Golden dataset notebook

notebooks/create_golden_dataset.ipynb

@@ -0,0 +1,749 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "vscode": {
+     "languageId": "plaintext"
+    }
+   },
+   "source": [
+    "# Create a golden dataset using RAGAS"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 127,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os\n",
+    "import getpass\n",
+    "from dotenv import load_dotenv\n",
+    "import getpass\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 128,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "\n",
+    "load_dotenv()\n",
+    "os.environ[\"LANGCHAIN_TRACING_V2\"] = \"true\"\n",
+    "\n",
+    "def set_api_key_if_not_present(key_name, prompt_message):\n",
+    "    if key_name not in os.environ or not os.environ[key_name]:\n",
+    "        os.environ[key_name] = getpass.getpass(prompt_message)\n",
+    "\n",
+    "set_api_key_if_not_present(\"OPENAI_API_KEY\", \"OpenAI API Key:\")\n",
+    "set_api_key_if_not_present(\"TAVILY_API_KEY\", \"TAVILY_API_KEY:\")\n",
+    "set_api_key_if_not_present(\"LANGCHAIN_API_KEY\", \"LANGCHAIN_API_KEY:\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Data loading"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "First, we're going to load all of our transcripts in."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 145,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "pstuts_transcripts_test_dev\n"
+     ]
+    }
+   ],
+   "source": [
+    "from ast import Dict\n",
+    "import json\n",
+    "\n",
+    "from pstuts_rag.loader import load_json_files\n",
+    "filenames = [\"../data/test.json\",\"../data/dev.json\", \"../data/train.json\"]\n",
+    "filenames = [\"../data/test.json\",\"../data/dev.json\"]\n",
+    "\n",
+    "from typing import List, Dict, Any\n",
+    "data:List[Dict[str,Any]] = await load_json_files(filenames)\n",
+    "\n",
+    "from pathlib import Path\n",
+    "\n",
+    "data_name = \"pstuts_transcripts_\"+\"_\".join(Path(path).stem for path in filenames )\n",
+    "    \n",
+    "print(data_name)\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "The following are all data keys. `group` indicates the filename the transcript was loaded from.\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 146,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Number of files: 22\n",
+      "File data fields: dict_keys(['video_id', 'title', 'desc', 'length', 'url', 'transcripts', 'qa', 'group'])\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(f\"Number of files: {len(data)}\")\n",
+    "print(f\"File data fields: {data[0].keys()}\" )\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "In each file, `transcripts` field is a list of transcript chunks"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 147,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Transcripts in first file: 58\n",
+      "Transcript keys: dict_keys(['sent_id', 'sent', 'begin', 'end'])\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(f\"Transcripts in first file: {len(data[0][\"transcripts\"])}\")\n",
+    "print(f\"Transcript keys: {data[0][\"transcripts\"][0].keys()}\" )"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Now, we will load the documents from transcripts.\n",
+    "\n",
+    "In this case, we are going to use the loader that loads 1 document per video.\n",
+    "(As opposed to the `VideoTranscriptChunkLoader` that loads 1 doc per chunk)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 148,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "# of documents: 22. # of videos: 22\n"
+     ]
+    }
+   ],
+   "source": [
+    "from pstuts_rag.loader import VideoTranscriptBulkLoader\n",
+    "\n",
+    "\n",
+    "loader = VideoTranscriptBulkLoader(json_payload=data)\n",
+    "docs = loader.load()\n",
+    "\n",
+    "print(f\"# of documents: {len(docs)}. # of videos: {len(data)}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Building the knowledge graph"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 149,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from ragas.llms import LangchainLLMWrapper\n",
+    "from ragas.embeddings import LangchainEmbeddingsWrapper\n",
+    "from langchain_openai import ChatOpenAI\n",
+    "from langchain_openai import OpenAIEmbeddings\n",
+    "from ragas.testset.graph import KnowledgeGraph\n",
+    "from ragas.testset.graph import Node, NodeType\n",
+    "from ragas.testset.transforms import default_transforms, apply_transforms\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 150,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "generator_llm = LangchainLLMWrapper(ChatOpenAI(model=\"gpt-4.1-mini\"))\n",
+    "generator_embeddings = LangchainEmbeddingsWrapper(OpenAIEmbeddings())\n",
+    "\n",
+    "transformer_llm = generator_llm\n",
+    "embedding_model = generator_embeddings\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 151,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "root = Path(\"../data\")\n",
+    "kg_filename = Path(f\"kg_{data_name}.json\")\n",
+    "kg_path = root.joinpath(kg_filename)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 152,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "../data/kg_pstuts_transcripts_test_dev.json does not contain a knowledge graph. Generating.\n",
+      "Initial size KnowledgeGraph(nodes: 22, relationships: 0)\n"
+     ]
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "a3a03e8639fb4148add8bd1997bb7e71",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Applying HeadlinesExtractor:   0%|          | 0/21 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "cd4c56f29d35411f9d8980bb2d1eda2d",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Applying HeadlineSplitter:   0%|          | 0/22 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "unable to apply transformation: 'headlines' property not found in this node\n"
+     ]
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "30b9923358a7484ba10ebc29c4416f9c",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Applying SummaryExtractor:   0%|          | 0/37 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Property 'summary' already exists in node '9c97f4'. Skipping!\n",
+      "Property 'summary' already exists in node '8a8e28'. Skipping!\n",
+      "Property 'summary' already exists in node 'dcd67c'. Skipping!\n",
+      "Property 'summary' already exists in node '876b13'. Skipping!\n",
+      "Property 'summary' already exists in node '8898c4'. Skipping!\n",
+      "Property 'summary' already exists in node '3f46e3'. Skipping!\n",
+      "Property 'summary' already exists in node '118e6c'. Skipping!\n",
+      "Property 'summary' already exists in node '9c28dd'. Skipping!\n",
+      "Property 'summary' already exists in node 'd9cbc0'. Skipping!\n",
+      "Property 'summary' already exists in node 'a0f2e0'. Skipping!\n",
+      "Property 'summary' already exists in node '8679a1'. Skipping!\n",
+      "Property 'summary' already exists in node '5403e1'. Skipping!\n",
+      "Property 'summary' already exists in node 'a71e28'. Skipping!\n",
+      "Property 'summary' already exists in node '767485'. Skipping!\n",
+      "Property 'summary' already exists in node '2b676d'. Skipping!\n",
+      "Property 'summary' already exists in node '4dd9a1'. Skipping!\n"
+     ]
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "cff631ccbbb9417ea47c879f39573e5a",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Applying CustomNodeFilter:   0%|          | 0/12 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "8c8dbe1d10054d8ea12c52eefc3ca12e",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Applying [EmbeddingExtractor, ThemesExtractor, NERExtractor]:   0%|          | 0/57 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Property 'summary_embedding' already exists in node '9c97f4'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node 'dcd67c'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '876b13'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node 'a0f2e0'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node 'd9cbc0'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '118e6c'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '8898c4'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '8679a1'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '8a8e28'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '3f46e3'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '9c28dd'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '5403e1'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node 'a71e28'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '767485'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '2b676d'. Skipping!\n",
+      "Property 'summary_embedding' already exists in node '4dd9a1'. Skipping!\n"
+     ]
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "f97f574c4fb047d99e37531c6a97c4c5",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Applying [CosineSimilarityBuilder, OverlapScoreBuilder]:   0%|          | 0/2 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "After transformations size KnowledgeGraph(nodes: 48, relationships: 695)\n",
+      "Saved to ../data/kg_pstuts_transcripts_test_dev.json.\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "KnowledgeGraph(nodes: 48, relationships: 695)"
+      ]
+     },
+     "execution_count": 152,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "kg = KnowledgeGraph()\n",
+    "\n",
+    "try:\n",
+    "    kg = kg.load(kg_path)\n",
+    "    print(f\"Loaded from {kg_path}.\")\n",
+    "except:\n",
+    "    print(f\"{kg_path} does not contain a knowledge graph. Generating.\")\n",
+    "    for doc in docs:\n",
+    "        kg.nodes.append(\n",
+    "            Node(\n",
+    "                type=NodeType.DOCUMENT,\n",
+    "                properties={\"page_content\": doc.page_content, \n",
+    "                            \"document_metadata\": doc.metadata}\n",
+    "            )\n",
+    "        )\n",
+    "    print(f\"Initial size {str(kg)}\")\n",
+    "    default_transforms = default_transforms(documents=docs, \n",
+    "                                            llm=transformer_llm, \n",
+    "                                            embedding_model=embedding_model)\n",
+    "    apply_transforms(kg, default_transforms)\n",
+    "    print(f\"After transformations size {str(kg)}\")\n",
+    "    kg.save(kg_path)\n",
+    "    print(f\"Saved to {kg_path}.\")\n",
+    "    \n",
+    "kg"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Test set generator"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 153,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from ragas.testset import TestsetGenerator\n",
+    "\n",
+    "personas = [\n",
+    "    Persona(\n",
+    "    name=\"Beginner Photoshop User\",\n",
+    "    role_description=(\"Beginner Photoshop user, learning to complete \"\n",
+    "                      \"simple tasks, use the tools in Photoshop \"\n",
+    "                      \"and navigate the graphical user interface\"),\n",
+    "),\n",
+    "    Persona(\n",
+    "    name=\"Photoshop trainer\",\n",
+    "    role_description=(\"Experienced trainer in Photoshop. Looking to develop\"\n",
+    "                      \"step-by-step guides for Photoshop beginners\"),\n",
+    ")\n",
+    "]\n",
+    "\n",
+    "generator = TestsetGenerator(llm=generator_llm, \n",
+    "                             embedding_model=embedding_model, \n",
+    "                             persona_list=personas,\n",
+    "                             knowledge_graph=kg)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 154,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from ragas.testset.synthesizers import default_query_distribution, SingleHopSpecificQuerySynthesizer, MultiHopAbstractQuerySynthesizer, MultiHopSpecificQuerySynthesizer\n",
+    "from ragas.testset.persona import Persona\n",
+    "query_distribution = [\n",
+    "        (SingleHopSpecificQuerySynthesizer(llm=generator_llm), 0.8),\n",
+    "        (MultiHopAbstractQuerySynthesizer(llm=generator_llm), 0.1),\n",
+    "        (MultiHopSpecificQuerySynthesizer(llm=generator_llm), 0.1),\n",
+    "]\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 159,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "655a6981a810458bb5f84e8df4d4da77",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Generating Scenarios:   0%|          | 0/3 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "a7007c4af6b641da889e286e03d51415",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Batch 1/1:   0%|          | 0/3 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "37b7421994104b0ebfbf593d16778425",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Generating Samples:   0%|          | 0/100 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "e118cb7a6eae4ba784738a5b076b16fd",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Batch 1/13:   0%|          | 0/8 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "\n",
+    "testset = generator.generate(\n",
+    "    testset_size=100, \n",
+    "    batch_size=8,\n",
+    "    num_personas=len(personas),\n",
+    "    query_distribution=query_distribution)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 160,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>user_input</th>\n",
+       "      <th>reference</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>How I can use Move tool to move many layers at...</td>\n",
+       "      <td>If you have the Move tool selected in Photosho...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>How I can use Windows key for select layers in...</td>\n",
+       "      <td>In Photoshop, when selecting layers to put int...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>How I select layers in Windows for group in Ph...</td>\n",
+       "      <td>In Windows, to select layers for grouping in P...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>how i make group in adobe photoshop to reduce ...</td>\n",
+       "      <td>In Adobe Photoshop, you can reduce clutter in ...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>Wut is Group 1 in Photoshop and how do I use it?</td>\n",
+       "      <td>Group 1 in Photoshop is a folder created by cl...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>...</th>\n",
+       "      <td>...</td>\n",
+       "      <td>...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>95</th>\n",
+       "      <td>How do you use the Rectangular Marquee tool to...</td>\n",
+       "      <td>To use the Rectangular Marquee tool to make a ...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>96</th>\n",
+       "      <td>How do you use the Select menu to deselect a s...</td>\n",
+       "      <td>After making a selection with the Rectangular ...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>97</th>\n",
+       "      <td>In Photoshop CC, how can a beginner add extra ...</td>\n",
+       "      <td>In Photoshop CC, to add extra canvas space to ...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>98</th>\n",
+       "      <td>How can I add extra pixels to just one side of...</td>\n",
+       "      <td>To add extra pixels to just one side of an ima...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>99</th>\n",
+       "      <td>How can a beginner Photoshop user use the Comm...</td>\n",
+       "      <td>A beginner Photoshop user can use the Command ...</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>100 rows × 2 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                                           user_input  \\\n",
+       "0   How I can use Move tool to move many layers at...   \n",
+       "1   How I can use Windows key for select layers in...   \n",
+       "2   How I select layers in Windows for group in Ph...   \n",
+       "3   how i make group in adobe photoshop to reduce ...   \n",
+       "4    Wut is Group 1 in Photoshop and how do I use it?   \n",
+       "..                                                ...   \n",
+       "95  How do you use the Rectangular Marquee tool to...   \n",
+       "96  How do you use the Select menu to deselect a s...   \n",
+       "97  In Photoshop CC, how can a beginner add extra ...   \n",
+       "98  How can I add extra pixels to just one side of...   \n",
+       "99  How can a beginner Photoshop user use the Comm...   \n",
+       "\n",
+       "                                            reference  \n",
+       "0   If you have the Move tool selected in Photosho...  \n",
+       "1   In Photoshop, when selecting layers to put int...  \n",
+       "2   In Windows, to select layers for grouping in P...  \n",
+       "3   In Adobe Photoshop, you can reduce clutter in ...  \n",
+       "4   Group 1 in Photoshop is a folder created by cl...  \n",
+       "..                                                ...  \n",
+       "95  To use the Rectangular Marquee tool to make a ...  \n",
+       "96  After making a selection with the Rectangular ...  \n",
+       "97  In Photoshop CC, to add extra canvas space to ...  \n",
+       "98  To add extra pixels to just one side of an ima...  \n",
+       "99  A beginner Photoshop user can use the Command ...  \n",
+       "\n",
+       "[100 rows x 2 columns]"
+      ]
+     },
+     "execution_count": 160,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "testset.to_pandas()[[\"user_input\", \"reference\"]]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 161,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Testset uploaded! View at https://app.ragas.io/dashboard/alignment/testset/94d276d0-95aa-4305-839c-c846af514e2e\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "'https://app.ragas.io/dashboard/alignment/testset/94d276d0-95aa-4305-839c-c846af514e2e'"
+      ]
+     },
+     "execution_count": 161,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "testset.upload()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 162,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "KnowledgeGraph(nodes: 48, relationships: 695)"
+      ]
+     },
+     "execution_count": 162,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "kg"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "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": "3.13.2"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

pyproject.toml

@@ -38,6 +38,7 @@ dependencies = [
     "pylint-venv>=3.0.4",
     "pyppeteer>=0.0.25",
     "grandalf>=0.8",
+    "jupyter-contrib-nbextensions>=0.7.0",
 ]
 authors = [{ name = "Marko Budisic", email = "[email protected]" }]
 license = "MIT"