L4-student.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "f057401b-a4ee-462a-9ba9-3e342e06eb1a",
   "metadata": {},
   "source": [
    "## Lab 4 - Cross-encoder re-ranking"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "5a5536f0-651c-40e7-aa15-27ee0cda80b7",
   "metadata": {
    "height": 64
   },
   "outputs": [],
   "source": [
    "from helper_utils import load_chroma, word_wrap, project_embeddings\n",
    "from chromadb.utils.embedding_functions import SentenceTransformerEmbeddingFunction\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3748b16d-d4a7-49c3-a48a-57dcfc42acd6",
   "metadata": {
    "height": 81
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4c5604447a1b4707b8320ce4f37e6ed6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       ".gitattributes:   0%|          | 0.00/1.23k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "67b7a3e940c44c85934eb1f86ba7cfc6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "1_Pooling/config.json:   0%|          | 0.00/190 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0449ff172fd24092b4c28f65a000adf5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "README.md:   0%|          | 0.00/10.7k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4c789f380a864726982cb4a6db2d509a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "config.json:   0%|          | 0.00/612 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "73f51fa1f2e34f05bc05dadde11b381b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "config_sentence_transformers.json:   0%|          | 0.00/116 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7de1609a2de84985a8a411446d2c976d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "data_config.json:   0%|          | 0.00/39.3k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a341ce1f1f914b808edc977cf9571194",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "model.safetensors:   0%|          | 0.00/90.9M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6fa80f40fae84aaf9b3a31b6a7e16759",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "model.onnx:   0%|          | 0.00/90.4M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ea19946d77b549b891e8449e53e9612c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "pytorch_model.bin:   0%|          | 0.00/90.9M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8c708f98f4f0443eb01556bb1ff49ee0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "sentence_bert_config.json:   0%|          | 0.00/53.0 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b8e51d9eac864836b785cdde091ea819",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "special_tokens_map.json:   0%|          | 0.00/112 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a84b201bdd9842fc9041a418697f0f53",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "tokenizer.json:   0%|          | 0.00/466k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6f7847cbad7a4f359e3f336f0e76d814",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "tokenizer_config.json:   0%|          | 0.00/350 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "002121c5ec4c467790a265402bbdcdf0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "train_script.py:   0%|          | 0.00/13.2k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5cc20a2184aa4e5ca844cc6b834a3cdd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "vocab.txt:   0%|          | 0.00/232k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fc209dc571db4534ad0bf746b85a0486",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "modules.json:   0%|          | 0.00/349 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "embedding_function = SentenceTransformerEmbeddingFunction()\n",
    "\n",
    "chroma_collection = load_chroma(filename='microsoft_annual_report_2022.pdf', collection_name='microsoft_annual_report_2022', embedding_function=embedding_function)\n",
    "chroma_collection.count()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68486bf8-37dd-4257-a23b-9ef50c47bcc5",
   "metadata": {},
   "source": [
    "# Re-ranking the long tail"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "888a86f8-2fe2-4682-bdaf-c15129ed1a32",
   "metadata": {
    "height": 149
   },
   "outputs": [],
   "source": [
    "query = \"What has been the investment in research and development?\"\n",
    "results = chroma_collection.query(query_texts=query, n_results=10, include=['documents', 'embeddings'])\n",
    "\n",
    "retrieved_documents = results['documents'][0]\n",
    "\n",
    "for document in results['documents'][0]:\n",
    "    print(word_wrap(document))\n",
    "    print('')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5665c695-22ea-4264-b1ac-5ba720b6d78b",
   "metadata": {
    "height": 47
   },
   "outputs": [],
   "source": [
    "from sentence_transformers import CrossEncoder\n",
    "cross_encoder = CrossEncoder('cross-encoder/ms-marco-MiniLM-L-6-v2')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c2a13d14-4484-46f0-8e67-277337f9d138",
   "metadata": {
    "height": 98
   },
   "outputs": [],
   "source": [
    "pairs = [[query, doc] for doc in retrieved_documents]\n",
    "scores = cross_encoder.predict(pairs)\n",
    "print(\"Scores:\")\n",
    "for score in scores:\n",
    "    print(score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8ba6c8c5-9ce4-44d0-9223-6fdd77871f87",
   "metadata": {
    "height": 64
   },
   "outputs": [],
   "source": [
    "print(\"New Ordering:\")\n",
    "for o in np.argsort(scores)[::-1]:\n",
    "    print(o+1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f54cc00-eebc-4294-91bf-1a2cdce51708",
   "metadata": {},
   "source": [
    "# Re-ranking with Query Expansion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "377a84aa-1d93-4e97-9b2d-d59c46355338",
   "metadata": {
    "height": 149
   },
   "outputs": [],
   "source": [
    "original_query = \"What were the most important factors that contributed to increases in revenue?\"\n",
    "generated_queries = [\n",
    "    \"What were the major drivers of revenue growth?\",\n",
    "    \"Were there any new product launches that contributed to the increase in revenue?\",\n",
    "    \"Did any changes in pricing or promotions impact the revenue growth?\",\n",
    "    \"What were the key market trends that facilitated the increase in revenue?\",\n",
    "    \"Did any acquisitions or partnerships contribute to the revenue growth?\"\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ba0ed8ca-6640-4c09-9cb3-9de5e7cf46dc",
   "metadata": {
    "height": 81
   },
   "outputs": [],
   "source": [
    "queries = [original_query] + generated_queries\n",
    "\n",
    "results = chroma_collection.query(query_texts=queries, n_results=10, include=['documents', 'embeddings'])\n",
    "retrieved_documents = results['documents']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "28bac3a2-0d29-48dc-9b48-2d9313239a25",
   "metadata": {
    "height": 132
   },
   "outputs": [],
   "source": [
    "# Deduplicate the retrieved documents\n",
    "unique_documents = set()\n",
    "for documents in retrieved_documents:\n",
    "    for document in documents:\n",
    "        unique_documents.add(document)\n",
    "\n",
    "unique_documents = list(unique_documents)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "db9f2758-0f5a-49e5-b1fa-517b91324575",
   "metadata": {
    "height": 64
   },
   "outputs": [],
   "source": [
    "pairs = []\n",
    "for doc in unique_documents:\n",
    "    pairs.append([original_query, doc])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aee59493-8a99-4da8-b94f-4747efcfc79d",
   "metadata": {
    "height": 47
   },
   "outputs": [],
   "source": [
    "scores = cross_encoder.predict(pairs)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f5eda9bc-ae76-4db6-9e0c-ae099d852d78",
   "metadata": {
    "height": 64
   },
   "outputs": [],
   "source": [
    "print(\"Scores:\")\n",
    "for score in scores:\n",
    "    print(score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b1183e75-4c65-422e-bc47-48010d8b29c9",
   "metadata": {
    "height": 64
   },
   "outputs": [],
   "source": [
    "print(\"New Ordering:\")\n",
    "for o in np.argsort(scores)[::-1]:\n",
    "    print(o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5fcd85cc-8898-41ed-a0aa-bd8a33fc565a",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c65337e9-85ee-47f7-89fd-7fe77cd0e1b2",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a7794092-4195-4cf3-9eab-11c9c05a26b9",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f2cab7a1-1be7-45f0-83b7-543e48f83901",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fe0343be-73c9-4aed-83b0-aba09569ac87",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6f0f3e33-e517-4f6b-8b38-c47c1e3d40b4",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a16fdcb1-57d0-4f04-af8f-7c7fc594d947",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "babe7893-9cbc-43c5-94ef-cbf8f5d68cf2",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "60a9524b-1085-4bdf-a161-39f11397dc1f",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d189f088-b58e-4583-9590-afdfa624cf87",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7b26a01a-4575-446b-b8dc-a8c5ab153172",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0950575b-b69d-46a3-8c91-c7af89f5c204",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3f123ad8-b2e8-4a25-8b42-a520ecaf566b",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "83c04587-d1de-419c-a213-2e3eb67dc33d",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d3155972-824e-4ebe-a692-2227c113c5a8",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d8144a4a-85f6-4800-87f9-36a1b6ceda1f",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8ff0b18e-12a0-4ac0-97dd-8618b22e7dbf",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "03ca7e7c-4b47-4652-9b46-a40b3dffa5e6",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f74e7d67-7f51-41c4-8e25-edbaa02d0bd8",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9188e886-d406-406f-b234-f5c3353a77a2",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3d3bb286-2694-4ed4-8466-46865e997ced",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2876084b-4038-4b0c-8ec8-8294a86adfc1",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3ac542e1-b094-431f-9611-cf7e36d3f0de",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bcd6114b-c09d-4173-a623-9a08aaf63e4b",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ad10ab65-b351-4f4b-b7d2-63474acfb9f9",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "800f3d81-cbdb-4ba4-8d49-85747fdfded8",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "37847448-c9f6-4f51-bf06-f7809964a8b2",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2dcefc87-0964-4b94-946b-2145781ad606",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5fc994bc-7b1e-476a-9df9-300a3e374882",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9ef5f5d5-acb7-4b0a-93ef-e61306708e69",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "44e4b33f-d8fb-4f3a-b884-8b43a3766583",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c2a480a2-2c29-4a01-80dd-ee41934b7901",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8127c2bf-0d15-4b62-b46a-f7a17ad2ec92",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "18ded129-a637-4269-a116-550fe9a90570",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c1d7ee44-7b29-483f-a3f2-cc9d8e18880e",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8e450dd8-9719-42c6-8c3c-33cac910e0a5",
   "metadata": {
    "height": 30
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.9.19"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}