kap10-openai-api.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/DJCordhose/buch-machine-learning-notebooks/blob/master/kap10-openai-api.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "912d7d82",
      "metadata": {
        "id": "912d7d82"
      },
      "source": [
        "# LLMS - OpenAI GPT API Basics\n",
        "\n",
        "* zum Ausführen ist ein OpenAI-Account notwendig: https://platform.openai.com/signup\n",
        "* hier https://platform.openai.com/api-keys den eigenen Key erzeugen und eintragen\n",
        "* Kosten\n",
        "  * https://openai.com/pricing\n",
        "  * https://platform.openai.com/usage\n",
        "* Installation und Quickstart: https://platform.openai.com/docs/quickstart?context=python  \n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "id": "BhKtZIx7BC8S",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BhKtZIx7BC8S",
        "outputId": "e7f50872-c8dd-4848-eba7-78fe7f3f9831"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {},
          "execution_count": 1
        }
      ],
      "source": [
        "import sys\n",
        "IN_COLAB = 'google.colab' in sys.modules # True if running in Google Colab.\n",
        "IN_COLAB"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "NnkgKmBYxFhH"
      },
      "outputs": [],
      "source": [
        "if IN_COLAB:\n",
        "    !pip install -q openai tiktoken"
      ],
      "id": "NnkgKmBYxFhH"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bha4IwxTxFhH"
      },
      "source": [
        "# Tokenizer\n",
        "\n",
        "* Tokenizer zerlegen Texte in Zahlen\n",
        "* diese Zahlen korrespondieren mit Token\n",
        "* Token könnten ganze Worte, Wortteile, Ziffern, Sonderzeichen oder Satzzeiten sein\n",
        "\n",
        "Links\n",
        "* https://github.com/openai/tiktoken\n",
        "* https://platform.openai.com/tokenizer"
      ],
      "id": "bha4IwxTxFhH"
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "xVppMHZdxFhH",
        "outputId": "795a1603-4da0-4e00-b1ed-12924b26e231",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[79178, 11, 46066, 0, 15]"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ],
      "source": [
        "import tiktoken\n",
        "\n",
        "# Dies ist der Tokenizer für GPT-4:\n",
        "enc = tiktoken.encoding_for_model(\"gpt-4\")\n",
        "assert enc.decode(enc.encode(\"Hallo, Welt\")) == \"Hallo, Welt\"\n",
        "enc.encode(\"Hallo, Welt!0\")"
      ],
      "id": "xVppMHZdxFhH"
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iNr6IVnwxFhI"
      },
      "source": [
        "# OpenAI Client"
      ],
      "id": "iNr6IVnwxFhI"
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "LIhBdwOXxFhI"
      },
      "outputs": [],
      "source": [
        "# TODO: den Key eintragen\n",
        "OPENAI_API_KEY = '...'\n",
        "\n",
        "\n",
        "assert OPENAI_API_KEY != '...', \"Bitte tragen Sie Ihren OpenAI API Key ein!\""
      ],
      "id": "LIhBdwOXxFhI"
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "IS-E6uzlxFhI"
      },
      "outputs": [],
      "source": [
        "from openai import OpenAI\n",
        "\n",
        "client = OpenAI(\n",
        "  api_key=OPENAI_API_KEY\n",
        ")\n",
        "models = client.models.list()"
      ],
      "id": "IS-E6uzlxFhI"
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "iqPajWTjxFhI",
        "outputId": "0b2c4ebb-25c3-4df2-caf3-f397b29db40f",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'openai', 'openai-internal', 'system'}"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ],
      "source": [
        "set([model.owned_by for model in models])"
      ],
      "id": "iqPajWTjxFhI"
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "LlX1o2bkxFhJ",
        "outputId": "f772123e-de1a-4c31-f3ae-dbdba73ea679",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'gpt-3.5-turbo',\n",
              " 'gpt-3.5-turbo-0301',\n",
              " 'gpt-3.5-turbo-0613',\n",
              " 'gpt-3.5-turbo-16k-0613',\n",
              " 'gpt-4',\n",
              " 'gpt-4-0613'}"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ],
      "source": [
        "set([model.id for model in models if model.owned_by == 'openai'])"
      ],
      "id": "LlX1o2bkxFhJ"
    },
    {
      "cell_type": "code",
      "source": [
        "set([model.id for model in models if model.owned_by == 'system'])"
      ],
      "metadata": {
        "id": "FACWSIE8mBEn",
        "outputId": "b39a334b-793f-429c-800b-5f1325605677",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "id": "FACWSIE8mBEn",
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'babbage-002',\n",
              " 'dall-e-2',\n",
              " 'dall-e-3',\n",
              " 'davinci-002',\n",
              " 'gpt-3.5-turbo-0125',\n",
              " 'gpt-3.5-turbo-1106',\n",
              " 'gpt-3.5-turbo-instruct',\n",
              " 'gpt-3.5-turbo-instruct-0914',\n",
              " 'gpt-4-0125-preview',\n",
              " 'gpt-4-1106-preview',\n",
              " 'gpt-4-1106-vision-preview',\n",
              " 'gpt-4-turbo',\n",
              " 'gpt-4-turbo-2024-04-09',\n",
              " 'gpt-4-turbo-preview',\n",
              " 'gpt-4-vision-preview',\n",
              " 'gpt-4o',\n",
              " 'gpt-4o-2024-05-13',\n",
              " 'text-embedding-3-large',\n",
              " 'text-embedding-3-small',\n",
              " 'tts-1-1106',\n",
              " 'tts-1-hd',\n",
              " 'tts-1-hd-1106'}"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "set([model.id for model in models if model.owned_by == 'openai-internal'])"
      ],
      "metadata": {
        "id": "uxhXGTvml2h1",
        "outputId": "b1701335-983c-414a-eeab-da861818f6ee",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "id": "uxhXGTvml2h1",
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'gpt-3.5-turbo-16k', 'text-embedding-ada-002', 'tts-1', 'whisper-1'}"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "jp397w5UxFhJ",
        "outputId": "58bded8d-8377-485b-9840-e592f776dc80",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 89
        }
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'Große Sprachmodelle sind mächtig aufgrund:\\n\\n1. **Skalierung**: Viele Parameter, die komplexe Muster erfassen.\\n2. **Datenvielfalt**: Training auf enormen, vielfältigen Datensätzen.\\n3. **Architektur**: Transformer-Architektur, die effektiv Kontext versteht.\\n4. **Feinabstimmung**: Anpassung an spezifische Aufgaben durch Fine-Tuning.\\n5. **Generalisierung**: Fähigkeit, Wissen zu transferieren und vielseitige Probleme zu lösen.'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 10
        }
      ],
      "source": [
        "# https://platform.openai.com/docs/api-reference/chat/create\n",
        "completion = client.chat.completions.create(\n",
        "  # model=\"gpt-3.5-turbo\",\n",
        "  model=\"gpt-4o\",\n",
        "\n",
        "  messages=[\n",
        "    {\"role\": \"system\", \"content\": \"Du bist ein Experte im Bereich Machine Learning. Deine Antworten sind kurz und gut.\"},\n",
        "    {\"role\": \"user\", \"content\": \"Warum sind große Sprachmodelle so mächtig?\"},\n",
        "  ],\n",
        "  logprobs=True,\n",
        "  top_logprobs=3\n",
        ")\n",
        "completion.choices[0].message.content"
      ],
      "id": "jp397w5UxFhJ"
    },
    {
      "cell_type": "code",
      "source": [
        "for token in completion.choices[0].logprobs.content:\n",
        "  print(\"---\")\n",
        "  print(token.token)\n",
        "  for top_token in token.top_logprobs:\n",
        "    print(f\"{top_token.token}: {top_token.logprob}\")"
      ],
      "metadata": {
        "id": "w-zEFMNV0ux2",
        "outputId": "5bd75c85-f5c2-4d85-8ec0-20303ea6a76d",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "id": "w-zEFMNV0ux2",
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "---\n",
            "Gro\n",
            "Gro: -0.000117370815\n",
            "Gr: -9.625117\n",
            "G: -11.125117\n",
            "---\n",
            "ße\n",
            "ße: -1.6524515e-05\n",
            "ß: -11.125016\n",
            "ßen: -13.375016\n",
            "---\n",
            " Sprach\n",
            " Sprach: -6.704273e-07\n",
            " Modelle: -14.750001\n",
            " sprach: -16.125\n",
            "---\n",
            "model\n",
            "model: -2.577686e-06\n",
            "module: -13.750003\n",
            "modell: -13.875003\n",
            "---\n",
            "le\n",
            "le: 0.0\n",
            "lek: -21.5\n",
            "lem: -22.125\n",
            "---\n",
            " sind\n",
            " sind: -0.024960347\n",
            " wie: -4.1499605\n",
            ",: -4.7749605\n",
            "---\n",
            " mä\n",
            " mä: -0.10221221\n",
            " so: -2.3522122\n",
            " deshalb: -7.2272124\n",
            "---\n",
            "chtig\n",
            "chtig: -5.5122365e-07\n",
            "cht: -14.750001\n",
            "chtigt: -16.625\n",
            "---\n",
            " aufgrund\n",
            ",: -0.25334388\n",
            " aufgrund: -2.1283438\n",
            " wegen: -2.3783438\n",
            "---\n",
            ":\n",
            "\n",
            "\n",
            " ihrer: -0.019424668\n",
            " mehr: -5.0194244\n",
            " von: -5.2694244\n",
            "---\n",
            "1\n",
            "1: -3.5716304e-05\n",
            "-: -10.250035\n",
            " : -16.125036\n",
            "---\n",
            ".\n",
            ".: -9.0883464e-07\n",
            "): -14.125001\n",
            "-: -16.75\n",
            "---\n",
            " **\n",
            " **: -0.0006212961\n",
            " ***: -8.500621\n",
            " Umfang: -8.500621\n",
            "---\n",
            "Sk\n",
            "D: -0.7418427\n",
            "Sk: -1.1168427\n",
            "Mass: -2.7418427\n",
            "---\n",
            "al\n",
            "al: -0.021226155\n",
            "alen: -4.146226\n",
            "ala: -5.271226\n",
            "---\n",
            "ierung\n",
            "ierung: -0.07180578\n",
            "ier: -2.8218057\n",
            "ene: -5.196806\n",
            "---\n",
            "**\n",
            "**: -0.15021011\n",
            ":**: -2.6502101\n",
            " der: -3.1502101\n",
            "---\n",
            ":\n",
            ":: -0.00011856285\n",
            ":\n",
            ": -9.375118\n",
            " -: -10.500118\n",
            "---\n",
            " Viele\n",
            " Sie: -0.94804233\n",
            " Mehr: -1.3230424\n",
            " Grö: -2.0730424\n",
            "---\n",
            " Parameter\n",
            " Parameter: -0.007063916\n",
            " Daten: -5.757064\n",
            " Param: -6.132064\n",
            "---\n",
            ",\n",
            " ermöglichen: -0.40318185\n",
            " erlaub: -1.7781818\n",
            " erf: -2.6531818\n",
            "---\n",
            " die\n",
            " die: -0.029338589\n",
            " um: -5.4043384\n",
            " was: -5.5293384\n",
            "---\n",
            " komplex\n",
            " komplex: -0.262723\n",
            " fe: -2.387723\n",
            " mehr: -4.262723\n",
            "---\n",
            "e\n",
            "e: -0.0043644905\n",
            "ere: -5.8793645\n",
            "es: -6.5043645\n",
            "---\n",
            " Muster\n",
            " Muster: -0.025570236\n",
            " Zusammen: -4.5255704\n",
            " Sprach: -4.9005704\n",
            "---\n",
            " erf\n",
            " erf: -1.0019282\n",
            " und: -1.0019282\n",
            " lernen: -2.2519283\n",
            "---\n",
            "assen\n",
            "assen: -1.3856493e-06\n",
            "ass: -13.750001\n",
            "ors: -15.125001\n",
            "---\n",
            ".\n",
            "\n",
            ".\n",
            ": -0.16222924\n",
            " können: -1.9122293\n",
            " und: -6.349729\n",
            "---\n",
            "2\n",
            "2: -0.00017624642\n",
            "1: -9.000176\n",
            "   \n",
            ": -10.125176\n",
            "---\n",
            ".\n",
            ".: 0.0\n",
            ":: -17.375\n",
            " **: -17.375\n",
            "---\n",
            " **\n",
            " **: -1.9361265e-07\n",
            "**: -16.125\n",
            " ***: -16.375\n",
            "---\n",
            "D\n",
            "D: -0.353438\n",
            "Training: -2.228438\n",
            "Vor: -2.353438\n",
            "---\n",
            "aten\n",
            "aten: -0.02398318\n",
            "iverse: -3.7739832\n",
            "iversity: -7.398983\n",
            "---\n",
            "vi\n",
            "vi: -0.5391582\n",
            "men: -1.1641582\n",
            "**: -2.5391583\n",
            "---\n",
            "elfalt\n",
            "elfalt: -6.9882217e-06\n",
            "elf: -11.875007\n",
            "abil: -19.875008\n",
            "---\n",
            "**\n",
            "**: -6.8854926e-05\n",
            ":**: -9.875069\n",
            " und: -11.000069\n",
            "---\n",
            ":\n",
            ":: -1.9361265e-07\n",
            " :: -17.0\n",
            "-: -17.375\n",
            "---\n",
            " Training\n",
            " Training: -0.06305617\n",
            " Train: -3.1880562\n",
            " Trainings: -4.313056\n",
            "---\n",
            " auf\n",
            " auf: -0.032885592\n",
            " mit: -3.5328856\n",
            " an: -5.7828856\n",
            "---\n",
            " enorm\n",
            " ries: -0.35161617\n",
            " umfang: -1.6016161\n",
            " großen: -2.9766161\n",
            "---\n",
            "en\n",
            "en: -0.024253389\n",
            " großen: -4.6492534\n",
            " vielfält: -4.7742534\n",
            "---\n",
            ",\n",
            ",: -0.7120002\n",
            " Daten: -1.7120001\n",
            " und: -1.8370001\n",
            "---\n",
            " vielfält\n",
            " vielfält: -0.11285905\n",
            " diversen: -2.862859\n",
            " vielseit: -3.737859\n",
            "---\n",
            "igen\n",
            "igen: -4.561021e-05\n",
            "ig: -10.625046\n",
            "igem: -11.125046\n",
            "---\n",
            " Dat\n",
            " Dat: -0.4810584\n",
            " Text: -1.1060584\n",
            " Daten: -2.9810584\n",
            "---\n",
            "ens\n",
            "ens: -7.564953e-05\n",
            "ensatz: -9.625075\n",
            "ense: -12.625075\n",
            "---\n",
            "ätzen\n",
            "ätzen: -0.0026081179\n",
            "ets: -6.0026083\n",
            "ätze: -9.002608\n",
            "---\n",
            ".\n",
            "\n",
            ".\n",
            ": -0.017228698\n",
            ",: -5.3922286\n",
            " für: -5.8922286\n",
            "---\n",
            "3\n",
            "3: -5.5122365e-07\n",
            "4: -15.375001\n",
            "2: -15.625001\n",
            "---\n",
            ".\n",
            ".: 0.0\n",
            " **: -18.1875\n",
            ":: -18.875\n",
            "---\n",
            " **\n",
            " **: -5.5122365e-07\n",
            "**: -14.875001\n",
            " ***: -15.625001\n",
            "---\n",
            "Arch\n",
            "Transfer: -0.6177183\n",
            "Arch: -1.1177182\n",
            "Transform: -3.7427182\n",
            "---\n",
            "itektur\n",
            "itektur: -0.014795026\n",
            "itekt: -4.264795\n",
            "itek: -7.389795\n",
            "---\n",
            "**\n",
            "**: -0.0012182132\n",
            " (: -7.1262183\n",
            "fort: -8.751218\n",
            "---\n",
            ":\n",
            ":: -2.6968896e-06\n",
            ":\n",
            ": -13.500003\n",
            " (: -14.125003\n",
            "---\n",
            " Transformer\n",
            " Transformer: -0.2412544\n",
            " Fortsch: -2.7412543\n",
            " Moderne: -3.6162543\n",
            "---\n",
            "-\n",
            "-: -0.028141242\n",
            "-bas: -4.278141\n",
            "-Str: -4.903141\n",
            "---\n",
            "Arch\n",
            "Arch: -0.067617334\n",
            "Model: -2.8176174\n",
            "Design: -5.4426174\n",
            "---\n",
            "itektur\n",
            "itektur: -0.055218045\n",
            "itekt: -2.930218\n",
            "itek: -8.180218\n",
            "---\n",
            ",\n",
            " ermöglicht: -0.4987161\n",
            ",: -1.9987161\n",
            " für: -2.498716\n",
            "---\n",
            " die\n",
            " die: -0.035070628\n",
            " effizient: -4.6600704\n",
            " effektiv: -4.9100704\n",
            "---\n",
            " effektiv\n",
            " Kontext: -1.1976988\n",
            " effizient: -1.4476988\n",
            " effiz: -1.8226988\n",
            "---\n",
            " Kontext\n",
            " Kontext: -0.1318316\n",
            " kon: -3.3818316\n",
            " lange: -4.1318316\n",
            "---\n",
            " versteht\n",
            " versteht: -0.6450137\n",
            " und: -1.7700137\n",
            " er: -1.7700137\n",
            "---\n",
            ".\n",
            "\n",
            ".\n",
            ": -0.017767191\n",
            " und: -4.142767\n",
            ".\n",
            "\n",
            ": -6.517767\n",
            "---\n",
            "4\n",
            "4: -7.465036e-06\n",
            "3: -12.250008\n",
            "5: -13.000008\n",
            "---\n",
            ".\n",
            ".: -3.1281633e-07\n",
            ":: -15.5\n",
            " **: -17.125\n",
            "---\n",
            " **\n",
            " **: -2.1008714e-06\n",
            "**: -13.250002\n",
            " ***: -15.250002\n",
            "---\n",
            "Fe\n",
            "Transfer: -0.3552681\n",
            "Fe: -2.105268\n",
            "Vor: -2.605268\n",
            "---\n",
            "in\n",
            "in: -0.20363887\n",
            "int: -1.7036389\n",
            "inet: -7.078639\n",
            "---\n",
            "ab\n",
            "ab: -0.049077805\n",
            "just: -3.1740777\n",
            "-T: -5.174078\n",
            "---\n",
            "st\n",
            "st: -1.0325924e-05\n",
            "stim: -12.2500105\n",
            "gleich: -12.3750105\n",
            "---\n",
            "immung\n",
            "immung: -0.0005008472\n",
            "immungen: -7.7505007\n",
            "imm: -9.750501\n",
            "---\n",
            "**\n",
            "**: -0.026496759\n",
            " (: -3.6514966\n",
            ":**: -9.526497\n",
            "---\n",
            ":\n",
            ":: -7.77952e-05\n",
            " (: -9.500078\n",
            " (\": -14.125078\n",
            "---\n",
            " Anpass\n",
            " Anpass: -0.07537365\n",
            " Möglichkeit: -3.0753736\n",
            " Fähigkeit: -4.8253736\n",
            "---\n",
            "ung\n",
            "ung: -0.16565222\n",
            "bar: -2.5406523\n",
            "barkeit: -3.4156523\n",
            "---\n",
            " an\n",
            " an: -0.15506725\n",
            " auf: -2.1550672\n",
            " für: -3.7800672\n",
            "---\n",
            " spezif\n",
            " spezif: -0.03968168\n",
            " spez: -3.7896817\n",
            " spezielle: -4.164682\n",
            "---\n",
            "ische\n",
            "ische: -3.4285862e-05\n",
            "ischen: -10.375034\n",
            "isch: -13.000034\n",
            "---\n",
            " Aufgaben\n",
            " Aufgaben: -0.0020234508\n",
            " Anwendungen: -6.877023\n",
            " An: -7.502023\n",
            "---\n",
            " durch\n",
            " durch: -0.50414604\n",
            ".\n",
            "\n",
            ": -2.254146\n",
            " und: -2.629146\n",
            "---\n",
            " Fine\n",
            " Fine: -0.53365684\n",
            " Transfer: -1.5336568\n",
            " Fe: -2.2836568\n",
            "---\n",
            "-T\n",
            "-T: -0.00761551\n",
            "-t: -4.8826156\n",
            " T: -11.632615\n",
            "---\n",
            "uning\n",
            "uning: -1.2664457e-06\n",
            "unning: -14.000001\n",
            "un: -15.625001\n",
            "---\n",
            ".\n",
            "\n",
            ".\n",
            "\n",
            ": -0.5644545\n",
            ".\n",
            ": -1.1894546\n",
            ".: -2.0644546\n",
            "---\n",
            "5\n",
            "5: -0.03832513\n",
            "<|end|>: -4.0383253\n",
            "   \n",
            ": -4.2883253\n",
            "---\n",
            ".\n",
            ".: -1.147242e-06\n",
            "**: -14.750001\n",
            ":: -14.875001\n",
            "---\n",
            " **\n",
            " **: -3.5313153e-06\n",
            "**: -13.000004\n",
            " ***: -14.250004\n",
            "---\n",
            "Gener\n",
            "Transfer: -0.63983476\n",
            "Gener: -1.7648348\n",
            "General: -2.014835\n",
            "---\n",
            "alis\n",
            "alis: -0.00049131457\n",
            "elle: -8.500491\n",
            "ative: -8.875491\n",
            "---\n",
            "ierung\n",
            "ierung: -0.16109268\n",
            "ierungs: -1.9110926\n",
            "ier: -7.1610928\n",
            "---\n",
            "**\n",
            "**: -2.7968066e-05\n",
            ":**: -11.000028\n",
            "skraft: -12.375028\n",
            "---\n",
            ":\n",
            ":: -3.2929079e-06\n",
            " :: -14.625003\n",
            " Fähigkeit: -15.250003\n",
            "---\n",
            " Fähigkeit\n",
            " Fähigkeit: -0.10773718\n",
            " Gute: -3.357737\n",
            " Über: -3.857737\n",
            "---\n",
            ",\n",
            ",: -0.0128491195\n",
            " zur: -4.3878493\n",
            " zu: -8.637849\n",
            "---\n",
            " Wissen\n",
            " Wissen: -0.23232016\n",
            " auf: -2.60732\n",
            " verschiedene: -3.35732\n",
            "---\n",
            " zu\n",
            " auf: -0.6875894\n",
            " aus: -1.4375894\n",
            " über: -1.5625894\n",
            "---\n",
            " transfer\n",
            " transfer: -0.55429673\n",
            " übertragen: -1.6792967\n",
            " ver: -2.3042967\n",
            "---\n",
            "ieren\n",
            "ieren: -3.0545007e-06\n",
            "eren: -14.125003\n",
            "ien: -14.375003\n",
            "---\n",
            " und\n",
            " und: -0.043827634\n",
            ".\n",
            "\n",
            ": -4.0438275\n",
            " über: -4.5438275\n",
            "---\n",
            " vielseit\n",
            " neue: -1.5565032\n",
            " auf: -1.6815032\n",
            " vielseit: -1.9315032\n",
            "---\n",
            "ige\n",
            "ige: -0.6328285\n",
            "ig: -0.7578285\n",
            "igen: -8.757829\n",
            "---\n",
            " Probleme\n",
            " Aufgaben: -0.010483979\n",
            " Probleme: -5.5104837\n",
            " Anwendungen: -5.8854837\n",
            "---\n",
            " zu\n",
            " zu: -0.0015277173\n",
            " an: -7.126528\n",
            " lösen: -7.251528\n",
            "---\n",
            " lösen\n",
            " lösen: -0.00053230225\n",
            " bearbeiten: -8.000532\n",
            " adress: -9.625532\n",
            "---\n",
            ".\n",
            ".: -0.38846865\n",
            ".\n",
            "\n",
            ": -1.1384686\n",
            ".\n",
            ": -6.5134687\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "_Sx00u1ulhw1"
      },
      "id": "_Sx00u1ulhw1",
      "execution_count": 11,
      "outputs": []
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.9.18"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}