llamaindex_milvus_rag_agent.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# first run installations and data download\n",
    "# ! pip install -U llama-index llama-index-vector-stores-milvus pymilvus llama-index-llms-openai llama-index-readers-file\n",
    "# ! mkdir -p './data/10k/'\n",
    "# ! curl 'https://raw.githubusercontent.com/run-llama/llama_index/main/docs/examples/data/10k/uber_2021.pdf' -O 'data/10k/uber_2021.pdf'\n",
    "# ! curl 'https://raw.githubusercontent.com/run-llama/llama_index/main/docs/examples/data/10k/lyft_2021.pdf' -O 'data/10k/lyft_2021.pdf'\n",
    "# ! mv 'lyft_2021.pdf' 'uber_2021.pdf' './data/10k/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Run the following once in your console to install all dependencies\n",
    "# ! pip install -r requirements.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from llama_index.core import (\n",
    "    SimpleDirectoryReader,\n",
    "    VectorStoreIndex,\n",
    "    StorageContext,\n",
    "    load_index_from_storage,\n",
    ")\n",
    "from llama_index.core.tools import QueryEngineTool, ToolMetadata\n",
    "\n",
    "from pymilvus import connections\n",
    "from llama_index.vector_stores.milvus import MilvusVectorStore"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "def is_api_key_valid():\n",
    "    api_key = os.environ.get('OPENAI_API_KEY')\n",
    "    return api_key is not None and api_key.strip() != \"\"\n",
    "\n",
    "if is_api_key_valid():\n",
    "    print(\"Found an OPENAI_API_KEY environment variable.\")\n",
    "else:\n",
    "    print(\"Error: The OPENAI_API_KEY environment variable is not.\")\n",
    "    print(\"Obtain a key from OpenAI, then set it as the OPENAI_API_KEY environment variable.\")\n",
    "    print(\"In bash, you can set the environment variable using the command 'export OPENAI_API_KEY=your_key_starting_SK-'\")\n",
    "    print(\"If using Visual Studio Code, make sure to restart Visual Studio Code after the above steps.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for use with milvus lite, make sure you have installed milvus; aka run the following in your console \"once\"\n",
    "\n",
    "# ! pip install milvus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for use with milvus lite, uncomment the following two lines and run them \"once\" to start the server\n",
    "milvuslite = True\n",
    "\n",
    "if milvuslite:\n",
    "    from milvus import default_server"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "if milvuslite:\n",
    "    if default_server.listen_port:\n",
    "        # If the server is already running, stop and restart it\n",
    "        default_server.stop()\n",
    "        default_server.cleanup()\n",
    "        default_server.start()\n",
    "    else:\n",
    "        # Otherwise, start the server\n",
    "        default_server.start()\n",
    "\n",
    "    print(f\"The Milvus Lite server is running on port {default_server.listen_port}\")\n",
    "else:\n",
    "    print(\"Milvus lite is not enabled. Adjust the value of 'milvuslite' in the earlier block if you want to run a local instance.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "try:\n",
    "    vector_store_lyft = MilvusVectorStore(dim=1536, collection_name=\"lyft\", overwrite=False)\n",
    "    vector_store_uber = MilvusVectorStore(dim=1536, collection_name=\"uber\", overwrite=False)\n",
    "    storage_context = StorageContext.from_defaults(\n",
    "        persist_dir=\"./storage/lyft\", vector_store=vector_store_lyft\n",
    "    )\n",
    "    lyft_index = load_index_from_storage(storage_context)\n",
    "    storage_context = StorageContext.from_defaults(\n",
    "        persist_dir=\"./storage/uber\", vector_store=vector_store_uber\n",
    "    )\n",
    "    uber_index = load_index_from_storage(storage_context)\n",
    "\n",
    "    index_loaded = True\n",
    "except:\n",
    "    index_loaded = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "if not index_loaded:\n",
    "    # load data\n",
    "    lyft_docs = SimpleDirectoryReader(\n",
    "        input_files=[\"./data/10k/lyft_2021.pdf\"]\n",
    "    ).load_data()\n",
    "    uber_docs = SimpleDirectoryReader(\n",
    "        input_files=[\"./data/10k/uber_2021.pdf\"]\n",
    "    ).load_data()\n",
    "\n",
    "    # build index\n",
    "    vector_store_lyft = MilvusVectorStore(dim=1536, collection_name=\"lyft\", overwrite=True)\n",
    "    vector_store_uber = MilvusVectorStore(dim=1536, collection_name=\"uber\", overwrite=True)\n",
    "    \n",
    "    # for milvus lite users\n",
    "    if milvuslite:\n",
    "        vector_store_lyft = MilvusVectorStore(host=\"localhost\", port=default_server.listen_port, dim=1536, collection_name=\"lyft\", overwrite=True)\n",
    "        vector_store_uber = MilvusVectorStore(host=\"localhost\", port=default_server.listen_port, dim=1536, collection_name=\"uber\", overwrite=True)\n",
    "    \n",
    "    storage_context_lyft = StorageContext.from_defaults(vector_store=vector_store_lyft)\n",
    "    storage_context_uber = StorageContext.from_defaults(vector_store=vector_store_uber)\n",
    "    lyft_index = VectorStoreIndex.from_documents(lyft_docs, storage_context=storage_context_lyft)\n",
    "    uber_index = VectorStoreIndex.from_documents(uber_docs, storage_context=storage_context_uber)\n",
    "\n",
    "    # persist index\n",
    "    lyft_index.storage_context.persist(persist_dir=\"./storage/lyft\")\n",
    "    uber_index.storage_context.persist(persist_dir=\"./storage/uber\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "lyft_engine = lyft_index.as_query_engine(similarity_top_k=3)\n",
    "uber_engine = uber_index.as_query_engine(similarity_top_k=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "query_engine_tools = [\n",
    "    QueryEngineTool(\n",
    "        query_engine=lyft_engine,\n",
    "        metadata=ToolMetadata(\n",
    "            name=\"lyft_10k\",\n",
    "            description=(\n",
    "                \"Provides information about Lyft financials for year 2021. \"\n",
    "                \"Use a detailed plain text question as input to the tool.\"\n",
    "            ),\n",
    "        ),\n",
    "    ),\n",
    "    QueryEngineTool(\n",
    "        query_engine=uber_engine,\n",
    "        metadata=ToolMetadata(\n",
    "            name=\"uber_10k\",\n",
    "            description=(\n",
    "                \"Provides information about Uber financials for year 2021. \"\n",
    "                \"Use a detailed plain text question as input to the tool.\"\n",
    "            ),\n",
    "        ),\n",
    "    ),\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from llama_index.core.agent import ReActAgent\n",
    "from llama_index.llms.openai import OpenAI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "llm = OpenAI(model=\"gpt-3.5-turbo-0613\")\n",
    "\n",
    "agent = ReActAgent.from_tools(\n",
    "    query_engine_tools,\n",
    "    llm=llm,\n",
    "    verbose=True,\n",
    "    # context=context\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "response = agent.chat(\"What was Lyft's revenue growth in 2021?\")\n",
    "print(str(response))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "hw_milvus",
   "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": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}