Update app.py

app.py

@@ -5,7 +5,7 @@ from langchain_community.vectorstores import Chroma
 from langchain.retrievers import MultiQueryRetriever
 from langchain.chains import ConversationalRetrievalChain
 from langchain.memory import ConversationBufferWindowMemory
-from langchain_community.llms import llamacpp, huggingface_pipeline
+from langchain_community.llms import llamacpp, huggingface_hub
 from langchain.prompts import PromptTemplate
 from langchain.chains import LLMChain
 from langchain.chains.question_answering import load_qa_chain
@@ -22,6 +22,15 @@ system_prompt = """You are a helpful assistant, you will use the provided contex
 Read the given context before answering questions and think step by step. If you can not answer a user question based on the provided context, inform the user.
 Do not use any other information for answering the user. Provide a detailed answer to the question."""
 
+def load_llmware_model():
+    return huggingface_hub.HuggingFaceHub(
+        repo_id = "",
+        verbose=True,
+        model_kwargs={
+            'temperature':0.03,
+            'n_batch':128,
+        }
+    )
 def load_quantized_model(model_id=None):
     MODEL_ID, MODEL_BASENAME = "TheBloke/zephyr-7B-beta-GGUF","zephyr-7b-beta.Q5_K_S.gguf"
     try:
@@ -70,6 +79,7 @@ with gr.Blocks() as demo:
         # llm = load_quantized_model(model_id=model_id) #type:ignore
         # ---------------------------------------------------------------------------------------------------
         llm = load_quantized_model()
+        llm_sm = load_llmware_model()
         # ---------------------------------------------------------------------------------------------------
         condense_question_prompt_template = PromptTemplate.from_template(_template)
         prompt_template = system_prompt + """
@@ -80,11 +90,11 @@ with gr.Blocks() as demo:
         memory = ConversationBufferWindowMemory(memory_key='chat_history', k=1, return_messages=True)
         retriever_from_llm = MultiQueryRetriever.from_llm(
                 retriever=db2.as_retriever(search_kwargs={'k':5}),
-                llm = llm,
+                llm = llm_sm,
         )
         qa2 = ConversationalRetrievalChain(
             retriever=retriever_from_llm,
-            question_generator= LLMChain(llm=llm, prompt=condense_question_prompt_template, memory=memory, verbose=True), #type:ignore
+            question_generator= LLMChain(llm=llm_sm, prompt=condense_question_prompt_template, memory=memory, verbose=True), #type:ignore
             combine_docs_chain=load_qa_chain(llm=llm, chain_type="stuff", prompt=qa_prompt, verbose=True), #type:ignore
             memory=memory,
             verbose=True,
@@ -134,4 +144,4 @@ with gr.Blocks() as demo:
 
 if __name__ == "__main__":
     demo.queue()
-    demo.launch(max_threads=10, debug=True)
+    demo.launch(max_threads=8, debug=True, show_error=True)