08-hybrid-search.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "import yaml, os, openai, textwrap\n",
    "from langchain.schema import Document\n",
    "from langchain.vectorstores import FAISS\n",
    "from langchain.vectorstores import Chroma\n",
    "from langchain.embeddings import HuggingFaceBgeEmbeddings\n",
    "from langchain.retrievers import BM25Retriever, EnsembleRetriever"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('cadentials.yaml') as f:\n",
    "    credentials = yaml.load(f, Loader=yaml.FullLoader)\n",
    "\n",
    "openai.api_key = credentials['AD_OPENAI_API_KEY']\n",
    "openai.api_base = credentials['AD_OPENAI_API_BASE']\n",
    "openai.api_type = credentials['AD_OPENAI_API_TYPE']\n",
    "openai.api_version = credentials['AD_OPENAI_API_VERSION']\n",
    "openai.engine = credentials['AD_ENGINE']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/1zuu/anaconda3/envs/llm/lib/python3.10/site-packages/bitsandbytes/cextension.py:34: UserWarning: The installed version of bitsandbytes was compiled without GPU support. 8-bit optimizers, 8-bit multiplication, and GPU quantization are unavailable.\n",
      "  warn(\"The installed version of bitsandbytes was compiled without GPU support. \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'NoneType' object has no attribute 'cadam32bit_grad_fp32'\n"
     ]
    }
   ],
   "source": [
    "model_name = \"BAAI/bge-small-en-v1.5\"\n",
    "encode_kwargs = {'normalize_embeddings': True} # set True to compute cosine similarity\n",
    "\n",
    "bge_embeddings = HuggingFaceBgeEmbeddings(\n",
    "                                        model_name=model_name,\n",
    "                                        model_kwargs={'device': 'mps'},\n",
    "                                        encode_kwargs=encode_kwargs\n",
    "                                        )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "doc_list = [\n",
    "            \"I like apples\",\n",
    "            \"I like oranges\",\n",
    "            \"Apples and oranges are fruits\",\n",
    "            \"I like computers by Apple\",\n",
    "            \"I love fruit juice\"\n",
    "            ]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## BM25 Retriever - Sparse retriever"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "bm25_retriever = BM25Retriever.from_texts(doc_list)\n",
    "bm25_retriever.k = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Document(page_content='I like computers by Apple'),\n",
       " Document(page_content='I love fruit juice')]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bm25_retriever.get_relevant_documents(\"Apple\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Document(page_content='I love fruit juice'),\n",
       " Document(page_content='I like computers by Apple')]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bm25_retriever.get_relevant_documents(\"a green fruit\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Embeddings - Dense retrievers FAISS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "faiss_vectorstore = FAISS.from_texts(doc_list, bge_embeddings)\n",
    "faiss_retriever = faiss_vectorstore.as_retriever(search_kwargs={\"k\": 2})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Document(page_content='Apples and oranges are fruits'),\n",
       " Document(page_content='I love fruit juice')]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "faiss_retriever.get_relevant_documents(\"A green fruit\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ensemble Retriever"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "ensemble_retriever = EnsembleRetriever(\n",
    "                                        retrievers=[\n",
    "                                                    bm25_retriever, \n",
    "                                                    faiss_retriever\n",
    "                                                    ],\n",
    "                                       weights=[0.5, 0.5]\n",
    "                                       )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Document(page_content='I love fruit juice'),\n",
       " Document(page_content='Apples and oranges are fruits'),\n",
       " Document(page_content='I like computers by Apple')]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "docs = ensemble_retriever.get_relevant_documents(\"A green fruit\")\n",
    "docs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Document(page_content='I like computers by Apple'),\n",
       " Document(page_content='I love fruit juice'),\n",
       " Document(page_content='I like apples')]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "docs = ensemble_retriever.get_relevant_documents(\"Apple Phones\")\n",
    "docs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}