Cleaned up codebase and populated README page.

README.md

@@ -1 +1,52 @@
 # rag-tutorial-v2
+
+This repository is a starter guide for learning how to use Retrieval-Augmented Generation (RAG) with Ollama. Follow the steps below to set up and run the project.
+
+## Prerequisites
+
+Ensure you have the following installed:
+- Python 3.8 or higher
+- pip (Python package installer)
+
+## Installation
+
+1. Clone the repository:
+    ```sh
+    git clone <repository-url>
+    cd rag-tutorial-v2
+    ```
+
+2. Install the required Python packages:
+    ```sh
+    pip install -r requirements.txt
+    ```
+
+## Setting Up the Database
+
+1. Place your PDF documents in the [data](http://_vscodecontentref_/0) directory.
+
+2. Populate the database with the documents:
+    ```sh
+    python populate_database.py --reset
+    ```
+
+## Querying the Database
+
+To query the database, run the [query_data.py](http://_vscodecontentref_/1) script with your query text:
+```sh
+python query_data.py "Your query text here"
+```
+## Project Structure
+
+- `__pycache__/`: Directory containing Python bytecode files.
+- `.gitignore`: Git ignore file to specify untracked files to ignore.
+- `chroma/`: Directory containing the Chroma database files.
+  - `77943a61-1aef-45fe-af55-a29fb829e25b/`: Subdirectory for Chroma database.
+  - `chroma.sqlite3`: SQLite database file for Chroma.
+- `data/`: Directory to store PDF documents.
+- `get_embedding_function.py`: Script to get the embedding function.
+- `populate_database.py`: Script to populate the Chroma database with documents.
+- `query_data.py`: Script to query the Chroma database.
+- `README.md`: This README file.
+- `requirements.txt`: List of required Python packages.
+- `test_rag.py`: Script to run tests on the RAG setup.

get_embedding_function.py

@@ -1,10 +1,10 @@
-from langchain_community.embeddings.ollama import OllamaEmbeddings
-from langchain_community.embeddings.bedrock import BedrockEmbeddings
+from langchain_ollama import OllamaEmbeddings
+from langchain_aws.embeddings.bedrock import BedrockEmbeddings
 
 
 def get_embedding_function():
-    embeddings = BedrockEmbeddings(
-        credentials_profile_name="default", region_name="us-east-1"
-    )
-    # embeddings = OllamaEmbeddings(model="nomic-embed-text")
+    # embeddings = BedrockEmbeddings(
+    #     credentials_profile_name="default", region_name="us-east-1"
+    # )
+    embeddings = OllamaEmbeddings(model="mxbai-embed-large")
     return embeddings

populate_database.py

@@ -1,11 +1,11 @@
 import argparse
 import os
 import shutil
-from langchain.document_loaders.pdf import PyPDFDirectoryLoader
+from langchain_community.document_loaders import PyPDFDirectoryLoader
 from langchain_text_splitters import RecursiveCharacterTextSplitter
 from langchain.schema.document import Document
 from get_embedding_function import get_embedding_function
-from langchain.vectorstores.chroma import Chroma
+from langchain_chroma import Chroma
 
 
 CHROMA_PATH = "chroma"
@@ -67,7 +67,6 @@ def add_to_chroma(chunks: list[Document]):
         print(f"👉 Adding new documents: {len(new_chunks)}")
         new_chunk_ids = [chunk.metadata["id"] for chunk in new_chunks]
         db.add_documents(new_chunks, ids=new_chunk_ids)
-        db.persist()
     else:
         print("✅ No new documents to add")
 

query_data.py

@@ -1,7 +1,7 @@
 import argparse
-from langchain.vectorstores.chroma import Chroma
+from langchain_chroma import Chroma
 from langchain.prompts import ChatPromptTemplate
-from langchain_community.llms.ollama import Ollama
+from langchain_ollama import OllamaLLM
 
 from get_embedding_function import get_embedding_function
 
@@ -40,7 +40,7 @@ def query_rag(query_text: str):
     prompt = prompt_template.format(context=context_text, question=query_text)
     # print(prompt)
 
-    model = Ollama(model="mistral")
+    model = OllamaLLM(model="mistral")
     response_text = model.invoke(prompt)
 
     sources = [doc.metadata.get("id", None) for doc, _score in results]

requirements.txt

@@ -1,5 +1,8 @@
 pypdf
 langchain
+langchain_ollama
+langchain_aws
+langchain_community
 chromadb # Vector storage
 pytest
-boto3
+boto3

test_rag.py

@@ -1,5 +1,5 @@
 from query_data import query_rag
-from langchain_community.llms.ollama import Ollama
+from langchain_ollama import OllamaLLM
 
 EVAL_PROMPT = """
 Expected Response: {expected_response}
@@ -29,7 +29,7 @@ def query_and_validate(question: str, expected_response: str):
         expected_response=expected_response, actual_response=response_text
     )
 
-    model = Ollama(model="mistral")
+    model = OllamaLLM(model="mistral")
     evaluation_results_str = model.invoke(prompt)
     evaluation_results_str_cleaned = evaluation_results_str.strip().lower()
 
@@ -47,3 +47,4 @@ def query_and_validate(question: str, expected_response: str):
         raise ValueError(
             f"Invalid evaluation result. Cannot determine if 'true' or 'false'."
         )
+