Merge pull request #938 from guardrails-ai/ml_based_validator_docs

ML Based validator concept doc

docs/concepts/ml_based_validators.ipynb

@@ -0,0 +1,143 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# ML Based Validators\n",
+    "\n",
+    "Although simple validation can often be handled with rule-based approaches, more complex scenarios require machine learning models for effective validation. \n",
+    "\n",
+    "ML based validators are able to handle more complex scenarios, providing some level of 'intelligence' to the validation method that is used."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "In the Guardrails Hub, we provide an easy way to filter and search for different validator types!"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "![Validator Hub Filtering](img/infra_filter.png)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "While different validators may be rules-based or machine learning based, it is still just as easy to implement both in your codebase."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from guardrails import Guard\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "vscode": {
+     "languageId": "shellscript"
+    }
+   },
+   "source": [
+    "```bash\n",
+    "guardrails hub install hub://guardrails/competitor_check\n",
+    "guardrails hub install hub://guardrails/regex_match\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Rules-based validators"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from guardrails.hub import RegexMatch\n",
+    "\n",
+    "guard = Guard().use(\n",
+    "    RegexMatch, \n",
+    "    regex=r\"^[a-zA-Z0-9_]+$\"\n",
+    ")\n",
+    "\n",
+    "result = guard(\n",
+    "    model=\"gpt-3.5-turbo\",\n",
+    "    messages=[\n",
+    "        {\"role\": \"system\", \"content\": \"You are a helpful assistant.\"},\n",
+    "        {\"role\": \"user\", \"content\": \"Tell me about the Apple Iphone.\"},\n",
+    "    ],\n",
+    "    max_tokens=1024,\n",
+    "    temperature=0,\n",
+    ")\n",
+    "print(result)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### ML Based Validator"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from guardrails.hub import CompetitorCheck\n",
+    "guard = Guard().use(\n",
+    "    CompetitorCheck, \n",
+    "    [\"Apple\"]\n",
+    ")\n",
+    "\n",
+    "result = guard(\n",
+    "    model=\"gpt-3.5-turbo\",\n",
+    "    messages=[\n",
+    "        {\"role\": \"system\", \"content\": \"You are a helpful assistant.\"},\n",
+    "        {\"role\": \"user\", \"content\": \"Tell me about the Apple Iphone.\"},\n",
+    "    ],\n",
+    "    max_tokens=1024,\n",
+    "    temperature=0,\n",
+    ")\n",
+    "print(result)"
+   ]
+  }
+ ],
+ "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": "3.11.8"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}