{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "## Ragas demonstration\n", "\n", "This notebook demonstrates how to evaluate a RAG system using the [Ragas evaluation framework](https://github.com/explodinggradients/ragas?tab=readme-ov-file) and import the resulting evaluation results into InspectorRAGet for analysis.\n", "\n", "### Installation" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Collecting git+https://github.com/explodinggradients/ragas\n", " Cloning https://github.com/explodinggradients/ragas to /private/var/folders/l5/fj0t2qmn44x0r042xw6cjr8w0000gn/T/pip-req-build-k70llt1h\n", " Running command git clone --filter=blob:none --quiet https://github.com/explodinggradients/ragas /private/var/folders/l5/fj0t2qmn44x0r042xw6cjr8w0000gn/T/pip-req-build-k70llt1h\n", " Resolved https://github.com/explodinggradients/ragas to commit d2486f117fd827dcfc3e196d4cf7798573c55b09\n", " Installing build dependencies ... \u001b[?25ldone\n", "\u001b[?25h Getting requirements to build wheel ... \u001b[?25ldone\n", "\u001b[?25h Preparing metadata (pyproject.toml) ... \u001b[?25ldone\n", "\u001b[?25hRequirement already satisfied: numpy in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (1.26.4)\n", "Requirement already satisfied: datasets in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (2.20.0)\n", "Requirement already satisfied: tiktoken in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (0.7.0)\n", "Requirement already satisfied: langchain in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (0.2.12)\n", "Requirement already satisfied: langchain-core in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (0.2.28)\n", "Requirement already satisfied: langchain-community in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (0.2.11)\n", "Requirement already satisfied: langchain-openai in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (0.1.20)\n", "Requirement already satisfied: openai>1 in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (1.39.0)\n", "Requirement already satisfied: pysbd>=0.3.4 in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (0.3.4)\n", "Requirement already satisfied: nest-asyncio in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (1.6.0)\n", "Requirement already satisfied: appdirs in ./.venv/lib/python3.9/site-packages (from ragas==0.1.13) (1.4.4)\n", "Requirement already satisfied: anyio<5,>=3.5.0 in ./.venv/lib/python3.9/site-packages (from openai>1->ragas==0.1.13) (4.4.0)\n", "Requirement already satisfied: distro<2,>=1.7.0 in ./.venv/lib/python3.9/site-packages (from openai>1->ragas==0.1.13) (1.9.0)\n", "Requirement already satisfied: httpx<1,>=0.23.0 in ./.venv/lib/python3.9/site-packages (from openai>1->ragas==0.1.13) (0.27.0)\n", "Requirement already satisfied: pydantic<3,>=1.9.0 in ./.venv/lib/python3.9/site-packages (from openai>1->ragas==0.1.13) (2.8.2)\n", "Requirement already satisfied: sniffio in ./.venv/lib/python3.9/site-packages (from openai>1->ragas==0.1.13) (1.3.1)\n", "Requirement already satisfied: tqdm>4 in ./.venv/lib/python3.9/site-packages (from openai>1->ragas==0.1.13) (4.66.5)\n", "Requirement already satisfied: typing-extensions<5,>=4.7 in ./.venv/lib/python3.9/site-packages (from openai>1->ragas==0.1.13) (4.12.2)\n", "Requirement already satisfied: filelock in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (3.15.4)\n", "Requirement already satisfied: pyarrow>=15.0.0 in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (17.0.0)\n", "Requirement already satisfied: pyarrow-hotfix in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (0.6)\n", "Requirement already satisfied: dill<0.3.9,>=0.3.0 in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (0.3.8)\n", "Requirement already satisfied: pandas in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (2.2.2)\n", "Requirement already satisfied: requests>=2.32.2 in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (2.32.3)\n", "Requirement already satisfied: xxhash in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (3.4.1)\n", "Requirement already satisfied: multiprocess in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (0.70.16)\n", "Requirement already satisfied: fsspec<=2024.5.0,>=2023.1.0 in ./.venv/lib/python3.9/site-packages (from fsspec[http]<=2024.5.0,>=2023.1.0->datasets->ragas==0.1.13) (2024.5.0)\n", "Requirement already satisfied: aiohttp in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (3.10.1)\n", "Requirement already satisfied: huggingface-hub>=0.21.2 in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (0.24.5)\n", "Requirement already satisfied: packaging in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (24.1)\n", "Requirement already satisfied: pyyaml>=5.1 in ./.venv/lib/python3.9/site-packages (from datasets->ragas==0.1.13) (6.0.1)\n", "Requirement already satisfied: SQLAlchemy<3,>=1.4 in ./.venv/lib/python3.9/site-packages (from langchain->ragas==0.1.13) (2.0.32)\n", "Requirement already satisfied: async-timeout<5.0.0,>=4.0.0 in ./.venv/lib/python3.9/site-packages (from langchain->ragas==0.1.13) (4.0.3)\n", "Requirement already satisfied: langchain-text-splitters<0.3.0,>=0.2.0 in ./.venv/lib/python3.9/site-packages (from langchain->ragas==0.1.13) (0.2.2)\n", "Requirement already satisfied: langsmith<0.2.0,>=0.1.17 in ./.venv/lib/python3.9/site-packages (from langchain->ragas==0.1.13) (0.1.96)\n", "Requirement already satisfied: tenacity!=8.4.0,<9.0.0,>=8.1.0 in ./.venv/lib/python3.9/site-packages (from langchain->ragas==0.1.13) (8.5.0)\n", "Requirement already satisfied: jsonpatch<2.0,>=1.33 in ./.venv/lib/python3.9/site-packages (from langchain-core->ragas==0.1.13) (1.33)\n", "Requirement already satisfied: dataclasses-json<0.7,>=0.5.7 in ./.venv/lib/python3.9/site-packages (from langchain-community->ragas==0.1.13) (0.6.7)\n", "Requirement already satisfied: regex>=2022.1.18 in ./.venv/lib/python3.9/site-packages (from tiktoken->ragas==0.1.13) (2024.7.24)\n", "Requirement already satisfied: aiohappyeyeballs>=2.3.0 in ./.venv/lib/python3.9/site-packages (from aiohttp->datasets->ragas==0.1.13) (2.3.4)\n", "Requirement already satisfied: aiosignal>=1.1.2 in ./.venv/lib/python3.9/site-packages (from aiohttp->datasets->ragas==0.1.13) (1.3.1)\n", "Requirement already satisfied: attrs>=17.3.0 in ./.venv/lib/python3.9/site-packages (from aiohttp->datasets->ragas==0.1.13) (24.1.0)\n", "Requirement already satisfied: frozenlist>=1.1.1 in ./.venv/lib/python3.9/site-packages (from aiohttp->datasets->ragas==0.1.13) (1.4.1)\n", "Requirement already satisfied: multidict<7.0,>=4.5 in ./.venv/lib/python3.9/site-packages (from aiohttp->datasets->ragas==0.1.13) (6.0.5)\n", "Requirement already satisfied: yarl<2.0,>=1.0 in ./.venv/lib/python3.9/site-packages (from aiohttp->datasets->ragas==0.1.13) (1.9.4)\n", "Requirement already satisfied: idna>=2.8 in ./.venv/lib/python3.9/site-packages (from anyio<5,>=3.5.0->openai>1->ragas==0.1.13) (3.7)\n", "Requirement already satisfied: exceptiongroup>=1.0.2 in ./.venv/lib/python3.9/site-packages (from anyio<5,>=3.5.0->openai>1->ragas==0.1.13) (1.2.2)\n", "Requirement already satisfied: marshmallow<4.0.0,>=3.18.0 in ./.venv/lib/python3.9/site-packages (from dataclasses-json<0.7,>=0.5.7->langchain-community->ragas==0.1.13) (3.21.3)\n", "Requirement already satisfied: typing-inspect<1,>=0.4.0 in ./.venv/lib/python3.9/site-packages (from dataclasses-json<0.7,>=0.5.7->langchain-community->ragas==0.1.13) (0.9.0)\n", "Requirement already satisfied: certifi in ./.venv/lib/python3.9/site-packages (from httpx<1,>=0.23.0->openai>1->ragas==0.1.13) (2024.7.4)\n", "Requirement already satisfied: httpcore==1.* in ./.venv/lib/python3.9/site-packages (from httpx<1,>=0.23.0->openai>1->ragas==0.1.13) (1.0.5)\n", "Requirement already satisfied: h11<0.15,>=0.13 in ./.venv/lib/python3.9/site-packages (from httpcore==1.*->httpx<1,>=0.23.0->openai>1->ragas==0.1.13) (0.14.0)\n", "Requirement already satisfied: jsonpointer>=1.9 in ./.venv/lib/python3.9/site-packages (from jsonpatch<2.0,>=1.33->langchain-core->ragas==0.1.13) (3.0.0)\n", "Requirement already satisfied: orjson<4.0.0,>=3.9.14 in ./.venv/lib/python3.9/site-packages (from langsmith<0.2.0,>=0.1.17->langchain->ragas==0.1.13) (3.10.6)\n", "Requirement already satisfied: annotated-types>=0.4.0 in ./.venv/lib/python3.9/site-packages (from pydantic<3,>=1.9.0->openai>1->ragas==0.1.13) (0.7.0)\n", "Requirement already satisfied: pydantic-core==2.20.1 in ./.venv/lib/python3.9/site-packages (from pydantic<3,>=1.9.0->openai>1->ragas==0.1.13) (2.20.1)\n", "Requirement already satisfied: charset-normalizer<4,>=2 in ./.venv/lib/python3.9/site-packages (from requests>=2.32.2->datasets->ragas==0.1.13) (3.3.2)\n", "Requirement already satisfied: urllib3<3,>=1.21.1 in ./.venv/lib/python3.9/site-packages (from requests>=2.32.2->datasets->ragas==0.1.13) (2.2.2)\n", "Requirement already satisfied: greenlet!=0.4.17 in ./.venv/lib/python3.9/site-packages (from SQLAlchemy<3,>=1.4->langchain->ragas==0.1.13) (3.0.3)\n", "Requirement already satisfied: python-dateutil>=2.8.2 in ./.venv/lib/python3.9/site-packages (from pandas->datasets->ragas==0.1.13) (2.9.0.post0)\n", "Requirement already satisfied: pytz>=2020.1 in ./.venv/lib/python3.9/site-packages (from pandas->datasets->ragas==0.1.13) (2024.1)\n", "Requirement already satisfied: tzdata>=2022.7 in ./.venv/lib/python3.9/site-packages (from pandas->datasets->ragas==0.1.13) (2024.1)\n", "Requirement already satisfied: six>=1.5 in ./.venv/lib/python3.9/site-packages (from python-dateutil>=2.8.2->pandas->datasets->ragas==0.1.13) (1.16.0)\n", "Requirement already satisfied: mypy-extensions>=0.3.0 in ./.venv/lib/python3.9/site-packages (from typing-inspect<1,>=0.4.0->dataclasses-json<0.7,>=0.5.7->langchain-community->ragas==0.1.13) (1.0.0)\n" ] } ], "source": [ "!pip install git+https://github.com/explodinggradients/ragas" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Requirement already satisfied: ipywidgets in ./.venv/lib/python3.9/site-packages (8.1.3)\n", "Requirement already satisfied: comm>=0.1.3 in ./.venv/lib/python3.9/site-packages (from ipywidgets) (0.2.2)\n", "Requirement already satisfied: ipython>=6.1.0 in ./.venv/lib/python3.9/site-packages (from ipywidgets) (8.18.1)\n", "Requirement already satisfied: traitlets>=4.3.1 in ./.venv/lib/python3.9/site-packages (from ipywidgets) (5.14.3)\n", "Requirement already satisfied: widgetsnbextension~=4.0.11 in ./.venv/lib/python3.9/site-packages (from ipywidgets) (4.0.11)\n", "Requirement already satisfied: jupyterlab-widgets~=3.0.11 in ./.venv/lib/python3.9/site-packages (from ipywidgets) (3.0.11)\n", "Requirement already satisfied: decorator in ./.venv/lib/python3.9/site-packages (from ipython>=6.1.0->ipywidgets) (5.1.1)\n", "Requirement already satisfied: jedi>=0.16 in ./.venv/lib/python3.9/site-packages (from ipython>=6.1.0->ipywidgets) (0.19.1)\n", "Requirement already satisfied: matplotlib-inline in ./.venv/lib/python3.9/site-packages (from ipython>=6.1.0->ipywidgets) (0.1.7)\n", "Requirement already satisfied: prompt-toolkit<3.1.0,>=3.0.41 in ./.venv/lib/python3.9/site-packages (from ipython>=6.1.0->ipywidgets) (3.0.47)\n", "Requirement already satisfied: pygments>=2.4.0 in ./.venv/lib/python3.9/site-packages (from ipython>=6.1.0->ipywidgets) (2.18.0)\n", "Requirement already satisfied: stack-data in ./.venv/lib/python3.9/site-packages (from ipython>=6.1.0->ipywidgets) (0.6.3)\n", "Requirement already satisfied: typing-extensions in ./.venv/lib/python3.9/site-packages (from ipython>=6.1.0->ipywidgets) (4.12.2)\n", "Requirement already satisfied: exceptiongroup in ./.venv/lib/python3.9/site-packages (from ipython>=6.1.0->ipywidgets) (1.2.2)\n", "Requirement already satisfied: pexpect>4.3 in ./.venv/lib/python3.9/site-packages (from ipython>=6.1.0->ipywidgets) (4.9.0)\n", "Requirement already satisfied: parso<0.9.0,>=0.8.3 in ./.venv/lib/python3.9/site-packages (from jedi>=0.16->ipython>=6.1.0->ipywidgets) (0.8.4)\n", "Requirement already satisfied: ptyprocess>=0.5 in ./.venv/lib/python3.9/site-packages (from pexpect>4.3->ipython>=6.1.0->ipywidgets) (0.7.0)\n", "Requirement already satisfied: wcwidth in ./.venv/lib/python3.9/site-packages (from prompt-toolkit<3.1.0,>=3.0.41->ipython>=6.1.0->ipywidgets) (0.2.13)\n", "Requirement already satisfied: executing>=1.2.0 in ./.venv/lib/python3.9/site-packages (from stack-data->ipython>=6.1.0->ipywidgets) (2.0.1)\n", "Requirement already satisfied: asttokens>=2.1.0 in ./.venv/lib/python3.9/site-packages (from stack-data->ipython>=6.1.0->ipywidgets) (2.4.1)\n", "Requirement already satisfied: pure-eval in ./.venv/lib/python3.9/site-packages (from stack-data->ipython>=6.1.0->ipywidgets) (0.2.3)\n", "Requirement already satisfied: six>=1.12.0 in ./.venv/lib/python3.9/site-packages (from asttokens>=2.1.0->stack-data->ipython>=6.1.0->ipywidgets) (1.16.0)\n" ] } ], "source": [ "!pip install ipywidgets" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "from datasets import Dataset \n", "import os\n", "from ragas import evaluate\n", "from ragas.metrics import faithfulness, answer_relevancy, context_precision, context_recall" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "os.environ[\"RAGAS_DO_NOT_TRACK\"] = \"true\"" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Import dataset\n", "\n", "For this example, we will be using the sample dataset provided in Ragas' quick start instructions." ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [], "source": [ "data_samples = {\n", " 'question': ['When was the first super bowl?', 'Who won the most super bowls?'],\n", " 'answer': ['The first superbowl was held on Jan 15, 1967', 'The most super bowls have been won by The New England Patriots'],\n", " 'contexts' : [['The First AFL–NFL World Championship Game was an American football game played on January 15, 1967, at the Los Angeles Memorial Coliseum in Los Angeles,'], \n", " ['The Green Bay Packers...Green Bay, Wisconsin.','The Packers compete...Football Conference']],\n", " 'ground_truth': ['The first superbowl was held on January 15, 1967', 'The New England Patriots have won the Super Bowl a record six times']\n", "}\n", "\n", "dataset = Dataset.from_dict(data_samples)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Run evaluation\n", "\n", "In this example, we run evaluation using `gpt-4o-mini` and the following Ragas evaluation metrics: faithfulness, answer relevance, context precision, and context recall. Consult the Ragas documentation on how to use different models or metrics.\n", "\n", "**Note: To use one of the OpenAI model for evaluation, you have to fill in your `OPENAI_API_KEY` below.** " ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [], "source": [ "os.environ[\"OPENAI_API_KEY\"] = \"provide-your-openai-api-key\"" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "d221210e378a4ad0a25748d8f8f095ba", "version_major": 2, "version_minor": 0 }, "text/plain": [ "Evaluating: 0%| | 0/8 [00:00\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
questionanswercontextsground_truthfaithfulnessanswer_relevancycontext_precisioncontext_recall
0When was the first super bowl?The first superbowl was held on Jan 15, 1967[The First AFL–NFL World Championship Game was...The first superbowl was held on January 15, 19671.00.9807141.01.0
1Who won the most super bowls?The most super bowls have been won by The New ...[The Green Bay Packers...Green Bay, Wisconsin....The New England Patriots have won the Super Bo...0.00.9430430.00.0
\n", "" ], "text/plain": [ " question \\\n", "0 When was the first super bowl? \n", "1 Who won the most super bowls? \n", "\n", " answer \\\n", "0 The first superbowl was held on Jan 15, 1967 \n", "1 The most super bowls have been won by The New ... \n", "\n", " contexts \\\n", "0 [The First AFL–NFL World Championship Game was... \n", "1 [The Green Bay Packers...Green Bay, Wisconsin.... \n", "\n", " ground_truth faithfulness \\\n", "0 The first superbowl was held on January 15, 1967 1.0 \n", "1 The New England Patriots have won the Super Bo... 0.0 \n", "\n", " answer_relevancy context_precision context_recall \n", "0 0.980714 1.0 1.0 \n", "1 0.943043 0.0 0.0 " ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from langchain_openai.chat_models import ChatOpenAI\n", "llm = ChatOpenAI(model=\"gpt-4o-mini\")\n", "\n", "score = evaluate(dataset,llm=llm, metrics=[faithfulness,answer_relevancy,context_precision,context_recall])\n", "df_score = score.to_pandas()\n", "\n", "display(df_score)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Create InspectorRAGet file" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We next generate the file for InspectorRAGet. We start by specifying the experiment metadata (experiment name, model, metrics)." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**Specify name of experiment**" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "name = \"Ragas Demo\"" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**Specify name of model that produced the answers:** Since we do not know where the answers in the `data_samples` came from, we use a dummy name. In your experiments, replace this with the name of the model that produced the answers." ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [], "source": [ "# models -> List[dict]\n", "models = [\n", " {\n", " \"model_id\": \"model_a\", # e.g., \"OpenAI/gpt-3.5-turbo\"\n", " \"name\": \"Model A\", # e.g., \"GPT-3.5-Turbo\"\n", " \"owner\": \"Owner A\" # e.g., \"OpenAI\"\n", " }\n", "]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**Specify metrics used:** Ceate a record for each metric used in the Ragas evaluation. Note that the `name` of each metric below should match the name of the metric used in the data frame `df_score` output by Ragas." ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [], "source": [ "# metrics -> List[dict]\n", "all_metrics = [\n", " {\n", " \"name\": \"faithfulness\",\n", " \"display_name\": \"Faithfulness\",\n", " \"description\": \"Faithfulness\",\n", " \"author\": \"algorithm\",\n", " \"type\": \"numerical\",\n", " \"aggregator\": \"average\",\n", " \"range\": [0, 1.0, 0.1]\n", " },\n", " {\n", " \"name\": \"answer_relevancy\",\n", " \"display_name\": \"Answer Relevancy\",\n", " \"description\": \"Answer Relevancy\",\n", " \"author\": \"algorithm\",\n", " \"type\": \"numerical\",\n", " \"aggregator\": \"average\",\n", " \"range\": [0, 1.0, 0.1]\n", " },\n", " {\n", " \"name\": \"context_precision\",\n", " \"display_name\": \"Context Precision\",\n", " \"description\": \"Context Precision\",\n", " \"author\": \"algorithm\",\n", " \"type\": \"numerical\",\n", " \"aggregator\": \"average\",\n", " \"range\": [0, 1.0, 0.1]\n", " },\n", " {\n", " \"name\": \"context_recall\",\n", " \"display_name\": \"Context Recall\",\n", " \"description\": \"Context Recall\",\n", " \"author\": \"algorithm\",\n", " \"type\": \"numerical\",\n", " \"aggregator\": \"average\",\n", " \"range\": [0, 1.0, 0.1]\n", " }\n", "]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "##### Compute document IDs" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [], "source": [ "doc_id_counter = 0\n", "\n", "doc_text_to_id = {}\n", "for index, row in df_score.iterrows():\n", " for c in row[\"contexts\"]:\n", " if c not in doc_text_to_id:\n", " doc_text_to_id[c] = doc_id_counter\n", " doc_id_counter += 1" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "##### Populate documents, tasks, and evaluations" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [], "source": [ "all_documents = []\n", "all_tasks = []\n", "all_evaluations = []\n", "\n", "# Populate documents\n", "for doc_text, doc_id in doc_text_to_id.items():\n", " document = {\n", " \"document_id\": f\"{doc_id}\",\n", " \"text\": f\"{doc_text}\"\n", " }\n", " all_documents.append(document)\n", "\n", "# Populate taks and evaluations\n", "for index, row in df_score.iterrows():\n", " instance = {\n", " \"task_id\": f\"{index}\",\n", " \"task_type\": \"rag\",\n", " \"contexts\": [ {\"document_id\": f\"{doc_text_to_id[c]}\"} for c in row[\"contexts\"] ],\n", " \"input\": [{\"speaker\": \"user\", \"text\": f\"{row['question']}\"}],\n", " \"targets\": [{\"text\": f\"{row['ground_truth']}\"}]\n", " }\n", " all_tasks.append(instance)\n", "\n", " evaluation = {\n", " \"task_id\": f\"{index}\",\n", " \"model_id\": f\"{models[0]['model_id']}\",\n", " \"model_response\": f\"{row['answer']}\",\n", " \"annotations\": {}\n", " }\n", " for metric in all_metrics:\n", " metric_name = metric[\"name\"]\n", " evaluation[\"annotations\"][metric_name] = {\n", " \"system\": {\n", " \"value\": row[metric_name],\n", " \"duration\": 0\n", " }\n", " }\n", " all_evaluations.append(evaluation)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "##### Write final json file to `ragas-inspectorraget-demo.json` in working directory" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [], "source": [ "import json\n", "\n", "output = {\n", " \"name\": name,\n", " \"models\": models,\n", " \"metrics\": all_metrics,\n", " \"documents\": all_documents,\n", " \"tasks\": all_tasks,\n", " \"evaluations\": all_evaluations,\n", "}\n", "\n", "with open(\n", " file=\"ragas-inspectorraget-demo.json\", mode=\"w\", encoding=\"utf-8\"\n", ") as fp:\n", " json.dump(output, fp, indent=4)" ] } ], "metadata": { "kernelspec": { "display_name": "bin", "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": "3.9.12" } }, "nbformat": 4, "nbformat_minor": 2 }