Update app.py

app.py

@@ -1,60 +1,39 @@
 import streamlit as st
+from transformers import pipeline
+from collections import deque
 
-st.title("Falcon QA Bot")
-
-# import chainlit as cl
-
-import os
-huggingfacehub_api_token = st.secrets["hf_token"]
-
-from langchain import HuggingFaceHub, PromptTemplate, LLMChain
-
-repo_id = "tiiuae/falcon-7b-instruct"
-llm = HuggingFaceHub(huggingfacehub_api_token=huggingfacehub_api_token, 
-                     repo_id=repo_id, 
-                     model_kwargs={"temperature":0.2, "max_new_tokens":2000})
-
-template = """
-You are an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions.
-
-{question}
-
-"""
-# input = st.text_input("What do you want to ask about", placeholder="Input your question here")
-
-
-# # @cl.langchain_factory
-# def factory():
-#     prompt = PromptTemplate(template=template, input_variables=['question'])
-#     llm_chain = LLMChain(prompt=prompt, llm=llm, verbose=True)
+# Configure system prompt
+system_prompt = "You are an AI assistant that specializes in helping with code-based questions and tasks. Feel free to ask anything related to coding!"
 
-#     return llm_chain
-
-
-prompt = PromptTemplate(template=template, input_variables=["question"])
-llm_chain = LLMChain(prompt=prompt,verbose=True,llm=llm)
-
-# result = llm_chain.predict(question=input)
+st.title("Falcon QA Bot")
 
-# print(result)
+@st.cache(allow_output_mutation=True)
+def get_qa_pipeline():
+    return pipeline("question-answering", model="tiiuae/falcon-7b-instruct", device=0)
 
 def chat(query):
-    # prompt = PromptTemplate(template=template, input_variables=["question"])
-    # llm_chain = LLMChain(prompt=prompt,verbose=True,llm=llm)
-
-    result = llm_chain.predict(question=query)
-
+    pipeline = get_qa_pipeline()
+    result = pipeline(question=query, max_length=2000, context=system_prompt)
     return result
 
-
-
-
 def main():
-    input = st.text_input("What do you want to ask about", placeholder="Input your question here")
-    if input:
-        output = chat(input)
-        st.write(output,unsafe_allow_html=True)
-
+    user_queue = deque()
+    
+    st.markdown('<style>div.row-widget.stRadio > div{flex-direction:row;}</style>', unsafe_allow_html=True)
+    
+    input = st.text_area("What do you want to ask about", value="", height=150, max_chars=500, key="input")
+    if st.button("Ask"):
+        if input:
+            user_queue.append(input)
+    
+    if user_queue:
+        current_user = user_queue[0]
+        st.text_area("System Prompt", value=system_prompt, height=150, disabled=True)
+        st.text_area("User Input", value=current_user, height=150, disabled=True)
+        with st.spinner("Generating response..."):
+            output = chat(current_user)
+        st.text_area("Falcon's Answer", value=output["answer"], height=150, disabled=True)
+        user_queue.popleft()
 
 if __name__ == '__main__':
-    main()
+    main()