fruits_(embeddings).ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2d_RMwa4CvWk",
        "outputId": "4a1786fe-3227-4498-bbb2-daaf7daf4f5c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting cohere==4.4.1\n",
            "  Downloading cohere-4.4.1-py3-none-any.whl (32 kB)\n",
            "Requirement already satisfied: aiohttp<4.0,>=3.0 in /usr/local/lib/python3.10/dist-packages (from cohere==4.4.1) (3.9.3)\n",
            "Collecting backoff<3.0,>=2.0 (from cohere==4.4.1)\n",
            "  Downloading backoff-2.2.1-py3-none-any.whl (15 kB)\n",
            "Requirement already satisfied: requests<3.0,>=2.0 in /usr/local/lib/python3.10/dist-packages (from cohere==4.4.1) (2.31.0)\n",
            "Requirement already satisfied: aiosignal>=1.1.2 in /usr/local/lib/python3.10/dist-packages (from aiohttp<4.0,>=3.0->cohere==4.4.1) (1.3.1)\n",
            "Requirement already satisfied: attrs>=17.3.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp<4.0,>=3.0->cohere==4.4.1) (23.2.0)\n",
            "Requirement already satisfied: frozenlist>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from aiohttp<4.0,>=3.0->cohere==4.4.1) (1.4.1)\n",
            "Requirement already satisfied: multidict<7.0,>=4.5 in /usr/local/lib/python3.10/dist-packages (from aiohttp<4.0,>=3.0->cohere==4.4.1) (6.0.5)\n",
            "Requirement already satisfied: yarl<2.0,>=1.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp<4.0,>=3.0->cohere==4.4.1) (1.9.4)\n",
            "Requirement already satisfied: async-timeout<5.0,>=4.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp<4.0,>=3.0->cohere==4.4.1) (4.0.3)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests<3.0,>=2.0->cohere==4.4.1) (3.3.2)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests<3.0,>=2.0->cohere==4.4.1) (3.6)\n",
            "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests<3.0,>=2.0->cohere==4.4.1) (2.0.7)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests<3.0,>=2.0->cohere==4.4.1) (2024.2.2)\n",
            "Installing collected packages: backoff, cohere\n",
            "\u001b[31mERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.\n",
            "llmx 0.0.15a0 requires openai, which is not installed.\n",
            "llmx 0.0.15a0 requires tiktoken, which is not installed.\u001b[0m\u001b[31m\n",
            "\u001b[0mSuccessfully installed backoff-2.2.1 cohere-4.4.1\n"
          ]
        }
      ],
      "source": [
        "!pip install cohere==4.4.1"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Zg5H7gd7V0Mu"
      },
      "outputs": [],
      "source": [
        "import cohere\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "\n",
        "api_key = '' # Paste your API key here. This is a trial key. These are free from Cohere.\n",
        "co = cohere.Client(api_key)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4_YQVuAbf3Xh"
      },
      "source": [
        "## Turn text into embeddings"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6HujlbFsIZjW"
      },
      "outputs": [],
      "source": [
        "oranges = np.array(co.embed([\"oranges\"], model=\"large\").embeddings)\n",
        "lemons = np.array(co.embed([\"lemons\"], model=\"large\").embeddings)\n",
        "limes = np.array(co.embed([\"limes\"], model=\"large\").embeddings)\n",
        "key_limes = np.array(co.embed([\"key_limes\"], model=\"large\").embeddings)\n",
        "tangerines = np.array(co.embed([\"tangerines\"], model=\"large\").embeddings)\n",
        "pineapples = np.array(co.embed([\"pineapples\"], model=\"large\").embeddings)\n",
        "mangoes = np.array(co.embed([\"mangoes\"], model=\"large\").embeddings)\n",
        "papayas = np.array(co.embed([\"papayas\"], model=\"large\").embeddings)\n",
        "kiwis = np.array(co.embed([\"kiwis\"], model=\"large\").embeddings)\n",
        "guavas = np.array(co.embed([\"guavas\"], model=\"large\").embeddings)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8UK87ZVm1pLd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b599b7a8-7440-44bc-88e1-52eb97e516e0"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[-1.6855469 , -2.0820312 ,  2.1914062 , ...,  1.6699219 ,\n",
              "         0.5576172 ,  0.19299316]])"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ],
      "source": [
        "oranges"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "as-OZGdGn4b5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f90e9e9d-9978-410f-b6fe-b90fa144c4d7"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "4096"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ],
      "source": [
        "len(oranges[0])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9pR9yDy1ot17"
      },
      "source": [
        "## Calculate dot products"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cAds6sY19zsx"
      },
      "source": [
        "- lemons\n",
        "- limes\n",
        "- kiwis"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8b-FViGlJQLe",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2efda072-243d-4f76-e047-ec3e8fbeefc9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Similarity between lemons and limes: 6950.971011710139\n",
            "Similarity between lemons and kiwis: 4482.204396531939\n",
            "Similarity between limes and kiwis: 4508.959219320356\n"
          ]
        }
      ],
      "source": [
        "from numpy import dot\n",
        "print(\"Similarity between lemons and limes:\", dot(lemons[0], limes[0]))\n",
        "print(\"Similarity between lemons and kiwis:\", dot(lemons[0], kiwis[0]))\n",
        "print(\"Similarity between limes and kiwis:\", dot(limes[0], kiwis[0]))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DWEpWsI0o8gn"
      },
      "source": [
        "## Calculate cosine similarities"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "D4ZkpwMSL_oH",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3992d5ce-bf3a-49a0-bf62-92b7471b7cfc"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Cosine similarity between lemons and limes: 0.7707548944485144\n",
            "Cosine similarity between lemons and kiwis: 0.506334924674642\n",
            "Cosine similarity between limes and kiwis: 0.5065402051392838\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics.pairwise import cosine_similarity\n",
        "print(\"Cosine similarity between lemons and limes:\", cosine_similarity(lemons, limes)[0][0])\n",
        "print(\"Cosine similarity between lemons and kiwis:\", cosine_similarity(lemons, kiwis)[0][0])\n",
        "print(\"Cosine similarity between limes and kiwis:\", cosine_similarity(limes, kiwis)[0][0])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "tZ7ls1JlkngY",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "629fb2ce-79b0-4c24-d9c4-87e216b4862a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Cosine similarity between lemons and lemons: 1.0\n",
            "Cosine similarity between limes and limes: 1.0\n",
            "Cosine similarity between kiwis and kiwis: 0.9999999999999997\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics.pairwise import cosine_similarity\n",
        "print(\"Cosine similarity between lemons and lemons:\", cosine_similarity(lemons, lemons)[0][0])\n",
        "print(\"Cosine similarity between limes and limes:\", cosine_similarity(limes, limes)[0][0])\n",
        "print(\"Cosine similarity between kiwis and kiwis:\", cosine_similarity(kiwis, kiwis)[0][0])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "sO3AxCUJq7AH",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 484
        },
        "outputId": "014a3ac5-97d7-443b-cefb-00b823bd9fc3"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[Text(0.5, 0, 'lemons'), Text(1.5, 0, 'limes'), Text(2.5, 0, 'kiwis')]"
            ]
          },
          "metadata": {},
          "execution_count": 9
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "embeddings = [lemons, limes, kiwis]\n",
        "dot_product_similarities = [[cosine_similarity(embeddings[i], embeddings[j])[0][0] for i in range(len(embeddings))] for j in range(len(embeddings))]\n",
        "ax = sns.heatmap(dot_product_similarities, vmin=0, vmax=1,cmap=\"Blues\",\n",
        "                 linewidths=1, linecolor='grey',\n",
        "                 xticklabels=[\"lemons\", \"limes\", \"kiwis\"],\n",
        "                 yticklabels=[\"lemons\", \"limes\", \"kiwis\"],\n",
        ")\n",
        "ax.set_xticklabels(ax.get_xticklabels(), rotation=45)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Kd6LZ6jyVkmj"
      },
      "source": [
        "## Calculate cosine similarities for sentences"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "nVI6pT9MTkP1"
      },
      "outputs": [],
      "source": [
        "lemons = \"Lemons are rich in vitamin C\"\n",
        "limes = \"Limes are tangy and acidic\"\n",
        "jordan = \"Michael Jordan played for the Chicago bulls\"\n",
        "\n",
        "lemons_emb = np.array(co.embed([lemons], model=\"large\").embeddings)\n",
        "limes_emb = np.array(co.embed([limes], model=\"large\").embeddings)\n",
        "jordan_emb = np.array(co.embed([jordan], model=\"large\").embeddings)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4Bjg8-SqoEuI",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0396835a-45be-4bea-c992-e10fc43c55b6"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "4096"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ],
      "source": [
        "len(lemons_emb[0])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "84LZgOlqTjQC",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 696
        },
        "outputId": "de1a4921-d6d8-4fbd-f1bc-32e988ba14f9"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[Text(0.5, 0, 'Lemons are rich in vitamin C'),\n",
              " Text(1.5, 0, 'Limes are tangy and acidic'),\n",
              " Text(2.5, 0, 'Michael Jordan played for the Chicago bulls')]"
            ]
          },
          "metadata": {},
          "execution_count": 12
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "embeddings = [lemons_emb, limes_emb, jordan_emb]\n",
        "dot_product_similarities = [[cosine_similarity(embeddings[i], embeddings[j])[0][0] for i in range(len(embeddings))] for j in range(len(embeddings))]\n",
        "ax = sns.heatmap(dot_product_similarities, vmin=0, vmax=1,cmap=\"Blues\",\n",
        "                 linewidths=1, linecolor='grey',\n",
        "                 xticklabels=[lemons, limes, jordan],\n",
        "                 yticklabels=[lemons, limes, jordan],\n",
        ")\n",
        "ax.set_xticklabels(ax.get_xticklabels(), rotation=45)"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python",
      "version": "3.10.0"
    },
    "vscode": {
      "interpreter": {
        "hash": "1fb8019e3560b882083e525615cf48e713d3a7345a15eb723d805e91aa410aac"
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}