Update app.py

app.py

@@ -52,7 +52,7 @@ hf_token = os.environ.get('hf_token')
 GEMINI_API_KEY = os.environ.get('GEMINI_API_KEY')
 fs_token = os.environ.get('fs_token')
 
-llm_name = "gpt-3.5-turbo"
+llm_name = "gpt-3.5-turbo-0301"
 
 vectordb = initialize.initialize()
 
@@ -61,45 +61,45 @@ vectordb = initialize.initialize()
 
 
 
-quantization_config = {
-    "load_in_4bit": True,
-    "bnb_4bit_compute_dtype": torch.float16,
-    "bnb_4bit_quant_type": "nf4",
-    "bnb_4bit_use_double_quant": True,
-}
-
-pipeline = pipeline(
-    "text-generation",
-    model=model_4bit,
-    tokenizer=tokenizer,
-    use_cache=True,
-    device='cpu',  # '0' is for GPU, 'cpu' for CPU
-    max_length=500,
-    do_sample=True,
-    top_k=5,
-    num_return_sequences=1,
-    eos_token_id=tokenizer.eos_token_id,
-    pad_token_id=tokenizer.eos_token_id,
-)
+# quantization_config = {
+#     "load_in_4bit": True,
+#     "bnb_4bit_compute_dtype": torch.float16,
+#     "bnb_4bit_quant_type": "nf4",
+#     "bnb_4bit_use_double_quant": True,
+# }
+
+# pipeline = pipeline(
+#     "text-generation",
+#     model=model_4bit,
+#     tokenizer=tokenizer,
+#     use_cache=True,
+#     device=0,  # '0' is for GPU, 'cpu' for CPU
+#     max_length=500,
+#     do_sample=True,
+#     top_k=5,
+#     num_return_sequences=1,
+#     eos_token_id=tokenizer.eos_token_id,
+#     pad_token_id=tokenizer.eos_token_id,
+# )
 
 
-llm = HuggingFacePipeline(pipeline=pipeline)
-model_id = "mistralai/Mistral-7B-Instruct-v0.1"
-model_4bit = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=quantization_config)
-tokenizer = AutoTokenizer.from_pretrained(model_id)
+# llm = HuggingFacePipeline(pipeline=pipeline)
+# model_id = "mistralai/Mistral-7B-Instruct-v0.1"
+# model_4bit = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=quantization_config)
+# tokenizer = AutoTokenizer.from_pretrained(model_id)
 
-# template = """[INST] You are a helpful, respectful and honest assistant. Answer exactly in few words from the context
-# Answer the question below from the context below:
-# {context}
-# {question} [/INST]
-# """
+# # template = """[INST] You are a helpful, respectful and honest assistant. Answer exactly in few words from the context
+# # Answer the question below from the context below:
+# # {context}
+# # {question} [/INST]
+# # """
 
 
-def chat_query(message, history):
+# def chat_query(message, history):
     
-    retrieverQA = RetrievalQA.from_chain_type(llm=llm, chain_type="retrieval", retriever=vectordb.as_retriever(), verbose=True)
-    result = retrieverQA.run()
-    return result
+#     retrieverQA = RetrievalQA.from_chain_type(llm=llm, chain_type="retrieval", retriever=vectordb.as_retriever(), verbose=True)
+#     result = retrieverQA.run()
+#     return result
 
 
 
@@ -107,23 +107,23 @@ def chat_query(message, history):
 #-------------------------------------------
 
     
-# def chat_query(question, history):
+def chat_query(question, history):
     
-#     llm = ChatOpenAI(model=llm_name, temperature=0.1, api_key = OPENAI_API_KEY)
+    llm = ChatOpenAI(model=llm_name, temperature=0.1, api_key = OPENAI_API_KEY)
 
-#     # Conversation Retrival Chain with Memory
-#     memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
-#     retriever=vectordb.as_retriever()
-#     qa = ConversationalRetrievalChain.from_llm(llm, retriever=retriever, memory=memory)
+    # Conversation Retrival Chain with Memory
+    memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
+    retriever=vectordb.as_retriever()
+    qa = ConversationalRetrievalChain.from_llm(llm, retriever=retriever, memory=memory)
 
-#     # Replace input() with question variable for Gradio
-#     result = qa({"question": question})
-#     return result['answer']
+    # Replace input() with question variable for Gradio
+    result = qa({"question": question})
+    return result['answer']
 
-#     # Chatbot only answers based on Documents
-#     # qa = VectorDBQA.from_chain_type(llm=OpenAI(openai_api_key = OPENAI_API_KEY, ), chain_type="stuff", vectorstore=vectordb)
-#     # result = qa.run(question)
-#     # return result
+    # Chatbot only answers based on Documents
+    # qa = VectorDBQA.from_chain_type(llm=OpenAI(openai_api_key = OPENAI_API_KEY, ), chain_type="stuff", vectorstore=vectordb)
+    # result = qa.run(question)
+    # return result