Upload 8 files

.gitattributes

@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+faiss_index/index.faiss filter=lfs diff=lfs merge=lfs -text
+PoliciesEn001.pdf filter=lfs diff=lfs merge=lfs -text

PoliciesEn001.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0593a94578ac36a2bbcee355bbd81b1514c730d13a251d93b6dd77e4171bc28c
+size 2497751

chain_setup.py

@@ -0,0 +1,26 @@
+# chain_setup.py
+
+from langchain.chains import ConversationalRetrievalChain
+from langchain_community.chat_models import ChatOllama
+from langchain.memory import ConversationBufferMemory
+
+def build_conversational_chain(vectorstore):
+    """
+    Creates a ConversationalRetrievalChain with a ChatOllama LLM and 
+    a ConversationBufferMemory for multi-turn Q&A.
+    """
+    llm = ChatOllama(model="qwen2.5:7b")
+
+    memory = ConversationBufferMemory(
+        memory_key="chat_history",
+        return_messages=True
+    )
+
+    qa_chain = ConversationalRetrievalChain.from_llm(
+        llm=llm,
+        retriever=vectorstore.as_retriever(search_type="similarity", search_kwargs={"k": 5}),
+        memory=memory,
+        verbose=True  # optional debug logs
+    )
+
+    return qa_chain

embedding.py

@@ -0,0 +1,13 @@
+# embedding.py
+
+from langchain_huggingface import HuggingFaceEmbeddings
+
+def load_embeddings():
+    """
+    Returns a HuggingFaceEmbeddings instance.
+    """
+    embeddings = HuggingFaceEmbeddings(
+        model_name="nomic-ai/nomic-embed-text-v1.5",
+        model_kwargs={"trust_remote_code": True}
+    )
+    return embeddings

faiss_index/index.faiss

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2040bb0dc2677cc22d40f55e1c3b92bc67a4a69c6cfb52c61dcd85e7092c47ed
+size 1044525

faiss_index/index.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb8b887e2cc00ef0dd40b895cd502b5028abc1e4198d86f8f354a6b8d13e8706
+size 245411

requirements.txt

@@ -0,0 +1,13 @@
+streamlit==1.25.0
+langchain
+langchain_community
+torch
+transformers
+sentence-transformers
+accelerate
+pypdf
+ollama
+langchain_experimental
+faiss-cpu
+langchain_huggingface
+einops

runtime.txt

@@ -0,0 +1 @@
+python-3.12

vectorstore.py

@@ -0,0 +1,32 @@
+# vectorstore.py
+
+import os
+from langchain_community.document_loaders import PyPDFLoader
+from langchain_experimental.text_splitter import SemanticChunker
+from langchain_community.vectorstores import FAISS
+
+def load_or_build_vectorstore(local_file: str, index_folder: str, embeddings):
+    """
+    Loads a local FAISS index if it exists; otherwise,
+    builds a new index from the specified PDF file.
+    """
+    if os.path.exists(index_folder):
+        print("Loading existing FAISS index from disk...")
+        vectorstore = FAISS.load_local(index_folder, embeddings, allow_dangerous_deserialization=True)
+    else:
+        print("Building a new FAISS index...")
+        loader = PyPDFLoader(local_file)
+        documents = loader.load()
+
+        text_splitter = SemanticChunker(
+            embeddings=embeddings,
+            breakpoint_threshold_type='percentile',
+            breakpoint_threshold_amount=90
+        )
+        chunked_docs = text_splitter.split_documents(documents)
+        print(f"Document split into {len(chunked_docs)} chunks.")
+
+        vectorstore = FAISS.from_documents(chunked_docs, embeddings)
+        vectorstore.save_local(index_folder)
+
+    return vectorstore