starter code

Author: alexeygrigorev

.gitignore

@@ -0,0 +1,3 @@
+.ipynb_checkpoints/
+__pycache__/
+minsearch.py

starter.ipynb

@@ -0,0 +1,195 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "d658f909-e679-41e9-9c4e-e0241c719049",
+   "metadata": {},
+   "source": [
+    "If you're not running in Saturn Cloud, you need to install these libraries:\n",
+    "\n",
+    "Make sure you use the latest versions\n",
+    "\n",
+    "```\n",
+    "pip install -U transformers accelerate bitsandbytes\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "506fab2a-a50c-42bd-a106-c83a9d2828ea",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "--2024-06-13 12:33:48--  https://raw.githubusercontent.com/alexeygrigorev/minsearch/main/minsearch.py\n",
+      "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...\n",
+      "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.\n",
+      "HTTP request sent, awaiting response... 200 OK\n",
+      "Length: 3832 (3.7K) [text/plain]\n",
+      "Saving to: 'minsearch.py'\n",
+      "\n",
+      "     0K ...                                                   100%  969K=0.004s\n",
+      "\n",
+      "2024-06-13 12:33:49 (969 KB/s) - 'minsearch.py' saved [3832/3832]\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "!rm -f minsearch.py\n",
+    "!wget https://raw.githubusercontent.com/alexeygrigorev/minsearch/main/minsearch.py"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "3ac947de-effd-4b61-8792-a6d7a133f347",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<minsearch.Index at 0x28d98e5ab10>"
+      ]
+     },
+     "execution_count": 2,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import requests \n",
+    "import minsearch\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)\n",
+    "\n",
+    "index = minsearch.Index(\n",
+    "    text_fields=[\"question\", \"text\", \"section\"],\n",
+    "    keyword_fields=[\"course\"]\n",
+    ")\n",
+    "\n",
+    "index.fit(documents)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "8f087272-b44d-4738-9ea2-175ec63a058b",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def search(query):\n",
+    "    boost = {'question': 3.0, 'section': 0.5}\n",
+    "\n",
+    "    results = index.search(\n",
+    "        query=query,\n",
+    "        filter_dict={'course': 'data-engineering-zoomcamp'},\n",
+    "        boost_dict=boost,\n",
+    "        num_results=5\n",
+    "    )\n",
+    "\n",
+    "    return results"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "742ab881-499a-4675-83c4-2013ea1377b9",
+   "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 = \"\"\n",
+    "    \n",
+    "    for doc in search_results:\n",
+    "        context = context + f\"section: {doc['section']}\\nquestion: {doc['question']}\\nanswer: {doc['text']}\\n\\n\"\n",
+    "    \n",
+    "    prompt = prompt_template.format(question=query, context=context).strip()\n",
+    "    return prompt\n",
+    "\n",
+    "def llm(prompt):\n",
+    "    response = client.chat.completions.create(\n",
+    "        model='gpt-4o',\n",
+    "        messages=[{\"role\": \"user\", \"content\": prompt}]\n",
+    "    )\n",
+    "    \n",
+    "    return response.choices[0].message.content"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "fe8bff3e-b672-42be-866b-f2d9bb217106",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def rag(query):\n",
+    "    search_results = search(query)\n",
+    "    prompt = build_prompt(query, search_results)\n",
+    "    answer = llm(prompt)\n",
+    "    return answer"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "091a77e6-936b-448e-a04b-bad1001f5bb0",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "21aa255e-c971-44ca-9826-a721df3ad063",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "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": "3.11.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}