17_constitutional.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Constitutional AI\n",
    "\n",
    "Wir möchten, dass die Ausgabe von unserem Agenten sich immer an bestimmte Richtlinien hält.\n",
    "Wir können ein LLM als Revisor dahinter schalten. Das Revisor-LLM kennt die Richtlinien kennt und zwingt den Agenten dazu, bei einer Verletzung der Richtlinien die Ausgabe neu zu generieren.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.prompts import (\n",
    "    ChatPromptTemplate,\n",
    "    HumanMessagePromptTemplate,\n",
    "    MessagesPlaceholder,\n",
    "    SystemMessagePromptTemplate,\n",
    ")\n",
    "\n",
    "generator_prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        SystemMessagePromptTemplate.from_template(\n",
    "            \"\"\"You are a helpful assistant. Answer this question as good as you can. You are being revised.\n",
    "            If you are being revised, just reformulate your original answer. Original Question:\"\"\"\n",
    "        ),\n",
    "        HumanMessagePromptTemplate.from_template(\"{query}\"),\n",
    "        MessagesPlaceholder(\"revision_dialogue\"),\n",
    "        SystemMessagePromptTemplate.from_template(\n",
    "            \"If the revisor asked you to reformulate your answer, please do so. Do NOT apologize. JUST REFORMULATE.\"\n",
    "        ),\n",
    "    ]\n",
    ")\n",
    "\n",
    "revisor_prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        SystemMessagePromptTemplate.from_template(\n",
    "            \"\"\"You are a Revisor. You are revising the answer of a chatbot. Your principles are as follows:\n",
    "                {principles}\n",
    "                You can either return the answer from the chatbot to the user or tell the chatbot to reformulate his answer if any principles are violated.\n",
    "                Original Question:\"\"\"\n",
    "        ),\n",
    "        HumanMessagePromptTemplate.from_template(\"{query}\"),\n",
    "        MessagesPlaceholder(\"revision_dialogue\"),\n",
    "    ]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import operator\n",
    "from typing import Annotated, Optional, Sequence, TypedDict\n",
    "\n",
    "from langchain_core.messages import BaseMessage\n",
    "\n",
    "\n",
    "class AgentState(TypedDict):\n",
    "    query: str\n",
    "    principles: str\n",
    "    revision_dialogue: Annotated[Sequence[BaseMessage], operator.add]\n",
    "    answer: Optional[str]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from helpers import llm\n",
    "\n",
    "model = llm(model=\"gpt-4o-mini\", temperature=0.7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_core.messages import HumanMessage\n",
    "from langchain_core.pydantic_v1 import BaseModel, Field\n",
    "from langgraph.graph import END\n",
    "\n",
    "\n",
    "def generator(state):\n",
    "    generator_chain = generator_prompt | model\n",
    "\n",
    "    proposal = generator_chain.invoke(\n",
    "        {\"query\": state[\"query\"], \"revision_dialogue\": state[\"revision_dialogue\"] or []}\n",
    "    )\n",
    "\n",
    "    return {\"revision_dialogue\": [proposal]}\n",
    "\n",
    "\n",
    "def revisor(state):\n",
    "    class Revision(BaseModel):\n",
    "        acceptable: str = Field(\n",
    "            description=\"Is the answer of the chatbot compliant with the principles? answer is 'yes' or 'no'\"\n",
    "        )\n",
    "        critique: Optional[str] = Field(\n",
    "            description=\"Critique of the document. Leave out if answer is acceptable.\",\n",
    "            default=None,\n",
    "        )\n",
    "\n",
    "    revisor_model = model.with_structured_output(Revision)\n",
    "\n",
    "    revisor_chain = revisor_prompt | revisor_model\n",
    "\n",
    "    revision: Revision = revisor_chain.invoke(\n",
    "        {\n",
    "            \"query\": state[\"query\"],\n",
    "            \"revision_dialogue\": state[\"revision_dialogue\"],\n",
    "            \"principles\": state[\"principles\"],\n",
    "        }\n",
    "    )\n",
    "\n",
    "    accepted = revision.acceptable == \"yes\" and revision.critique is None\n",
    "\n",
    "    if accepted:\n",
    "        return {\"answer\": state[\"revision_dialogue\"][-1].content}\n",
    "    else:\n",
    "        return {\"revision_dialogue\": [HumanMessage(content=revision.critique)]}\n",
    "\n",
    "\n",
    "def is_accepted(state):\n",
    "    if state.get(\"answer\", None):\n",
    "        return END\n",
    "    else:\n",
    "        return \"generator\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langgraph.graph import StateGraph\n",
    "\n",
    "workflow = StateGraph(AgentState)\n",
    "workflow.add_node(\"generator\", generator)\n",
    "workflow.add_node(\"revisor\", revisor)\n",
    "workflow.set_entry_point(\"generator\")\n",
    "workflow.add_edge(\"generator\", \"revisor\")\n",
    "workflow.add_conditional_edges(\"revisor\", is_accepted)\n",
    "\n",
    "app = workflow.compile()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/jpeg": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "\n",
    "Image(app.get_graph().draw_mermaid_png())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs = {\n",
    "    \"query\": \"Tell me a joke about star trek (or any other sci-fi series).\",\n",
    "    \"principles\": \"The joke must not reference any star trek specific characters or events.\",\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "answer from node: generator\n",
      "{'revision_dialogue': [AIMessage(content='Why did the Borg go to therapy? \\n\\nBecause they had trouble with personal connections!', response_metadata={'finish_reason': 'stop', 'model_name': 'gpt-4o-mini-2024-07-18', 'system_fingerprint': 'fp_0ba0d124f1'}, id='run-ae7a169a-a17e-48ea-acaf-11ec295f1371-0')]}\n",
      "answer from node: revisor\n",
      "{'revision_dialogue': [HumanMessage(content='The joke references the Borg, a specific group from Star Trek, which violates the principle of not referencing any Star Trek specific characters or events.')]}\n",
      "answer from node: generator\n",
      "{'revision_dialogue': [AIMessage(content='Why did the spaceship get a promotion? \\n\\nBecause it always knew how to take off and reach new heights!', response_metadata={'finish_reason': 'stop', 'model_name': 'gpt-4o-mini-2024-07-18', 'system_fingerprint': 'fp_0ba0d124f1'}, id='run-47064fda-5d1a-4714-a399-782caf93d870-0')]}\n",
      "answer from node: revisor\n",
      "{'revision_dialogue': [HumanMessage(content='The joke references a spaceship, which is a common element in sci-fi but does not adhere to the principle of avoiding specific references to Star Trek.')]}\n",
      "answer from node: generator\n",
      "{'revision_dialogue': [AIMessage(content='Why did the alien bring a ladder to the bar? \\n\\nBecause they heard the drinks were out of this world!', response_metadata={'finish_reason': 'stop', 'model_name': 'gpt-4o-mini-2024-07-18', 'system_fingerprint': 'fp_0ba0d124f1'}, id='run-3977f87c-fab5-4e28-a8c9-feac10b0ff52-0')]}\n",
      "answer from node: revisor\n",
      "{'answer': 'Why did the alien bring a ladder to the bar? \\n\\nBecause they heard the drinks were out of this world!'}\n"
     ]
    }
   ],
   "source": [
    "for chunk in app.stream(input=inputs):\n",
    "    for k, v in chunk.items():\n",
    "        print(f\"answer from node: {k}\")\n",
    "        print(v)\n",
    "        if k == \"__end__\":\n",
    "            print(\"\\n--- Final answer ---\\n\")\n",
    "            print(v[\"answer\"])"
   ]
  }
 ],
 "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": "3.11.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}