Update app.py

app.py

@@ -1,63 +1,88 @@
-import gradio as gr
-from huggingface_hub import InferenceClient
+from langchain.chains import ConversationalRetrievalChain
+from langchain.memory import ConversationBufferMemory, ConversationSummaryMemory
+from langchain.chat_models import ChatOpenAI
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
+from langchain.text_splitter import CharacterTextSplitter, RecursiveCharacterTextSplitter
+from langchain.embeddings.openai import OpenAIEmbeddings
+from langchain.vectorstores import Chroma
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+import os
+import sys
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
 
-    messages.append({"role": "user", "content": message})
+from langchain.schema import (
+    AIMessage,
+    HumanMessage,
+    SystemMessage
+)
 
-    response = ""
+from langchain.prompts import ChatPromptTemplate
+from langchain.prompts.chat import SystemMessagePromptTemplate, HumanMessagePromptTemplate
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+#embeddings = HuggingFaceEmbeddings(model_name='sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2')
+embeddings  = OpenAIEmbeddings(api_key='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')
 
-        response += token
-        yield response
+#vectordb=Chroma.from_documents(document_chunks,embedding=embeddings, persist_directory='./ai_vocacional_v2')
+vectordb = Chroma(persist_directory="./ai_vocacional_v2", embedding_function=embeddings)
 
+llm=ChatOpenAI(temperature=0, model_name='gpt-4o-mini', api_key='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')
+
+memory = ConversationBufferMemory(
+    memory_key='chat_history',
+    return_messages=True)
+
+general_system_template = r"""
+Eres un asistente AI avanzado especializado en asesorar a alumnos con su ingreso a la universidad. 
+Toma los siguientes documentos de contexto {context} y responde únicamente basado en este contexto.
 """
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
 
+general_user_template = "Pregunta:```{question}```"
+messages = [
+            SystemMessagePromptTemplate.from_template(general_system_template),
+            HumanMessagePromptTemplate.from_template(general_user_template)
+]
+qa_prompt = ChatPromptTemplate.from_messages( messages )
+
+pdf_qa = ConversationalRetrievalChain.from_llm(
+            llm = llm,
+            retriever=vectordb.as_retriever(search_kwargs={'k':16})
+            , combine_docs_chain_kwargs={'prompt': qa_prompt},
+            memory = memory#,max_tokens_limit=4000
+            )
+
+#Clarification: Si la pregunta del usuario es vaga o le faltan detalles importantes para dar una respuseta, debes realizar preguntas de clarificación para entender sus necesidades y darle la asistencia adecuada.
+#Constraints: Debes responder solamente con la informacion disponible y saludar solo una vez. En caso no tengas una respuesta o no estés seguro, no inventes respuesta.
+
+import gradio as gr
+# Define chat interface
+with gr.Blocks() as demo:
+    chatbot = gr.Chatbot()
+    msg = gr.Textbox()
+    clear = gr.Button("Clear")
+    chat_history = []
+
+    def user(query, chat_history):
+        print("User query:", query)
+        print("Chat history:", chat_history)
+
+        # Convert chat history to list of tuples
+        chat_history_tuples = []
+        for message in chat_history:
+            chat_history_tuples.append((message[0], message[1]))
+
+        # Get result from QA chain
+        result = pdf_qa({"question": query})#, "chat_history": chat_history_tuples})
+
+        # Append user message and response to chat history
+        chat_history.append((query, result["answer"]))
+        print("Updated chat history:", chat_history)
+
+        return gr.update(value=""), chat_history
+
+
+    msg.submit(user, [msg, chatbot], [msg, chatbot], queue=False)
+    clear.click(lambda: None, None, chatbot, queue=False)
 
 if __name__ == "__main__":
     demo.launch()