mediacharacters.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "cabcb941-f2ef-489c-b05f-746ed39b5713",
   "metadata": {},
   "outputs": [],
   "source": [
    "from datasets import load_dataset, disable_caching\n",
    "from enum import Enum\n",
    "from ollama import chat, pull\n",
    "from pydantic import BaseModel, Field\n",
    "from typing import List, Dict\n",
    "import json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4114ecaa-f71c-43b7-b659-b63ef6dc805d",
   "metadata": {},
   "outputs": [],
   "source": [
    "OLLAMA_MODEL = \"hermes3\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "054d0174-a342-4c43-9afa-93b29626c59b",
   "metadata": {},
   "outputs": [],
   "source": [
    "class Gender(str, Enum):\n",
    "    male = \"male\"\n",
    "    female = \"female\"\n",
    "    agender = \"agender\"\n",
    "    nonbinary = \"nonbinary\"\n",
    "    either = \"character has no inherent gender\"\n",
    "\n",
    "class MediaCharacter(BaseModel):\n",
    "    character_name: str = Field(..., description=\"The name of the character\")\n",
    "    character_gender: Gender = Field(..., description=\"The typical gender of the character, if it can vary answer that it has no inherent gender\")\n",
    "    series_name: str = Field(..., description=\"The series the character normally comes from, with preference to the oldest series\")\n",
    "    genre: List[str] = Field(..., description=\"Two or three lower-case genres the series falls into, if not known then say \\\"unknown\\\"\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "e416b0d4-69a0-4d57-af84-fe92ccaaa23d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\"character_name\":\"Twilight Sparkle\",\"character_gender\":\"female\",\"series_name\":\"My Little Pony: Friendship is Magic\",\"genre\":[\"Fantasy\",\"Adventure\"]}\n",
      "{\"character_name\":\"Fluttershy\",\"character_gender\":\"female\",\"series_name\":\"My Little Pony: Friendship Is Magic\",\"genre\":[\"Animation\",\"Comedy\"]}\n",
      "{\"character_name\":\"Megumin\",\"character_gender\":\"female\",\"series_name\":\"Konosuba\",\"genre\":[\"comedy\",\"fantasy\"]}\n",
      "{\"character_name\":\"Monika\",\"character_gender\":\"female\",\"series_name\":\"Doki Doki Literature Club\",\"genre\":[\"visual novel\",\"horror\"]}\n",
      "{\"character_name\":\"Ashe\",\"character_gender\":\"female\",\"series_name\":\"League of Legends\",\"genre\":[\"fantasy\",\"game\"]}\n",
      "{\"character_name\":\"Claptrap\",\"character_gender\":\"male\",\"series_name\":\"Borderlands\",\"genre\":[\"action\",\"shooter\",\"sci-fi\"]}\n",
      "{\"character_name\":\"Caterpie\",\"character_gender\":\"agender\",\"series_name\":\"Pokémon\",\"genre\":[\"Adventure\",\"Children\"]}\n"
     ]
    }
   ],
   "source": [
    "entries = [\n",
    "    \"Twilight Sparkle from My Little Pony: Friendship is Magic\",\n",
    "    \"Fluttershy\",\n",
    "    \"Megumin in Konosuba\",\n",
    "    \"Monika from DDLC\",\n",
    "    \"Ashe from league\",\n",
    "    \"Claptrap Borderlands\",\n",
    "    \"Caterpie from Viridian Forest\",\n",
    "]\n",
    "results = []\n",
    "\n",
    "for x in entries:\n",
    "    response = chat(\n",
    "        messages = [\n",
    "            {\n",
    "                \"role\": \"system\",\n",
    "                \"content\": f\"\"\"You are a helpful assistant that answers in JSON. Here's the json schema you must adhere to:\n",
    "<schema>\n",
    "{MediaCharacter.model_json_schema()}\n",
    "</schema>\n",
    "\n",
    "You have been given the answer to the following question: What character from which media are you based on?\"\"\"\n",
    "            },\n",
    "            {\n",
    "                \"role\": \"user\",\n",
    "                \"content\": x,\n",
    "            }\n",
    "        ],\n",
    "        model=OLLAMA_MODEL,\n",
    "        format=MediaCharacter.model_json_schema(),\n",
    "    )\n",
    "\n",
    "    character = MediaCharacter.model_validate_json(response.message.content)\n",
    "    results.append(character.model_dump_json())\n",
    "\n",
    "for x in results:\n",
    "    print(x)"
   ]
  }
 ],
 "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": "3.12.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}