Update utils/llm_ag.py

utils/llm_ag.py

@@ -1,60 +1,62 @@
-import requests
-import os
-from dotenv import load_dotenv
-
-load_dotenv()
-
-# Get the API key from environment variable
-GROQ_API_KEY = "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"
-if not GROQ_API_KEY:
-    raise ValueError("GROQ_API_KEY is not set in the .env file")
-
-def intiate_convo(user_query, image_description, additional_text, model="mixtral-8x7b-32768"):
-    # Prepare the message payload
-    messages = [
-        {
-            "role": "system",
-            "content": """You are a AI Assistant for training. Given an image description, additional context, and a user query, respond with a detailed long answer with steps, ,be polite. 
-            IMPORTANT: When referring to the image, subtly acknowledge it by saying "as I see here" rather than explicitly mentioning "image" or "photo." 
-            Your tone should be natural and conversational. Keep it detailed , engaging, and relevant to the query, using both the image description and the additional context as reference points."""
-        },
-        {
-            "role": "user",
-            "content": f"Image description: {image_description}. Additional context: {additional_text}. User query: {user_query}. Provide a detaile response like an ai assistant."
-        }
-    ]
-
-    # Make the API request
-    response = requests.post(
-        "https://api.groq.com/openai/v1/chat/completions",
-        json={
-            "model": model,
-            "messages": messages,
-            "max_tokens": 32768,
-            "stop": None,
-            "stream": False
-        },
-        headers={
-            "Authorization": f"Bearer {GROQ_API_KEY}",
-            "Content-Type": "application/json"
-        },
-        timeout=60
-    )
-
-    # Process the response
-    if response.status_code == 200:
-        result = response.json()
-        answer = result["choices"][0]["message"]["content"]
-        return answer
-    else:
-        return f"Error from LLM API: {response.status_code} - {response.text}"
-
-# # Example usage
-# # Define the inputs
-# user_query = "Can you tell me more about the person in this description?"
-# image_description = """The main subject of the image is a person with dark complexion, short black hair, and white-framed glasses, wearing a dark-colored shirt or jacket. They are looking directly at the camera with a subtle expression."""
-# additional_text = """This individual is a software engineer specializing in AI development. They are known for their expertise in computer vision and enjoy photography as a hobby."""
-
-# # Get the LLM response
-# response = intiate_convo(user_query, image_description, additional_text)
-# print(response)
+import requests
+import os
+from dotenv import load_dotenv
+
+load_dotenv()
+
+# Get the API key from environment variable
+GROQ_API_KEY = "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"
+if not GROQ_API_KEY:
+    raise ValueError("GROQ_API_KEY is not set in the .env file")
+
+def intiate_convo(user_query, image_description, additional_text, model="mixtral-8x7b-32768"):
+    # Prepare the message payload
+    messages = [
+        {
+            "role": "system",
+            "content": """You are a AI Assistant for training. Given an image description, additional context, and a user query, respond with a PRECISE answer WITH THE HELP OF ADDITIONAL CONTEXT,be polite. 
+            IMPORTANT: When referring to the image, subtly acknowledge it by saying "as I see here" rather than explicitly mentioning "image" or "photo." 
+            Your tone should be natural and conversational.relevant to the query, using both the image description and the additional context as reference points.
+            Provide only the answer.
+            """
+        },
+        {
+            "role": "user",
+            "content": f"Image description: {image_description}. Additional context: {additional_text}. User query: {user_query}. Provide a detaile response like an ai assistant."
+        }
+    ]
+
+    # Make the API request
+    response = requests.post(
+        "https://api.groq.com/openai/v1/chat/completions",
+        json={
+            "model": model,
+            "messages": messages,
+            "max_tokens": 32768,
+            "stop": None,
+            "stream": False
+        },
+        headers={
+            "Authorization": f"Bearer {GROQ_API_KEY}",
+            "Content-Type": "application/json"
+        },
+        timeout=60
+    )
+
+    # Process the response
+    if response.status_code == 200:
+        result = response.json()
+        answer = result["choices"][0]["message"]["content"]
+        return answer
+    else:
+        return f"Error from LLM API: {response.status_code} - {response.text}"
+
+# # Example usage
+# # Define the inputs
+# user_query = "Can you tell me more about the person in this description?"
+# image_description = """The main subject of the image is a person with dark complexion, short black hair, and white-framed glasses, wearing a dark-colored shirt or jacket. They are looking directly at the camera with a subtle expression."""
+# additional_text = """This individual is a software engineer specializing in AI development. They are known for their expertise in computer vision and enjoy photography as a hobby."""
+
+# # Get the LLM response
+# response = intiate_convo(user_query, image_description, additional_text)
+# print(response)