01-intro/homework-1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import requests \n",
    "\n",
    "docs_url = 'https://github.com/DataTalksClub/llm-zoomcamp/blob/main/01-intro/documents.json?raw=1'\n",
    "docs_response = requests.get(docs_url)\n",
    "documents_raw = docs_response.json()\n",
    "\n",
    "documents = []\n",
    "\n",
    "for course in documents_raw:\n",
    "    course_name = course['course']\n",
    "\n",
    "    for doc in course['documents']:\n",
    "        doc['course'] = course_name\n",
    "        documents.append(doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ObjectApiResponse({'acknowledged': True, 'shards_acknowledged': True, 'index': 'course_homework'})"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from elasticsearch import Elasticsearch\n",
    "\n",
    "es_client = Elasticsearch(\"http://localhost:9200\")\n",
    "index_name = \"course_homework\"\n",
    "index_settings = {\n",
    "    \"settings\": {\n",
    "        \"number_of_shards\": 1,\n",
    "        \"number_of_replicas\": 0\n",
    "    },\n",
    "    \"mappings\": {\n",
    "        \"properties\": {\n",
    "            \"text\": {\"type\": \"text\"},\n",
    "            \"section\": {\"type\": \"text\"},\n",
    "            \"question\": {\"type\": \"text\"},\n",
    "            \"course\": {\"type\": \"keyword\"} \n",
    "        }\n",
    "    }\n",
    "}\n",
    "\n",
    "es_client.indices.create(index=index_name, body=index_settings)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Q2. Indexing the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ahmedk/Courses/Bootcamps/llm-zoomcamp/.venv/lib/python3.9/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n",
      "100%|██████████| 948/948 [00:05<00:00, 161.44it/s]\n"
     ]
    }
   ],
   "source": [
    "from tqdm.auto import tqdm\n",
    "for doc in tqdm(documents):\n",
    "    es_client.index(index=index_name, document=doc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def elastic_search(query):\n",
    "\n",
    "    search_query = {\n",
    "        \"size\": 1,\n",
    "        \"query\": {\n",
    "            \"bool\": {\n",
    "                \"must\": {\n",
    "                    \"multi_match\": {\n",
    "                        \"query\": query,\n",
    "                        \"fields\": [\"question^4\", \"text\"],\n",
    "                        \"type\": \"best_fields\"\n",
    "                    }\n",
    "                },\n",
    "            }\n",
    "        }\n",
    "    }\n",
    "    return es_client.search(index=index_name, body=search_query)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def find_scores(response):\n",
    "    result_scores = []\n",
    "    for hit in response['hits']['hits']:\n",
    "        result_scores.append(hit[\"_score\"])\n",
    "    return result_scores"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Q3. Searching"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top ranking result score: 81.68475\n"
     ]
    }
   ],
   "source": [
    "query = \"How do I execute a command in a running docker container?\"\n",
    "results = elastic_search(query)\n",
    "print(\"Top ranking result score:\", find_scores(results)[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Q4. Filtering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def elastic_search(query):\n",
    "\n",
    "    search_query = {\n",
    "        \"size\": 3,\n",
    "        \"query\": {\n",
    "            \"bool\": {\n",
    "                \"must\": {\n",
    "                    \"multi_match\": {\n",
    "                        \"query\": query,\n",
    "                        \"fields\": [\"question^4\", \"text\"],\n",
    "                        \"type\": \"best_fields\"\n",
    "                    }\n",
    "                },\n",
    "                \"filter\": {\n",
    "                    \"term\": {\n",
    "                        \"course\": \"machine-learning-zoomcamp\"\n",
    "                    }\n",
    "                }\n",
    "            }\n",
    "        }\n",
    "    }\n",
    "    return es_client.search(index=index_name, body=search_query)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3rd question from the response How do I copy files from a different folder into docker container’s working directory?\n"
     ]
    }
   ],
   "source": [
    "response = elastic_search(query)\n",
    "print(f\"3rd question from the response\", response['hits']['hits'][2]['_source'][\"question\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_results(response):\n",
    "    result_docs = []\n",
    "    for hit in response['hits']['hits']:\n",
    "        result_docs.append(hit[\"_source\"])\n",
    "    return result_docs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Q5. Building a prompt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_prompt(query, search_results):\n",
    "    prompt_template = \"\"\"\n",
    "    You're a course teaching assistant. Answer the QUESTION based on the CONTEXT from the FAQ database.\n",
    "    Use only the facts from the CONTEXT when answering the QUESTION.\n",
    "\n",
    "    QUESTION: {question}\n",
    "\n",
    "    CONTEXT:\n",
    "    {context}\n",
    "    \"\"\".strip()\n",
    "\n",
    "    context_template = \"\"\"\n",
    "    Q: {question}\n",
    "    A: {text}\n",
    "    \"\"\".strip()\n",
    "\n",
    "    context = \"\\n\\n\".join(\n",
    "        context_template.format(\n",
    "            question=doc['question'], text=doc['text'])for doc in search_results)\n",
    "    return prompt_template.format(question=query, context=context)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "prompt length is: 1490\n"
     ]
    }
   ],
   "source": [
    "response = elastic_search(query)\n",
    "results = get_results(response)\n",
    "prompt = build_prompt(query, results)\n",
    "print(\"prompt length is:\", len(prompt))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Q6. Tokens"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def llm(prompt):\n",
    "    from openai import OpenAI\n",
    "\n",
    "    client = OpenAI()\n",
    "    response = client.chat.completions.create(\n",
    "        model='gpt-4o',\n",
    "        messages=[{'role':'user', 'content': prompt}],\n",
    "    )\n",
    "    return response.choices[0].message.content"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of tokens in our prompt is  329\n"
     ]
    }
   ],
   "source": [
    "import tiktoken\n",
    "encoding = tiktoken.encoding_for_model(\"gpt-4o\")\n",
    "num_sent_tokens = len(encoding.encode(prompt))\n",
    "print(\"Number of tokens in our prompt is \", num_sent_tokens)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Bonus: generating the answer (ungraded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'To execute a command in a running Docker container, follow these steps:\\n\\n1. Use the `docker ps` command to list all running containers and find the container ID of the specific container you want to execute a command in.\\n   ```sh\\n   docker ps\\n   ```\\n\\n2. Use the `docker exec -it` command followed by the container ID and the command you want to execute. For example, to start a bash session in the container, you would use:\\n   ```sh\\n   docker exec -it <container-id> bash\\n   ```\\n\\nThese steps will allow you to run commands in your desired container.'"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def rag(query):\n",
    "    response = elastic_search(query)\n",
    "    results = get_results(response)\n",
    "    prompt = build_prompt(query, results)\n",
    "    return llm(prompt)\n",
    "openai_resp = rag(query)\n",
    "num_recv_token = len(encoding.encode(openai_resp))\n",
    "openai_resp"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Bonus: calculating the costs (ungraded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Average cost is 4.5$'"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "avg_cost = (150*1000)*(0.005/1000) + (250*1000)*(0.015/1000)\n",
    "f\"Average cost is {avg_cost}$\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Homework cost is 0.00355$'"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q5_q6_cost = num_sent_tokens*(0.005/1000) + num_recv_token*(0.015/1000)\n",
    "f\"Homework cost is {q5_q6_cost}$\""
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.9.19"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}