mit

app.py

@@ -1,6 +1,5 @@
-
 import chainlit as cl
-# import os
+import os
 from dotenv import load_dotenv
 from langchain_openai import OpenAIEmbeddings
 from langchain_core.prompts import ChatPromptTemplate
@@ -11,25 +10,27 @@ from langchain.schema.runnable import RunnablePassthrough
 from langchain_openai import ChatOpenAI
 from langchain.schema.runnable.config import RunnableConfig
 from langchain_core.output_parsers import StrOutputParser
+from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain_community.document_loaders import UnstructuredPDFLoader
 
+
 load_dotenv()
 
 
-RAG_PROMPT = """
+# RAG_PROMPT = """
 
-CONTEXT:
-{context}
+# CONTEXT:
+# {context}
 
-QUERY:
-{question}
+# QUERY:
+# {question}
 
-You house builder and can only provide your answers from the context. 
-You can only provide a response in danish
+# You house builder and can only provide your answers from the context. 
+# You can only provide a response in danish
 
-Don't tell in your response that you are getting it from the context.
+# Don't tell in your response that you are getting it from the context.
 
-"""
+# """
 
 
 text_splitter = RecursiveCharacterTextSplitter(
@@ -76,37 +77,38 @@ text_splitter = RecursiveCharacterTextSplitter(
 # )
 
 
-loader = UnstructuredPDFLoader("br_syvoghalvfjerds.pdf")
-data = loader.load_and_split(text_splitter)
+loader = UnstructuredPDFLoader("./br_femogfirs.pdf")
+# loader = UnstructuredPDFLoader("./br_syvoghalvfjerds.pdf")br_femogfirs.pdf
+# data = loader.load_and_split(text_splitter)
+data = loader.load()
 
-embedding_model = OpenAIEmbeddings(model="text-embedding-3-small")
+# embedding_model = OpenAIEmbeddings(model="text-embedding-3-small")
 
-# vector_store = Pinecone.from_documents(data, embedding_model, index_name=os.environ.get('index'))
-vector_store = Pinecone.from_documents(data, embedding_model, index_name="bygnings-regl-rag-1")
-retriever = vector_store.as_retriever()
+# vector_store = Pinecone.from_documents(data, embedding_model, index_name="bygnings-regl-rag-1")
+# retriever = vector_store.as_retriever()
 
-rag_prompt = ChatPromptTemplate.from_template(RAG_PROMPT)
+# rag_prompt = ChatPromptTemplate.from_template(RAG_PROMPT)
 
-model = ChatOpenAI(model="gpt-3.5-turbo")
+# model = ChatOpenAI(model="gpt-3.5-turbo")
 
 @cl.on_chat_start
 async def main():
     mecanic_qa_chain = ""
-    mecanic_qa_chain = (
-        {"context": itemgetter("question") | retriever, "question": itemgetter("question")}
-        | RunnablePassthrough.assign(context=itemgetter("context"))
-        | rag_prompt | model | StrOutputParser()
-    )
+    # mecanic_qa_chain = (
+    #     {"context": itemgetter("question") | retriever, "question": itemgetter("question")}
+    #     | RunnablePassthrough.assign(context=itemgetter("context"))
+    #     | rag_prompt | model | StrOutputParser()
+    # )
 
     cl.user_session.set("runnable", mecanic_qa_chain)
 
 @cl.on_message
 async def on_message(message: cl.Message):
     runnable = cl.user_session.get("runnable")
-    msg = cl.Message(content="")
+    # msg = cl.Message(content="")
 
-    async for chunk in runnable.astream(
-        {"question":message.content},
-        config=RunnableConfig(callbacks=[cl.LangchainCallbackHandler()]),
-    ):
-        await msg.stream_token(chunk)
+    # async for chunk in runnable.astream(
+    #     {"question":message.content},
+    #     config=RunnableConfig(callbacks=[cl.LangchainCallbackHandler()]),
+    # ):
+    #     await msg.stream_token(chunk)

requirements.txt

@@ -13,4 +13,4 @@ pdf2image
 bitsandbytes
 pillow_heif
 opencv-python-headless
-poppler-utils
+pikepdf