Update app.py

app.py

@@ -80,16 +80,15 @@ from transformers import AutoTokenizer
 tokenizer = AutoTokenizer.from_pretrained(llm_model)
 
 print("check2")
-pipe = pipeline(model = llm_model, tokenizer = tokenizer, task="text-generation",chat = qa_chat_prompt)
+pipe = pipeline(model = llm_model, tokenizer = tokenizer, task = "question-answering", temperature=0.2)
 #"question-answering", "conversational"
 
 print("check3")
-
-#chain = qa_chat_prompt | pipe
-
+chain = pipe(qa_chat_prompt,retriever) 
+print("check3A")
 import gradio as gr
 #ragdemo = gr.load("models/HuggingFaceH4/zephyr-7b-beta")
-ragdemo = gr.Interface.from_pipeline(pipe)
+ragdemo = gr.Interface.from_pipeline(chain)
 print("check4")
 ragdemo.launch()
 print("check5")