core.memory_management import reset_memory, get_last_interaction, update_last_interaction
 from app.core.fact_management import get_user_fact, save_user_fact
 from app.core.conversation_retrieval import get_similar_conversations
@@ -40,16 +39,18 @@ def query_gemini_api(prompt: str) -> str:
     url = f"https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key={GEMINI_API_KEY}"
     headers = {"Content-Type": "application/json"}
     payload = {
-        "contents": [{"parts": [{"text": prompt}]}]
+        "prompt": {"text": prompt},
+        "temperature": 0.7,
+        "maxOutputTokens": 256
     }
 
     try:
         response = requests.post(url, headers=headers, json=payload)
         response.raise_for_status()
         data = response.json()
-        return data.get("contents", [{}])[0].get("parts", [{}])[0].get("text", "")
-    except Exception as e:
-        logger.error(f"🚨 Error querying Gemini API: {e}")
+        return data.get("candidates", [{}])[0].get("output", "")
+    except requests.exceptions.RequestException as e:
+        logger.error(f"🚨 Error querying Gemini API: {e}, Response: {response.text if 'response' in locals() else 'No response'}")
         return "⚠️ An error occurred while generating a response."
 
 # Replace generate_response and query_huggingface with query_gemini_api
@@ -140,8 +141,7 @@ async def chat_with_arina(request: Request, request_body: ChatRequest):
         reset_memory()
         return {"response": "✅ Memory wiped."}
 
-    # Extract facts and retrieve user-specific data
-    extract_and_store_facts(user_input)
+    # Retrieve user-specific data
     user_name = get_user_fact("name")
 
     # Retrieve past relevant conversations