{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "\n", " \"Open\n", "\n", "\n", "## Welcome to Okareo!\n", "\n", "Get your API token from [https://app.okareo.com/](https://app.okareo.com/) and set it in the cell below. 👇\n", " (Note: You will also need an OpenAI key.)\n", "\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "OKAREO_API_KEY = \"\"\n", "OPENAI_API_KEY = \"\"" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "%pip install okareo\n", "%pip install openai" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "# Simple summarization prompt using OpenAI'a GPT 3.5 Turbo model\n", "from openai import OpenAI\n", "\n", "client = OpenAI(api_key=OPENAI_API_KEY)\n", "\n", "def get_turbo_summary(messages, model=\"gpt-3.5-turbo\", \n", " temperature=0, max_tokens=500):\n", " response = client.chat.completions.create(\n", " model=model,\n", " messages=messages,\n", " temperature=temperature, \n", " max_tokens=max_tokens,\n", " )\n", " return response.choices[0].message.content\n", "\n", "USER_PROMPT_TEMPLATE = \"{input}\"\n", "\n", "SUMMARIZATION_CONTEXT_TEMPLATE = \"\"\"\n", "You will be provided with text.\n", "Summarize the text in 1 simple sentence.\n", "\"\"\"\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "# Evaluate the scenario and model combination and then get a link to the results on Okareo\n", "import random\n", "import string\n", "import os\n", "import tempfile\n", "\n", "from okareo import Okareo\n", "from okareo.model_under_test import OpenAIModel\n", "from okareo_api_client.models.test_run_type import TestRunType\n", "\n", "okareo = Okareo(OKAREO_API_KEY)\n", "random_string = ''.join(random.choices(string.ascii_letters, k=5))\n", "mut_name = f\"OpenAI Summarization Model - {random_string}\"\n", "eval_name = f\"Summarization Run - {random_string}\"\n", "\n", "# Register the model to use in the test run\n", "model_under_test = okareo.register_model(\n", " name=mut_name,\n", " model=OpenAIModel(\n", " model_id=\"gpt-3.5-turbo\",\n", " temperature=0,\n", " system_prompt_template=SUMMARIZATION_CONTEXT_TEMPLATE,\n", " user_prompt_template=USER_PROMPT_TEMPLATE,\n", " ),\n", ")\n", "\n", "# Create a scenario to evaluate the model with\n", "# Get jsonl file from Okareo's SDK repo\n", "webbizz_articles = os.popen('curl https://raw.githubusercontent.com/okareo-ai/okareo-python-sdk/main/examples/webbizz_10_articles.jsonl').read()\n", "temp_dir = tempfile.gettempdir()\n", "file_path = os.path.join(temp_dir, \"webbizz_10_articles.jsonl\")\n", "with open(file_path, \"w+\") as file:\n", " lines = webbizz_articles.split('\\n')\n", " # Use the first 3 json objects to make a scenario set with 3 scenarios\n", " for i in range(3):\n", " file.write(f\"{lines[i]}\\n\")\n", "scenario = okareo.upload_scenario_set(file_path=file_path, scenario_name=\"Webbizz Articles Scenario\")\n", "\n", "# make sure to clean up tmp file\n", "os.remove(file_path)\n", "\n", "# Run the evaluation\n", "evaluation = model_under_test.run_test(\n", " name=eval_name,\n", " scenario=scenario,\n", " api_key=OPENAI_API_KEY,\n", " test_run_type=TestRunType.NL_GENERATION,\n", " calculate_metrics=True,\n", ")\n", "\n", "print(f\"See results in Okareo: {evaluation.app_link}\")" ] } ], "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": "3.11.2" } }, "nbformat": 4, "nbformat_minor": 2 }