Update run_model.py

run_model.py

@@ -1,5 +1,5 @@
-# from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
 import os
+from langchain_google_genai import ChatGoogleGenerativeAI
 
 SYSTEM_PROMPT = "You are helpful AI assistant. You will answer user queries truthfully. If you don't know answer to any question you will say so. You will never promote or use any offensive or racist language."
 
@@ -22,41 +22,4 @@ def generate_response(prompt: str, context: str="", history: list = []):
     response = gemma_model.invoke(message)
 
     return response.content
-    
-
-
-# MODEL_NAME = "microsoft/Phi-3.5-mini-instruct"
-
-
-# model = AutoModelForCausalLM.from_pretrained(
-#     MODEL_NAME, 
-#     device_map="auto", 
-#     torch_dtype=torch.float16,
-# )
-# tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
-
-# print(f"Microsoft Phi-3.5-mini-instruct Downloaded Successfully !")
-
-
-# pipe = pipeline(
-#     "text-generation",
-#     model=model,
-#     tokenizer=tokenizer,
-#     max_new_tokens=64, 
-#     temperature=0.4, 
-#     do_sample=True
-# )
-
-# def generate_response(prompt: str, context: str = "", history: list = []) -> str:
-
-
-#     BASE_MESSAGE = [{"role" : "system", "content" : SYSTEM_PROMPT}] + history
-
-#     if context == "":
-#         message = BASE_MESSAGE + [{"role" : "user", "content" : prompt}]
-#     else:
-#         message = BASE_MESSAGE + [{"role" : "user", "content" : f"Context : {context}\nQuery : {prompt}"}]
-
-#     outputs = pipe(message)
-    
-#     return outputs[0]["generated_text"]
+