RAG using Chroma Langchain

README.md

@@ -18,4 +18,8 @@
     HF_PASS=your-password
     ```
 
-Now you can run the chatbot and interact with it.
+Now you can run the chatbot and interact with it.
+
+
+
+https://github.com/langchain-ai/langchain/issues/6628#issuecomment-1935374689

main.py

@@ -10,39 +10,75 @@ from langchain_core.runnables import RunnablePassthrough
 from langchain_core.documents import Document
 from langchain_core.prompts import ChatPromptTemplate
 from langchain_core.output_parsers import StrOutputParser
+
 # from langchain_community.chains import 
 from langchain_community.chat_models import ChatOllama
 from langchain_chroma import Chroma
 from hugchat import hugchat
+# from langchain.callbacks import SystemMessage
 from hugchat.login import Login
 import dotenv
 from utils import HuggingChat
-from langchain import PromptTemplate
+from langchain_core.prompts import PromptTemplate
+from langchain_community.embeddings import HuggingFaceEmbeddings
+import langchain
+langchain.debug = True
 
 dotenv.load_dotenv()
 
 
 class GradioApp:
     def __init__(self):
+
+        self.history = []
         # self.llm = ChatOllama(model="phi3:3.8b", base_url="http://localhost:11434", num_gpu=32)
         
 
-        template = """
-        You are a helpful health assistant. These Human will ask you a questions about their pregnancy health.
-        Use following piece of context to answer the question.
-        If you don't know the answer, just say you don't know.
-        Keep the answer within 2 sentences and concise.
+#         template = """
+# You are a helpful health assistant. These Human will ask you a questions about their pregnancy health.
+# Use following piece of context to answer the question.
+# If you don't know the answer, just say you don't know.
+# Keep the answer within 2 sentences and concise.
+
+# Context: {context}
+# Question: {question}
+# Answer: """
+
+
+        self.template = """
+You are a helpful AI bot that guides the customer or user through the website content and provides the user with exact details they want.
+You help everyone by answering questions, and improve your answers from previous answers in History.
+Don't try to make up an answer, if you don't know, just say that you don't know.
+Answer in the same language the question was asked.
+Answer in a way that is easy to understand.
+Try to limit the answer to 3-4 sentences.
+Do not say "Based on the information you provided, ..." or "I think the answer is...". Just answer the question directly in detail.
 
-        Context: {context}
-        Question: {question}
-        Answer:
+History: {chat_history}
 
-        """
+Context: {context}
 
+Question: {question}
+Answer: 
+"""
+        self.prompt = PromptTemplate(
+            template=self.template,
+            input_variables=["chat_history","context", "question"]
+        )
+        self.db = Chroma(persist_directory="./pragetx_chroma", embedding_function=HuggingFaceEmbeddings())  
         self.llm = HuggingChat(email = os.getenv("HF_EMAIL") , psw = os.getenv("HF_PASS") )
-        self.chain = (self.llm | StrOutputParser())
+        self.chain = (
+            {"chat_history": self.chat_history, "context": self.db.as_retriever(k=1), "question": RunnablePassthrough()} |
+            self.prompt |
+            self.llm | 
+            StrOutputParser())
+    def chat_history(self, history):
+        print(self.history)
+        print("\n".join(f"##Human: {x[0]}\n{'##Bot: '+x[1] if x[1] else ''}" for x in self.history))
+        return "\n".join(f"##Human: {x[0]}\n{'##Bot: '+x[1] if x[1] else ''}" for x in self.history)
 
     def user(self,user_message, history):
+        self.history = history + [[user_message, None]]
         return "", history + [[user_message, None]]
     
     def bot(self,history):
@@ -53,7 +89,8 @@ class GradioApp:
             history[-1][1] += chunks
             yield history
         history[-1][1] = history[-1][1] or ""
-        history[-1][1] += self.chain.invoke(prompt)
+        self.history = history
+        # history[-1][1] += self.chain.invoke(prompt)
         print(history[-1][1])
         print(history)
         return history

setup.py

@@ -6,7 +6,7 @@ from langchain_community.embeddings import HuggingFaceEmbeddings
 
 loader = TextLoader('./pragetx.md')
 documents = loader.load()
-text_splitter = CharacterTextSplitter(chunk_size=4000, chunk_overlap=4)
+text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=4)
 docs = text_splitter.split_documents(documents)
 
 embeddings = HuggingFaceEmbeddings()